Sorry for interrupting, this website uses cookies to improve your experience. We'll assume you're ok with this, but you can opt-out if you wish. Learn more
Accept & Close

Methodologies

US

Mintel is an independent market analysis company that prides itself on supplying objective information on a whole range of markets and marketing issues.

There are five main sources of research that are used in the compilation of Mintel reports:

  • Consumer research
  • Social media research
  • Desk research
  • Trade research
  • Statistical forecasting

Reports are written and managed by analysts with experience in the relevant markets.

Mintel analyzes and interprets data from a variety of sources. Sources are identified below each Figure, table and graph. Data sourced as 'Mintel' are derived from multiple sources, then interpreted and expanded by Mintel analysts. When referenced as 'estimated' the information is either not finalized in the original source or has been extrapolated by Mintel analysts.


Consumer research

 

In-depth consumer research examines how social, economic, cultural and psychological influences affect attitudes and purchasing decisions. Mintel combines exclusive primary research with syndicated data to provide an accurate and unique analysis. For additional analysis of consumer survey data, or with questions about our research methodology, please contact Mintel at 312.932.0400.

Primary Data Analysis

For each report Mintel develops custom primary research questions and uses specialty research firms for data collection. Mintel uses best in class consumer research strategies to ensure data are of the highest quality.

Sampling

Online surveys

Mintel uses set quotas based on gender, age, household income, and region to ensure that survey samples are proportionally representative of the entire U.S. adult internet population. 

Specific quotas for a sample of 2,000 adults aged 18+ are shown below:

 

 

 

Age groups by gender % N
Male, 18-24 6.5% 130
Male, 25-34 9.5% 189
Male, 35-44 8.7% 175
Male, 45-54 8.7% 173
Male, 55-64 7.9% 157
Male, 65-74 4.6% 92
Male, 75+ 2.3% 47
: : :
: : :
Region % N
     
Northeast  17.9% 357
Midwest 21.4% 427
South  37.1% 742
West 23.7% 474
     
Total* 100 2,000

 

Household income % N
     
Less than $25,000 13.5% 269
$25,000 - $49,999 20.1% 402
$50,000 - $74,999 18.1% 362
$75,000 - $99,999 14.5% 290
$100,000 and over 33.9% 677
     
: : :

 

Children in the household % N
     
Household with children aged 5 and under 12.7% 253
Household with children aged 6-11 12.7% 253
Household with children aged 12-17 12.7% 253
Household with no children 62.0% 1,241
     
Total 100 2,000

To ensure an adequate representation of these groups in our survey results and to allow for more realistic interpretation of our reported findings, Hispanic and African American respondents are over-sampled relative to the overall population.

Race  % N
     
White 74.1% 1,482
Black 15.1% 302
Asian 6.0% 119
Other race 4.9% 97
     
Total 100 2,000

 

Ethnicity % N
     
Hispanic 15.6% 312
Not Hispanic 84.4% 1,688
     
Total 100 2,000

When necessary, Mintel uses specialist panels or targeted sample to capture responses from hard to reach demographic groups or target markets.

Lightspeed GMI

Founded in 1996, Lightspeed GMI’s double opt-in U.S. online consumer panel contains approximately 1.27 million U.S. consumers. Lightspeed GMI recruits its panelists through many different sources including web advertising, permission-based databases and partner-recruited panels.

Secondary Data Analysis

In addition to exclusively commissioned surveys, Mintel gathers syndicated data from the most respected consumer research firms. This allows Mintel analysts to form objective and cohesive analyses of consumer attitudes and behavior.

Simmons National Consumer Studies

Mintel reports frequently draw on the Simmons National Consumer surveys conducted by Simmons Research. The Simmons National Consumer Study (NCS) is a comprehensive survey of American consumers aged 18 and older. It provides single-source measurement of major media, products, services, and in-depth consumer demographic and lifestyle/psychographic characteristics.

  • 25,000 Adults 18+
  • Two-phase data collection
    • Phase 1: A telephone placement interview for a self-administered mail survey is conducted with any adult aged 18 or over in the household
    • Phase 2: Self-administered survey is mailed to each qualified household member 
  • Upfront cash incentives/sweepstakes offer
  • All qualified household members aged 18 or over participate by completing their own personal booklets
  • Principal shopper completes one Household Survey
  • Conducted and released quarterly - Winter, Spring, Summer and Fall
  • Ability to look at full-year and quarterly data

 

The Simmons National Hispanic Study (NHCS) is the only national, multi-media syndicated research instrument targeting the Hispanic market, and is particularly valuable in identifying media habits, product and service usage and attitudes and opinions among this segment.

 

  • 7,500 Hispanic adults 18+
  • Two-phase data collection
    • Phase 1: A telephone placement interview for a self-administered mail survey is conducted with any adult aged 18 or over in the household
    • Phase 2: Self-administered survey is mailed to each qualified household member 
  • Survey offered in Spanish or English - respondent’s choice
  • Incentive/sweepstakes offer
  • Conducted and released quarterly - Winter, Spring, Summer and Fall
  • Ability to look at full-year and quarterly data

 

The samples for the Kids and Teens Studies are taken from the same households participating in the adult study. The Kids and Teens Studies provide in-depth insight into these consumer segments to understand their effect on the marketplace, and how and where to reach them.

 

  • 2,500 Teens 12-17 and 2,600 Kids 6-11
  • Both samples gathered from within NCS participating households
  • Upfront incentive/sweepstakes offer
  • All teens or kids in household participate by completing their own personal booklets
  • Released twice a year—Spring and Fall data releases

 

In some instances Mintel uses Experian’s Mosaic segmentation system to further analyze Simmons NHCS data. Mosaic is a household-based segmentation system, which classifies all U.S. households and neighborhoods into 71 unique Mosaic segments and 19 groupings that share similar demographic and socioeconomic characteristics. Descriptive content is sourced from Simmons NCS/NHCS data.

 

% of U.S. Households

A - Power Elite

4.68

A01 - American Royalty

1.01

A02 - Platinum Prosperity

.79

A03 - Kids and Cabernet

0.69

A04 - Picture Perfect Families

0.79

A05 - Couples with Clout

0.66

A06 - Jet Set Urbanites

0.73

: :
: :

As of the Fall 2015 Simmons National Consumer Study, all of the Mosaic groups and types are listed below:


Qualitative Research

 

Revelation by FocusVision

FocusVision provides Mintel with qualitative bulletin board software. This allows the creation of Internet-based, 'virtual' venues where participants recruited from Mintel's online surveys gather and engage in interactive, text-based discussions led by Mintel moderators.

 

Further Analysis

 

Mintel employs numerous quantitative data analysis techniques to enhance the value of our consumer research. The techniques used vary form one report to another.  Below describes some of the more commonly used techniques. 

 

Repertoire Analysis

 

This is used to create consumer groups based on reported behaviour or attitudes. Consumer responses of the same value (or list of values) across a list of survey items are tallied into a single variable. The repertoire variable summarizes the number of occurrences in which the value or values appear among a list of survey items.  For example, a repertoire of brand purchasing might produce groups of those that purchase 1-2 brands, 3-4 brands and 5 or more brands. Each subgroup should be large enough (ie N=75+) to analyze. 

 

Cluster Analysis

 

This technique assigns a set of individual people in to groups called clusters on the basis of one or more question responses, so that respondents within the same cluster are in some sense closer or more similar to one another than to respondents that were grouped into a different cluster.

 

Correspondence Analysis

 

This is a statistical visualization method for picturing the associations between rows (image, attitudes) and columns (brands, products, segments, etc) of a two-way contingency table. It allows us to display brand images (and/or consumer attitudes towards brands) related to each brand covered in this survey in a joint space that is easy to understand. The significance of the relationship between a brand and its associated image is measured using the Chi-square test. If two brands have similar response patterns regarding their perceived images, they are assigned similar scores on underlying dimensions and will then be displayed close to each other in the perceptual map.

 

CHAID analysis

 

CHAID (Chi-squared Automatic Interaction Detection), a type of decision tree analysis, is used to highlight key target groups in a sample by identifying which sub-groups are more likely to show a particular characteristic. This analysis subdivides the sample into a series of subgroups that share similar characteristics towards a specific response variable and allows us to identify which combinations have the highest response rates for the target variable. It is commonly used to understand and visualize the relationship between a variable of interest such as “interest in trying a new product” and other characteristics of the sample, such as demographic composition.

 

Key Driver Analysis

 

Key driver analysis can be a useful tool in helping to prioritize focus between different factors which may impact key performance indicators (eg satisfaction, likelihood to switch providers, likelihood to recommend a brand, etc). Using correlations analysis or regression analysis provides an understanding of which factors or attributes of a market have the strongest association or “link” with a positive performance on key performance indicators (KPIs). Hence, factors or attributes are identified which are relatively more critical in a market category compared to others and ensures that often limited resources can be allocated to focusing on the main market drivers.

 

TURF Analysis

 

TURF (Total Unduplicated Reach & Frequency) analysis identifies the mix of features, attributes, or messages that will attract the largest number of unique respondents.  It is typically used when the number of features or attributes must be or should be limited, but the goal is still to reach the widest possible audience. By identifying the Total Unduplicated Reach, it is possible to maximize the number of people who find one or more of their preferred features or attributes in the product line. The resulting output from TURF is additive, with each additional feature increasing total reach. The chart is read from left to right, with each arrow indicating the incremental change in total reach when adding a new feature. The final bar represents the maximum reach of the total population when all shown features are offered. 


Social Media Research

 

To complement our exclusive consumer research, Mintel tracks and analyses social media data for inclusion in Mintel reports. Using Infegy’s Atlas software, Mintel 'listens in' on online conversations across a range of social platforms including Facebook, Twitter, consumer forums and the wider web.

 

Atlas provides rich consumer insight via the analysis of commentary posted publicly on the internet. The system performs comprehensive and broad collection of data from millions of internet sources, working to ensure a faithful and extensive sampling of feedback from the widest range of individuals. The dataset contains commentary posted in real time, as well as a substantial archive dating back through 2007.


Trade research

 

Informal

Mintel conducts informal trade research for all reports. This involves contacting key players in the trade not only to gain information concerning their own operations, but also to obtain explanations and views of the strategic issues pertinent to the market being researched in order to address current brand and marketing issues. To ensure accuracy, Mintel sends draft copies of reports to key industry representatives for review, taking their feedback into consideration before publishing the report. Comments, where appropriate, are incorporated into the report.

 

Formal

Internally, Mintel's analysts undertake extensive trade interviews with selected key experts in the field for the majority of reports. The purpose of these interviews is to assess key issues in the market place in order to ensure that any research undertaken takes these into account.

 

In addition, using experienced external researchers, trade research is undertaken for some reports. This takes the form of full trade interview questionnaires and direct quotes are included in the report and analysed by experts in the field. This gives a valuable insight into a range of trade views of topical issues. Direct quotations are included in the reports, giving valuable insight into a range of trade views on topical issues.

 

Desk Research

Mintel has an internal team of market analysts who monitor: government statistics, consumer and trade association statistics, manufacturer sponsored reports, annual company reports and accounts, directories and press articles from around the world and online databases. The latter are extracted from hundreds of publications and websites, both U.S. and overseas. All information is cross-referenced for immediate access. Data from other published sources are the latest available at the time of writing the report. This information is supplemented by an extensive library of Mintel's reports produced since 1972 globally and added to each year by the 500+ reports which are produced annually.

 

In addition to in-house sources, researchers also occasionally use outside libraries or data from Trade Associations. Other information is also gathered from store and exhibition visits across the U.S., as well as using other databases within the Mintel Group, such as the Global New Product Database (GNPD), which monitors FMCG sales promotions.

 

Intelligence gathered through desk research is used to guide research and enrich data findings.


Statistical Forecasting

 

Statistical modelling

For the majority of reports, Mintel produces five-year forecasts based on an advanced statistical technique known as ‘multivariate time series auto-regression’ using the statistical software package SPSS.

 

The model is based on historical market size data taken from Mintel’s own market size database and supplemented by published macroeconomic and demographic data from various private and public sources including the Federal Reserve Board, the U.S. Commerce Department, the Census Bureau, the Council of Economic Advisers, and the Congressional Budget Office.

 

The model searches for relationships between actual market sizes and a selection of relevant and significant macroeconomic and demographic factors (independent variables) to identify those predictors having the most influence on the market.

 

Factors used in a forecast are stated in the relevant report section alongside an interpretation of their role in explaining the development in demand for the product or market in question.

 

Qualitative insight

At Mintel we understand that historic data is limited in its capacity to act as the only force behind the future state of markets. Thus, rich qualitative insights from industry experts regarding past and future events that may impact the market play a crucial role in our post statistical modeling evaluation process.

 

As a result, the Mintel forecast allows for additional factors or market conditions outside of the capacity of the data analysis to impact the market forecast model; using a rigorous statistical process complemented by in-depth market knowledge and expertise.

 

 

   

The Mintel fan chart

 

Forecasts of future economic outcomes are always subject to uncertainty. In order to raise awareness amongst our clients and to illustrate this uncertainty, Mintel has introduced a new way of displaying market size forecasts in the form of a fan-chart.

 

Next to historical market sizes and a current year estimate, the fan chart illustrates the probability of various outcomes for the market value/volume over the next five years.

 

At a 95% confidence interval, we are saying that 95 out of 100 times the forecast will fall within these outer limits, which we call the best and worst case forecasts. These, based on the statistically driven forecast, are the highest (best case) and lowest (worst case) market sizes the market is expected to achieve.

 

Over the next five years, the widening bands successively show the developments that occur within 95%, 90%, 70% and 50% probability intervals. Statistical processes predict the central forecast to fall within the darker shaded area which illustrates 50% probability, i.e. a 5 in 10 chance.

 

In general, based on our current knowledge of given historic market size data as well as projections for key macro- and socio-economic measures that were used to create the forecast, we can assume that in 95% of the time the actual market size will fall within the purple shaded fan. In 5% of all cases this model might not be correct due to random errors and the actual market size will fall out of these boundaries.

 

Weather analogy

To illustrate uncertainty in forecasting in an everyday example, let us assume the following weather forecast was produced based on the meteorologists’ current knowledge of the previous weather condition during the last few days, atmospheric observations, incoming weather fronts etc.

 

 

    

Now, how accurate is this forecast and how certain can we be that the temperature on Saturday will indeed be 15°C?

To state that the temperature in central London on Saturday will rise to exactly 15°C is possible but one can’t be 100% certain about that fact.

To say the temperature on Saturday will be between 13°C and 17°C is a broader statement and much more probable.

In general, we can say that based on the existing statistical model, one can be 95% certain that the temperature on Saturday will be between 13°C and 17°C, and respectively 50% certain it will be between about 14.5°C and 15.5°C. Again, only in 5% of all cases this model might not be correct due to random errors and the actual temperature on Saturday will fall out of these boundaries and thus will be below 13°C or above 17°C.

(To learn more about uncertainty in weather forecasts visit: http://research.metoffice.gov.uk/research/nwp/ensemble/uncertainty.html)

US Multicultural

Mintel is an independent market analysis company that prides itself on supplying objective information on a range of markets and marketing issues.

Five main sources of research are used in the compilation of Mintel multicultural reports:

  • Consumer research
  • Social media research
  • Desk research
  • Trade research
  • Statistical forecasting

 Reports are written and managed by multicultural analysts with experience in the relevant multicultural segment.

Mintel analyzes and interprets data from a variety of sources. Sources are identified below each figure, table and graph. Data sourced as 'Mintel' are derived from multiple sources, then interpreted and expanded by Mintel analysts. When referenced as 'estimated' the information is either not finalized in the original source or has been extrapolated by Mintel analysts.


Consumer research

 

In-depth consumer research examines how social, economic, cultural and psychological influences affect attitudes and purchasing decisions of African Americans, Asian Americans, and Hispanics. Mintel combines exclusive primary research with syndicated data to provide an accurate and unique analysis. For additional analysis of consumer survey data, or with questions about our research methodology, please contact Mintel at 312.932.0400.

Primary Data Analysis

For each report, Mintel develops custom primary research questions which are fielded via specialty research firms.

Sampling

Online surveys

Mintel uses set quotas to ensure an accurate reflection of African-Americans, Asian-Americans, or Hispanics based on that segment’s internet population.

Specific quotas for a sample of 1,000 African-American, 750 Asian American, and 1,000 Hispanic adults aged 18+ are shown below. Please note that Asian Americans constitute a small proportion of the US population, therefore we target and report on fewer Asian-Americans.

African American studies’ quotas

Age groups by gender % N
Male, 18-24 9% 93
Male, 25-34 10% 99
Male, 35-44 9% 89
Male, 45-54 7% 72
Male, 55-64 5% 52
Male, 65+ 4% 39

 

   
: : :
: : :

 

  % N
Region

 

 

     
Northeast  16% 156
Midwest 15% 153
South  58% 581
West 11% 110
     
: : :

 

 Household income % N
     
Less than $25,000 26% 258
$25,000 - $49,999 24% 244
$50,000 - $74,999 16% 162
$75,000 - $99,999 10% 96
$100,000 and over 24% 240
     
: : :

 

Asian American studies’ quotas

Age groups by gender % N
Male, 18-24 10.0% 75
Male, 25-34 9.6% 72
Male, 35-44 9.7% 73
Male, 45-54 9.5% 71
Male, 55-64 4.6% 34
Male, 65+ 4.4% 33

 

   
: : :
: : :

 

Ethnic heritage % N
     
Eastern Asia (Japan, China, Korea, etc.) 46.8% 351
Southeast Asia (Indonesia, Philippines, Thailand, Vietnam, etc.) 31.5% 236
South Asia (India, Pakistan, Bangladesh, etc.) 21.7% 163
     
Total 100 750
Region % N
     
Northeast  24.1% 181
Midwest 11.6% 87
South  24.4% 183
West 39.8% 299
     
Total* 100 750

 

Household income % N
     
Less than $25,000 17.8% 133
$25,000 - $49,999 18.5% 139
$50,000 - $74,999 17.6% 132
$75,000 - $99,999 12.5% 94
$100,000 and over 33.7% 252
     
: : :

 

Hispanic studies’ quotas

Age groups by gender % N
Male, 18-24 11% 105
Male, 25-34 15% 146
Male, 35-44 12% 117
Male, 45-54 7% 75
Male, 55-64 4% 37
Male, 65+ 3% 26

 

   
: : :
: : :

 

Preferred language % N
     
Only English 25% 246
Mostly English but some Spanish 35% 352
Mostly Spanish but some English 23% 232
 Only Spanish  17%  170
     
Total* 100 1,000

 

Region % N
     
Northeast  15% 150
Midwest 9% 93
South  36% 361
West 40% 396
     
Total* 100 1,000

 

Household income % N
     
Less than $25,000 17% 174
$25,000 - $49,999 26% 262
$50,000 - $74,999 20% 201
$75,000 - $99,999 15% 148
$100,000 and over 21% 215
     
: : :

 

Hispanic origin % N
     
Mexico  61% 607
Puerto Rico 10% 101
Cuba  4% 45
Other Spanish/Hispanic/Latino heritage 25% 247
     
Total 100 1,000

 

Lightspeed GMI

Founded in 1996, Lightspeed GMI’s double opt-in U.S. online consumer panel contains approximately 1.27 million U.S. consumers. Lightspeed GMI recruits its panelists through many different sources including web advertising, permission-based databases and partner-recruited panels.

Survey Sampling International (SSI)

Mintel partners with SSI to collect data for Hispanic reports. Founded in 1977, SSI is an industry leader in general population, specialized, and Hispanic samples.    

Secondary Data Analysis

In addition to exclusively commissioned surveys, Mintel gathers syndicated data from the most respected consumer research firms. This allows Mintel analysts to form objective and cohesive analyses of consumer attitudes and behavior.

Simmons National Consumer Studies

Mintel reports frequently draw on the Simmons National Consumer surveys conducted by Simmons Research. The Simmons National Consumer Study (NCS) is a comprehensive survey of American consumers aged 18 and older. It provides single-source measurement of major media, products, services, and in-depth consumer demographic and lifestyle/psychographic characteristics.

  • 20,000 Adults 18+
  • Two-phase data collection
    • Phase 1: A telephone placement interview for a self-administered mail survey is conducted with any adult aged 18 or over in the household
    • Phase 2: Self-administered survey is mailed to each qualified household member 
  • Upfront cash incentives/sweepstakes offer
  • All qualified household members aged 18 or over participate by completing their own personal booklets
  • Principal shopper completes one Household Survey
  • Conducted and released quarterly - Winter, Spring, Summer and Fall
  • Ability to look at full-year and quarterly data

The Simmons National Hispanic Study (NHCS) is the only national, multi-media syndicated research instrument targeting the Hispanic market, and is particularly valuable in identifying media habits, product and service usage and attitudes and opinions among this segment.

  • 7,500 Hispanic adults 18+
  • Two-phase data collection
    • Phase 1: A telephone placement interview for a self-administered mail survey is conducted with any adult aged 18 or over in the household
    • Phase 2: Self-administered survey is mailed to each qualified household member 
  • Survey offered in Spanish or English - respondent’s choice
  • Incentive/sweepstakes offer
  • Conducted and released quarterly - Winter, Spring, Summer and Fall
  • Ability to look at full-year and quarterly data
  • The samples for the Kids and Teens Studies are taken from the same households participating in the adult study. The Kids and Teens Studies provide in-depth insight into these consumer segments to understand their effect on the marketplace, and how and where to reach them.
  • 2,500 Teens 12-17 and 2,600 Kids 6-11
  • Both samples gathered from within NCS participating households
  • Upfront incentive/sweepstakes offer
  • All teens or kids in household participate by completing their own personal booklets
  • Released twice a year—Spring and Fall data releases

In some instances Mintel uses Experian’s Mosaic segmentation system to further analyze Simmons NHCS data. Mosaic is a household-based segmentation system, which classifies all U.S. households and neighborhoods into 71 unique Mosaic segments and 19 groupings that share similar demographic and socioeconomic characteristics. Descriptive content is sourced from Simmons NCS/NHCS data.

As of the Fall 2015 Simmons National Consumer Study, all of the Mosaic groups and types are listed below:

 

% of U.S. Households
A - Power Elite 4.68
A01 - American Royalty 1.01
A02 - Platinum Prosperity .79
A03 - Kids and Cabernet 0.69
A04 - Picture Perfect Families 0.79
A05 - Couples with Clout 0.66
A06 - Jet Set Urbanites 0.73
: :
: :

Click here to shrink table

 

Qualitative Research 

Revelation by FocusVision . FocusVision provides Mintel with qualitative bulletin board software. This allows the creation of Internet-based, ‘virtual’ venues where participants recruited from Mintel’s online surveys gather and engage in interactive, text-based discussions led by Mintel moderators.

 

Further Analysis

Mintel employs numerous quantitative data analysis techniques to enhance the value of our consumer research. The techniques used vary form one report to another.  Below describes some of the more commonly used techniques. 

 

Repertoire Analysis

 

This is used to create consumer groups based on reported behaviour or attitudes. Consumer responses of the same value (or list of values) across a list of survey items are tallied into a single variable. The repertoire variable summarizes the number of occurrences in which the value or values appear among a list of survey items.  For example, a repertoire of brand purchasing might produce groups of those that purchase 1-2 brands, 3-4 brands and 5 or more brands. Each subgroup should be large enough (ie N=75+) to analyze. 

 

Cluster Analysis

 

This technique assigns a set of individual people in to groups called clusters on the basis of one or more question responses, so that respondents within the same cluster are in some sense closer or more similar to one another than to respondents that were grouped into a different cluster.

 

Correspondence Analysis

 

This is a statistical visualization method for picturing the associations between rows (image, attitudes) and columns (brands, products, segments, etc) of a two-way contingency table. It allows us to display brand images (and/or consumer attitudes towards brands) related to each brand covered in this survey in a joint space that is easy to understand. The significance of the relationship between a brand and its associated image is measured using the Chi-square test. If two brands have similar response patterns regarding their perceived images, they are assigned similar scores on underlying dimensions and will then be displayed close to each other in the perceptual map.

 

CHAID analysis

 

CHAID (Chi-squared Automatic Interaction Detection), a type of decision tree analysis, is used to highlight key target groups in a sample by identifying which sub-groups are more likely to show a particular characteristic. This analysis subdivides the sample into a series of subgroups that share similar characteristics towards a specific response variable and allows us to identify which combinations have the highest response rates for the target variable. It is commonly used to understand and visualize the relationship between a variable of interest such as “interest in trying a new product” and other characteristics of the sample, such as demographic composition.

 

Key Driver Analysis

 

Key driver analysis can be a useful tool in helping to prioritize focus between different factors which may impact key performance indicators (eg satisfaction, likelihood to switch providers, likelihood to recommend a brand, etc). Using correlations analysis or regression analysis provides an understanding of which factors or attributes of a market have the strongest association or “link” with a positive performance on key performance indicators (KPIs). Hence, factors or attributes are identified which are relatively more critical in a market category compared to others and ensures that often limited resources can be allocated to focusing on the main market drivers.

 

TURF Analysis

 

TURF (Total Unduplicated Reach & Frequency) analysis identifies the mix of features, attributes, or messages that will attract the largest number of unique respondents.  It is typically used when the number of features or attributes must be or should be limited, but the goal is still to reach the widest possible audience. By identifying the Total Unduplicated Reach, it is possible to maximize the number of people who find one or more of their preferred features or attributes in the product line. The resulting output from TURF is additive, with each additional feature increasing total reach. The chart is read from left to right, with each arrow indicating the incremental change in total reach when adding a new feature. The final bar represents the maximum reach of the total population when all shown features are offered.

 

Social Media Research

To complement our exclusive consumer research, Mintel tracks and analyses social media data for inclusion in Mintel Reports. Using Infegy’s Social Radar software, Mintel 'listens in' on online conversations across a range of social platforms including Facebook, Twitter, consumer forums and the wider web.

Social Radar provides rich consumer insight via the analysis of commentary posted publicly on the internet. The system performs comprehensive and broad collection of data from millions of internet sources, working to ensure a faithful and extensive sampling of feedback from the widest range of individuals. The dataset contains commentary posted in real time, as well as a substantial archive dating back through 2007.

To perform the analysis, related commentary is read by machine, with specialized proprietary software able to break down and understand human language and complex grammatical structures. This process can then extract topical information, sentiment and tonality, emotional expressiveness and more with accuracy and speed.


Trade research

 

Informal

Mintel conducts informal trade research for all reports. This involves contacting key players in the trade not only to gain information concerning their own operations, but also to obtain explanations and views of the strategic issues pertinent to the market being researched in order to address current brand and marketing issues. To ensure accuracy, Mintel sends draft copies of reports to key industry representatives for review, taking their feedback into consideration before publishing the report. Comments, where appropriate, are incorporated into the report.

Formal

Internally, Mintel’s analysts undertake extensive trade interviews with selected key experts in the field for the majority of reports. The purpose of these interviews is to assess key issues in the market place in order to ensure that any research undertaken takes these into account.

In addition, using experienced external researchers, trade research is undertaken for some reports. This takes the form of full trade interview questionnaires and direct quotes are included in the report and analysed by experts in the field. This gives a valuable insight into a range of trade views of topical issues. Direct quotations are included in the reports, giving valuable insight into a range of trade views on topical issues.

Desk Research

Mintel has an internal team of market analysts who monitor: government statistics, consumer and trade association statistics, manufacturer sponsored reports, annual company reports and accounts, directories and press articles from around the world and online databases. The latter are extracted from hundreds of publications and websites, both U.S. and overseas. All information is cross-referenced for immediate access. Data from other published sources are the latest available at the time of writing the report. This information is supplemented by an extensive library of Mintel’s reports produced since 1972 globally and added to each year by the 500+ reports which are produced annually.

In addition to in-house sources, researchers also occasionally use outside libraries or data from Trade Associations. Other information is also gathered from store and exhibition visits across the U.S., as well as using other databases within the Mintel Group, such as the Global New Product Database (GNPD), which monitors FMCG sales promotions.

Intelligence gathered through desk research is used to guide research and enrich data findings.


Statistical forecasting

 

Statistical modelling

For the majority of reports, Mintel produces five-year forecasts based on an advanced statistical technique known as ‘multivariate time series auto-regression’ using the statistical software package SPSS.

The model is based on historical market size data taken from Mintel’s own market size database and supplemented by published macroeconomic and demographic data from various private and public sources including the Federal Reserve Board, the U.S. Commerce Department, the Census Bureau, the Council of Economic Advisers, and the Congressional Budget Office.

The model searches for relationships between actual market sizes and a selection of relevant and significant macroeconomic and demographic determinants (independent variables) to identify those predictors having the most influence on the market.

Factors used in a forecast are stated in the relevant report section alongside an interpretation of their role in explaining the development in demand for the product or market in question.

Qualitative insight

At Mintel we understand that historic data is limited in its capacity to act as the only force behind the future state of markets. Thus, rich qualitative insights from industry experts regarding future events that might impact upon various markets play an invaluable role in our post statistical modeling evaluation process.

As a result, the Mintel forecast complements a rigorous statistical process with in-depth market knowledge and expertise to allow for additional factors or market conditions outside of the capacity of the statistical forecast. 

 

 

  

The Mintel fan chart

Forecasts of future economic outcomes are always subject to uncertainty. In order to raise awareness amongst our clients and to illustrate this uncertainty, Mintel has introduced a new way of displaying market size forecasts in the form of a fan-chart.

Next to historical market sizes and a current year estimate, the fan chart illustrates the probability of various outcomes for the market value/volume over the next five years.

At a 95% confidence interval, we are saying that 95 out of 100 times the forecast will fall within these outer limits, which we call the best and worst case forecasts. These, based on the statistically driven forecast, are the highest (best case) and lowest (worst case) market sizes the market is expected to achieve.

Over the next five years, the widening bands successively show the developments that occur within 95%, 90%, 70% and 50% probability intervals. Statistical processes predict the central forecast to fall within the darker shaded area which illustrates 50% probability, i.e. a 5 in 10 chance.

In general, based on our current knowledge of given historic market size data as well as projections for key macro- and socio-economic measures that were used to create the forecast, we can assume that in 95% of the time the actual market size will fall within the purple shaded fan. In 5% of all cases this model might not be correct due to random errors and the actual market size will fall out of these boundaries.

Weather analogy

To illustrate uncertainty in forecasting in an everyday example, let us assume the following weather forecast was produced based on the meteorologists’ current knowledge of the previous weather condition during the last few days, atmospheric observations, incoming weather fronts etc.

 

 

   

Now, how accurate is this forecast and how certain can we be that the temperature on Saturday will indeed be 15°C?

To state that the temperature in central London on Saturday will rise to exactly 15°C is possible but one can’t be 100% certain about that fact.

To say the temperature on Saturday will be between 13°C and 17°C is a broader statement and much more probable.

In general, we can say that based on the existing statistical model, one can be 95% certain that the temperature on Saturday will be between 13°C and 17°C, and respectively 50% certain it will be between about 14.5°C and 15.5°C. Again, only in 5% of all cases this model might not be correct due to random errors and the actual temperature on Saturday will fall out of these boundaries and thus will be below 13°C or above 17°C.

( To learn more about uncertainty in weather forecasts visit: http://research.metoffice.gov.uk/research/nwp/ensemble/uncertainty.html )

UK

Mintel is an independent market analysis company that prides itself on supplying objective information on a whole range of markets and marketing issues.

There are six main sources of research that are used in the compilation of Mintel reports:

  • Consumer research
  • Brand & social media research
  • Desk research
  • Trade research
  • Statistical forecasting
  • Mintel’s exclusive archive of over 40 years of analysis and expertise.

Mintel reports are written and managed by analysts with experience in the relevant markets.


Consumer research

 

Exclusive and original quantitative consumer research is commissioned for almost all Mintel reports. In addition, qualitative research is also undertaken for a large proportion of reports in the form of online discussion groups. Mintel invests a considerable sum each year in consumer research, and the purchaser of a Mintel report benefits, as the price of an individual report is less than the cost of the original research alone. The research brings an up-to-date and unique insight into topical issues of importance.

Consumer research is conducted among a nationally representative sample of either adults or internet users and is generally carried out by Lightspeed GMI (online), Ipsos Mori (face to face), while other suppliers are used on an ad hoc basis as required. The results are only available in Mintel reports.

Sampling and weighting

 

Face to Face Surveys

 

Ipsos Mori

Ipsos MORI Capibus uses a two-stage random location sample design which generates a very high quality sample representative of the Great Britain adult population. Interviews are conducted via c170-180 sampling points, randomly selected every week, and CACI ACORN is employed to set interlocking quota controls specific to each interviewer location. This ensures consistent accurate representation of the locations interviewed every week. By using this proven sample design, all sub-sectors of the population are represented – at a national and regional level.

All information collected on Capibus is then weighted to reflect the known profile of the adult population in Great Britain. Capibus uses a rim weighting system which weights to mid-2010 census and NRS defined profiles for age, social grade, region, ethnicity and working status - within gender. Additional profiles used include tenure and car in household, for example.

Because the sampling process is repeated every week, the Capibus sample is matched wave on wave, making it ideal for taking successive measurements on the same issue.

Calculation of socio-economic grade in Ipsos MORI Capibus surveys

Socio-economic grade is classified according to the occupation of the chief income earner in the household. A number of questions are asked by the interviewer in order to assign social grade accurately. The interviewer probes the respondent for information about the occupation of the chief income earner, the type of organisation he or she works for, job actually done, job title/rank/grade, and whether the chief income earner is self-employed. Additionally, questions are asked about the number of people working at the place of employment and whether the chief income earner is responsible for anyone, together with confirmation of qualifications. Once the interviewer is satisfied that sufficient information has been gathered in order to determine social grade, their estimate is recorded and this is later double checked for accuracy by experts in social grading coding at Ipsos MORI when it can be amended if necessary.

 

Online Surveys

 

Lightspeed GMI

Founded in 1999, GMI’s double opt-in online consumer panel has reach to approximately 350,000 consumers in Great Britain.  Re-branded as Lightspeed GMI in September 2014 (after its acquisition by Kantar in 2011), it delivers uniquely identified online respondents via extensive use of fraud detection and location-verification technology at multiple points in the research cycle, from initial registration through survey fielding and incentive redemption. Lightspeed GMI panellists are profiled on a wide variety of attri­butes to deliver the specific hard-to-reach demographics.

To ensure our surveys are nationally representative of internet users, Mintel sets quotas for each age group, split by gender. Specific quotas for a sample of 2,000 adults aged 16+ are shown below:

 

 

%

N

Age groups by gender

 

 

 

 

 

16-19 Men

3.3

67

16-19 Women

3.2

64

 

 

 

20-24 Men

4.6

92

20-24 Women

4.5

89

: : :
: : :

 

Mintel also sets quotas on region and socio-economic group. Specific quotas for a sample of 2,000 adults aged 16+ are shown below:

 

 

%

N

Region

   

 

 

 

North East

4.1

83

North West

11.3

227

Yorkshire & Humberside

8.5

170

East Midlands

7.4

148

West Midlands

9.1

182

: : :
: : :

 

 

%

N

Socio-economic grade

   

 

 

 

AB

22.3

446

C1

30.9

618

C2

20.9

418

DE

25.9

518

 

 

 

: : :

 

Calculation of socio-economic grade in Lightspeed GMI surveys 

To calculate panellist’s socio-economic grade Lightspeed GMI ask a set of screening questions at registration to their panel. These questions include employment status and profession (both panellists own and the household’s chief income earner). Once the panellist has completed these initial questions they are invited to take an additional set of follow-up questions that allow Lightspeed GMI to further fine tune their socio-economic grade. Panellists are required to retake the socio-economic screening questions once a year to ensure that any changes in circumstance are accounted for.

 

Qualitative Research

 

FocusVision Revelation

FocusVision provides Mintel with qualitative bulletin board software 'Revelation'. This allows the creation of Internet-based, ‘virtual’ venues where participants recruited from Mintel’s online surveys gather and engage in interactive, text-based discussions led by Mintel moderators.

Further Analysis

Mintel employs numerous quantitative data analysis techniques to enhance the value of our consumer research. The techniques used vary form one report to another.  Below describes some of the more commonly used techniques.

Repertoire Analysis

This is used to create consumer groups based on reported behaviour or attitudes. Consumer responses of the same value (or list of values) across a list of survey items are tallied into a single variable. The repertoire variable summarises the number of occurrences in which the value or values appear among a list of survey items.  For example, a repertoire of brand purchasing might produce groups of those that purchase 1-2 brands, 3-4 brands and 5 or more brands. Each subgroup should be large enough (ie N=75+) to analyse.

Cluster Analysis

This technique assigns a set of individual people in to groups called clusters on the basis of one or more question responses, so that respondents within the same cluster are in some sense closer or more similar to one another than to respondents that were grouped into a different cluster.

Correspondence Analysis

This is a statistical visualisation method for picturing the associations between rows (image, attitudes) and columns (brands, products, segments, etc.) of a two-way contingency table. It allows us to display brand images (and/or consumer attitudes towards brands) related to each brand covered in this survey in a joint space that is easy to understand. The significance of the relationship between a brand and its associated image is measured using the Chi-square test. If two brands have similar response patterns regarding their perceived images, they are assigned similar scores on underlying dimensions and will then be displayed close to each other in the perceptual map.

CHAID analysis

CHAID (Chi-squared Automatic Interaction Detection), a type of decision tree analysis, is used to highlight key target groups in a sample by identifying which sub-groups are more likely to show a particular characteristic. This analysis subdivides the sample into a series of subgroups that share similar characteristics towards a specific response variable and allows us to identify which combinations have the highest response rates for the target variable. It is commonly used to understand and visualise the relationship between a variable of interest such as “interest in trying a new product” and other characteristics of the sample, such as demographic composition.

Key Driver Analysis

Key driver analysis can be a useful tool in helping to prioritise focus between different factors which may impact key performance indicators (eg satisfaction, likelihood to switch providers, likelihood to recommend a brand, etc). Using correlations analysis or regression analysis we can get an understanding of which factors or attributes of a market have the strongest association or “link” with a positive performance on key performance indicators (KPIs). Hence, we are able to identify which factors or attributes are relatively more critical in a market category compared to others and ensures that often limited resources can be allocated to focusing on the main market drivers.

 

TURF Analysis

TURF (Total Unduplicated Reach & Frequency) analysis identifies the mix of features, attributes, or messages that will attract the largest number of unique respondents.  It is typically used when the number of features or attributes must be or should be limited, but the goal is still to reach the widest possible audience. By identifying the Total Unduplicated Reach, it is possible to maximize the number of people who find one or more of their preferred features or attributes in the product line. The resulting output from TURF is additive, with each additional feature increasing total reach. The chart is read from left to right, with each arrow indicating the incremental change in total reach when adding a new feature. The final bar represents the maximum reach of the total population when all shown features are offered. 


Brand & Social Media Research

 

Mintel’s brand research includes more in-depth knowledge about brands covered in relevant markets.  Brands are evaluated on a number of areas including usage, commitment, dynamism, differentiation, satisfaction, image, recommendation and attitudes. When evaluating attitudes, up to ten brand attitude statements are selected and tested, depending on the actual consumer market. 

To complement its exclusive consumer research, Mintel tracks social media data for inclusion in selected reports. Using Infegy’s Atlas software, Mintel analyses online conversations across a range of social platforms including Twitter, consumer forums and the wider web. 

Atlas provides rich consumer insight via the analysis of commentary posted publicly on the internet. The system performs comprehensive and broad collection of data from millions of internet sources, working to ensure a faithful and extensive sampling of feedback from the widest range of individuals. The dataset contains commentary posted in real time, as well as a substantial archive dating back to 2007.


Trade research

 

Informal

Trade research is undertaken for all reports. This involves contacting relevant players in the trade, not only to gain information concerning their own operations, but also to obtain explanations and views of the strategic issues pertinent to the market being researched. 

Formal

Internally, Mintel’s analysts undertake extensive trade interviews with selected key experts in the field for the majority of reports. The purpose of these interviews is to assess key issues in the market place in order to ensure that any research undertaken takes these into account.

In addition, using experienced external researchers, trade research is undertaken for some reports. This takes the form of full trade interview questionnaires and direct quotes are included in the report and analysed by experts in the field. This gives a valuable insight into a range of trade views of topical issues.

Desk research

Mintel has an internal team of market analysts who monitor: government statistics, consumer and trade association statistics, manufacturer sponsored reports, annual company reports and accounts, directories, press articles from around the world and online databases. The latter are extracted from hundreds of publications and websites, both British and overseas. All information is cross-referenced for immediate access.

Data from other published sources are the latest available at the time of writing the report.

This information is supplemented by an extensive library of Mintel’s reports produced since 1972 and added to each year by the 500+ reports which are produced annually.

In addition to in-house sources, researchers also occasionally use outside libraries such as the British Library or the Department of Trade and Industry. Other information is also gathered from store and exhibition visits across Europe, as well as using other databases within the Mintel Group, such as the Global New Product Database (GNPD), which monitors FMCG sales promotions.

All analysts have access to Mintel’s Market Size and Macroeconomic Databases – a database containing many areas of consumer spending and retail sales as well as macroeconomic and demographic factors which impinge on consumer spending patterns, going back some 20 years.

The database is used in conjunction with an SPSS forecasting program which uses weighted historical correlations of market dynamics, with independent variables, to produce future spending scenarios.


Statistical forecasting

 

Statistical modelling

For the majority of reports, Mintel produces five-year forecasts based on an advanced statistical technique known as ‘multivariate time series auto-regression’ using the statistical software package SPSS.

Historical market size data feeding into each forecast are collated in Mintel’s own market size database and supplemented by macro- and socio-economic data sourced from organisations such as the Office for National Statistics, HM Treasury, the Bank of England and the Economist Intelligence Unit. 

Within the forecasting process, the model searches for, and analyses relationships between, actual market sizes and a selection of key economic and demographic determinants (independent variables) in order to identify those predictors having the most influence on the market.

Factors used in a forecast are stated in the relevant report section alongside an interpretation of their role in explaining the development in demand for the product or market in question.

Qualitative insight

At Mintel we understand that historic data is limited in its capacity to act as the only force behind the future state of markets. Thus, rich qualitative insights from industry experts regarding future events that might impact upon various markets play an invaluable role in our post statistical modeling evaluation process.

As a result, the Mintel forecast complements a rigorous statistical process with in-depth market knowledge and expertise to allow for additional factors or market conditions outside of the capacity of the statistical forecast. 

 

 

Forecasts of future economic outcomes are always subject to uncertainty. In order to raise awareness amongst our clients and to illustrate this uncertainty, Mintel has introduced a new way of displaying market size forecasts in the form of a fan-chart.

Next to historical market sizes and a current year estimate, the fan chart illustrates the probability of various outcomes for the market value/volume over the next five years.

At a 95% confidence interval, we are saying that 95 out of 100 times the forecast will fall within these outer limits, which we call the best and worst case forecasts. These, based on the statistically driven forecast, are the highest (best case) and lowest (worst case) market sizes the market is expected to achieve.

Over the next five years, the widening bands successively show the developments that occur within 95%, 90%, 70% and 50% probability intervals. Statistical processes predict the central forecast to fall within the darker shaded area which illustrates 50% probability, i.e. a 5 in 10 chance.

A general conclusion: Based on our current knowledge of given historic market size data as well as projections for key macro- and socio-economic measures that were used to create the forecast, we can assume that in 95% of the time the actual market size will fall within the purple shaded fan. In 5% of all cases this model might not be correct due to random errors and the actual market size will fall out of these boundaries.

Weather analogy

To illustrate uncertainty in forecasting in an everyday example, let us assume the following weather forecast was produced based on the meteorologists’ current knowledge of the previous weather condition during the last few days, atmospheric observations, incoming weather fronts etc.

 

 

Now, how accurate is this forecast and how certain can we be that the temperature on Saturday will indeed be 15°C?

To state that the temperature in central London on Saturday will rise to exactly 15°C is possible but one can’t be 100% certain about that fact.

To say the temperature on Saturday will be between 13°C and 17°C is a broader statement and much more probable.

In general, we can say that based on the existing statistical model, one can be 95% certain that the temperature on Saturday will be between 13°C and 17°C, and respectively 50% certain it will be between about 14.5°C and 15.5°C. Again, only in 5% of all cases this model might not be correct due to random errors and the actual temperature on Saturday will fall out of these boundaries and thus will be below 13°C or above 17°C.

(To learn more about uncertainty in weather forecasts visit: http://research.metoffice.gov.uk/research/nwp/ensemble/uncertainty.html)

UK B2B

Mintel is an independent market analysis company that prides itself on supplying objective information on a whole range of markets and marketing issues.

There are four main sources of research that are used in the compilation of Mintel’s business-to-business (B2B) reports:

  • Desk Research
  • Trade Research
  • Statistcal forecasting
  • Mintel's exclusive archive of over 40 years analysis and expertise
  • Consumer research (selected reports only)

Mintel reports are written and managed by analysts with experience in the relevant markets.


Trade research

 

Informal

Trade research is undertaken for all reports. This involves contacting relevant players in the trade, not only to gain information concerning their own operations, but also to obtain explanations and views of the strategic issues pertinent to the market being researched. Such is Mintel’s concern with accuracy that draft copies of reports are sent to industry representatives, to get their feedback and avoid any misrepresentation of the market. These comments are incorporated into reports prior to final publication.

Formal

Internally, Mintel’s analysts undertake extensive trade interviews with selected key experts in the field for the majority of reports. The purpose of these interviews is to assess key issues in the market place in order to ensure that any research undertaken takes these into account.

In addition, using experienced external researchers, trade research is undertaken for some reports. This takes the form of full trade interview questionnaires and direct quotes are included in the report and analysed by experts in the field. This gives a valuable insight into a range of trade views on topical issues.

Desk research

Mintel has an internal team of desk researchers who monitor: government statistics, consumer and trade association statistics, manufacturer sponsored reports, annual company reports and accounts, directories, press articles from around the world and online databases. The latter are extracted from hundreds of publications and websites, both British and overseas. All information is cross-referenced for immediate access.

Data from other published sources are the latest available at the time of writing the report.

In addition to in-house sources, researchers also occasionally use outside libraries such as the British Library or the Department of Trade and Industry.


Consumer research

 

Exclusive and original quantitative consumer research is commissioned for almost all Mintel reports. In addition, qualitative research is also undertaken for a large proportion of reports in the form of online discussion groups. Mintel invests a considerable sum each year in consumer research, and the purchaser of a Mintel report benefits, as the price of an individual report is less than the cost of the original research alone. The research brings an up-to-date and unique insight into topical issues of importance.

Consumer research is conducted among a nationally representative sample of either adults or internet users and is generally carried out by Lightspeed GMI (online), Ipsos Mori (face to face), while other suppliers are used on an ad hoc basis as required. The results are only available in Mintel reports.

Sampling and weighting

 

Face to Face Surveys

 

Ipsos Mori

Ipsos MORI Capibus uses a two-stage random location sample design which generates a very high quality sample representative of the Great Britain adult population. Interviews are conducted via c170 sampling points, randomly selected every week, and MOSAIC is employed to set interlocking quota controls specific to each interviewer location. This ensures consistent accurate representation of the locations interviewed every week. By using this proven sample design, all sub-sectors of the population are represented – at a national and regional level.

All information collected on Capibus is then weighted to reflect the known profile of the adult population in Great Britain. Capibus uses a rim weighting system which weights to mid-2010 census and NRS defined profiles for age, social grade, region, ethnicity and working status - within gender. Additional profiles used include tenure and car in household, for example.

Because the sampling process is repeated every week, the Capibus sample is matched wave on wave, making it ideal for taking successive measurements on the same issue.

 

Online Surveys

 

Lightspeed GMI

Founded in 1999, GMI’s double opt-in online consumer panel has reach to approximately 350,000 consumers in Great Britain.  Re-branded as Lightspeed GMI in September 2014 (after its acquisition by Kantar in 2011), it delivers uniquely identified online respondents via extensive use of fraud detection and location-verification technology at multiple points in the research cycle, from initial registration through survey fielding and incentive redemption. Lightspeed GMI panellists are profiled on a wide variety of attri­butes to deliver the specific hard-to-reach demographics.

To ensure our surveys are nationally representative of internet users, Mintel sets quotas for each age group, split by gender. Specific quotas for a sample of 2,000 adults aged 16+ are shown below:

 

 

%

N

Age groups by gender

 

 

 

 

 

16-19 men

3.6

72

16-19 women

3.4

68

 

 

 

20-24 men

4.9

98

20-24 women

4.8

95

: : :
: : :

 

Mintel also sets quotas on region and socio-economic group. Specific quotas for a sample of 2,000 adults aged 16+ are shown below:

 

 

%

N

Region

   

 

 

 

North East

4.2

84

North West

11.5

229

Yorkshire & Humberside

8.6

172

East Midlands

7.4

148

West Midlands

9.1

182

: : :
: : :

 

 

%

N

Social economic group

   

 

 

 

AB

22.3

446

C1

30.9

618

C2

20.9

418

DE

25.9

518

 

 

 

: : :

 

 

Definitions

 

Socio-economic group

 

Socio-economic groups are based on the head of household or chief income earner and are defined as follows:

Socio-economic group

Occupation of chief income earner

 

 

A

Higher managerial, administrative or professional

B

Intermediate managerial, administrative or professional

C1

Supervisory or clerical, and junior managerial, administrative or professional

C2

Skilled manual workers

D

Semi and unskilled manual workers

E

All those entirely dependent on the state long term, through sickness, unemployment, old age or other reasons

 

Retired persons who have a company pension or private pension, or who have private means are graded on their previous occupation.

Students in higher education living at home are graded on the occupation of the head of the household. Students living away from home are graded C1 (no account is taken of casual or vacation jobs).

 

Qualitative Research

 

Toluna/ 2020 Research

Toluna in partnership with 20/20 Research provide Mintel with qualitative bulletin board software. This allows the creation of Internet-based, ‘virtual’ venues where participants recruited from Mintel’s online surveys gather and engage in interactive, text-based discussions led by Mintel moderators.

Further Analysis

Mintel employs numerous quantitative data analysis techniques to enhance the value of our consumer research. The techniques used vary form one report to another.  Below describes some of the more commonly used techniques.

Repertoire Analysis

This is used to create consumer groups based on reported behaviour or attitudes. Consumer responses of the same value (or list of values) across a list of survey items are tallied into a single variable. The repertoire variable summarises the number of occurrences in which the value or values appear among a list of survey items.  For example, a repertoire of brand purchasing might produce groups of those that purchase 1-2 brands, 3-4 brands and 5 or more brands. Each subgroup should be large enough (ie N=75+) to analyse.

Cluster Analysis

This technique assigns a set of individual people in to groups called clusters on the basis of one or more question responses, so that respondents within the same cluster are in some sense closer or more similar to one another than to respondents that were grouped into a different cluster.

Correspondence Analysis

This is a statistical visualisation method for picturing the associations between rows (image, attitudes) and columns (brands, products, segments, etc.) of a two-way contingency table. It allows us to display brand images (and/or consumer attitudes towards brands) related to each brand covered in this survey in a joint space that is easy to understand. The significance of the relationship between a brand and its associated image is measured using the Chi-square test. If two brands have similar response patterns regarding their perceived images, they are assigned similar scores on underlying dimensions and will then be displayed close to each other in the perceptual map.

CHAID analysis

CHAID (Chi-squared Automatic Interaction Detection), a type of decision tree analysis, is used to highlight key target groups in a sample by identifying which sub-groups are more likely to show a particular characteristic. This analysis subdivides the sample into a series of subgroups that share similar characteristics towards a specific response variable and allows us to identify which combinations have the highest response rates for the target variable. It is commonly used to understand and visualise the relationship between a variable of interest such as “interest in trying a new product” and other characteristics of the sample, such as demographic composition.

Key Driver Analysis

Key driver analysis can be a useful tool in helping to prioritise focus between different factors which may impact key performance indicators (eg satisfaction, likelihood to switch providers, likelihood to recommend a brand, etc). Using correlations analysis or regression analysis we can get an understanding of which factors or attributes of a market have the strongest association or “link” with a positive performance on key performance indicators (KPIs). Hence, we are able to identify which factors or attributes are relatively more critical in a market category compared to others and ensures that often limited resources can be allocated to focusing on the main market drivers.


Statistical forecasting

 

Statistical modelling

Historical market size data feeding into each forecast are collated in Mintel’s own market size database and supplemented by macro- and socio-economic data sourced from organisations such as the Office for National Statistics, HM Treasury and the Bank of England. Forecasts are modelled in the unique Mintel database, based on trade insight and expectations, combined with historic performance, taking into account a combination of economic factors, end use market activity, the regulatory environment where applicable, industry expectations and technological developments.

Qualitative insight

Crucially the vast majority of reports also draw on the Mintel quantitative and qualitative understanding and analysis of the relevant end use markets. At Mintel we understand that historic data is limited in its capacity to act as the only force behind the future state of markets. Thus, rich qualitative insights from industry experts regarding future events that might impact upon various markets play an invaluable role in our post statistical modelling evaluation process.

As a result, the Mintel forecast complements a rigorous statistical process with in-depth market knowledge and expertise to allow for additional factors or market conditions outside of the capacity of the statistical forecast.

 

 

   

Europe

For Europe, Mintel surveys five major European markets - France, Germany, Spain, Italy and Poland - with samples of either 1,000 or 2,000 adults aged 16+. The surveys are conducted online, via Lightspeed GMI, with ten waves of research a year.

  • Our consumer research is based on a random sample of consumers from each of the five countries.
  • The research is representative of the online population in each country. Mintel applies a quota-sampling approach with quotas based on age, gender and region (see more details below).
  • For category-focused reports (eg chocolate confectionery), consumers will typically be asked about usage, frequency and location of purchase, consumption occasion, brand usage and a series of attitudinal statements about the category.

Sample sizes by demographic and geographies

 

Mintel’s consumer research is representative of the online population of each country by region. Specific quotas for a sample of 2,000 respondents are shown below:

France

 

 

%

N

 

 

 

Region

   

 

 

 

Île de France

18.8

376

Bassin Parisien

16.9

337

Nord - Pas-de-Calais

6.4

127

Est

8.4

168

: : :
: : :

 

 

Spain

 

 

%

N

 

 

 

Region

   

 

 

 

Noroeste

9.4

188

Noreste

9.6

191

Comunidad de Madrid

13.8

275

Centro

12.2

243

: : :
: : :

 

Germany

 

 

%

N

Region

   

 

 

 

Baden-Württemberg

13.2

264

Bayern

15.7

313

Berlin

4.3

85

Brandenburg

3.1

61

Bremen

0.8

16

: : :
: : :

 

 

Italy

 

 

%

N

Region

   

 

 

 

Nord-Ovest

26.6

531

Sud

23.3

465

Isole

11.1

222

Nord-Est

19.2

384

Centro

19.9

398

: : :
: : :

 

 

Poland

 

 

%

N

Region

   

 

 

 

Centralny

20.5

410

Poludniowy

20.8

415

Wschodni

17.4

348

Pólnocno-Zachodni

16.2

323

Poludniowo-Zachodni

10.1

201

: : :
: : :

 

To also receive representative and comparable age data, Mintel has defined age segments for each country. Specific quotas for each age group, split by gender, are shown below:

France

 

 

%

N

Age groups by gender

 

 

 

 

 

16-19 men

3.7

74

16-19 women

3.5

70

 

 

 

20-24 men

4.4

87

20-24 women

4.3

85

: : :
: : :

 

Spain

 

 

%

N

Age groups by gender

 

 

 

 

 

16-19 men

3.1

61

16-19 women

2.9

57

 

 

 

20-24 men

4.1

82

20-24 women

4.0

79

: : :
: : :

 

Germany

 

 

%

N

Age groups by gender

 

 

 

 

 

16-19 men

2.9

57

16-19 women

2.7

54

 

 

 

20-24 men

4.0

80

20-24 women

3.8

76

: : :
: : :

 

Italy

 

 

%

N

 

 

 

Age groups by gender

 

 

 

 

 

16-19 men

3.4

68

16-19 women

3.2

63

 

 

 

20-24 men

4.5

90

: : :
: : :

 

Poland

 

 

%

N

 

 

 

Age groups by gender

 

 

 

 

 

16-19 men

3.9

78

16-19 women

3.7

74

 

 

 

20-24 men

5.9

118

: : :
: : :

 





Our research partner - Lightspeed GMI

 

Founded in 1999, Lightspeed GMI’s double opt-in online consumer panel has reach of approximately 210,000 consumers in Germany, 310,000 consumers in France, 180,000 consumers in Italy, 170,000 consumers in Spain and 35,000 consumers in Poland.  Lightspeed GMI delivers uniquely identified online respondents via extensive use of fraud detection and location-verification technology at multiple points in the research cycle, from initial registration through survey fielding and incentive redemption. Lightspeed GMI panellists are profiled on a wide variety of attributes to deliver the specific hard-to-reach demographics.  

China

For its China report series, Mintel has commissioned exclusive consumer research through QQsurvey China, a Chinese licensed market survey agent. Online consumer research is run in ten cities, completing 300 interviews per city with a total sample size of 3,000.

  • Our consumer research is based on a random sample of internet respondents from a panel recruited by QQSurvey China online research (see more details below)
  • In each wave, we survey four major tier 1 cities, i.e. Shanghai, Beijing, Guangzhou and Chengdu
  • For tier 2 and tier 3 cities, we rotate amongst a selection of cities based on size and economic development (see below)
  • The research is not representative of the population as a whole, and is not being analysed as such. Mintel applies a quota-sampling approach with quotas on age, sex and monthly household income in these cities
  • Our sample data can only be considered as indicative of urban consumers in those regions rather than representative of China as a whole
  • For category reports (e.g. chocolate confectionery), consumers will typically be asked about usage, frequency, location of purchase, consumption occasion, brand usage and a series of attitudinal statements about the category
  • Lifestyle reports cover a broader range of attitudinal and behavioural topics

Confidence levels

 

Statistical confidence levels of +/- 2% or 3% can be applied to the data, depending on sample size and percentage of respondents. For example, if 20% of a total sample of 1,000 adults say that they do something, you can be 95% certain that the figure for the population lies between 17% and 23%. For a sample of 3,000 adults, you can be 95% certain that the figure lies between 18% and 22%.

Consumer research is stored in a database supervised by Mintel's statisticians. Additional analysis of information too abundant to be included in published reports can be made available.


Sample sizes by city

 

When the overall population of a city is large enough (> 20,000), sample size is not determined by the size of population. It is only when the population becomes quite small (e.g. less than 10,000) that the sample size calculation is affected.

As a result a sample size of 300 per city across all 10 cities in our survey was set. Statistically, this enables us to apply a confidence level at 95% with a margin of error of 5.66%.


Sampling methodology and sampling structure

 

According to government figures, there are 645 cities in China. These cities are very different in terms of size, economic development, culture, history and lifestyle. To meet Mintel’s client interests (by region, by tier), ten cities are selected in each wave of research based on their geographic coverage and level of economic development (GDP and per capita income).

The table below shows an example of cities covered:

Region  Tier 1 cities  Tier 2 cities  Tier 3 cities  Total 
North China  Beijing  Shenyang    2
East China  Shanghai  Hangzhou  Ningbo 
Middle China    Wuhan   
West China  Chengdu     
South China  Guangzhou  Fuzhou  Dongguan 
Total  4  4  2  10 

Note 1: tier 2 and tier 3 cities in the table above are shown as examples only

Note 2: Mintel defines the tier levels of cities in China as follows:

  • Tier 1: Major economic hubs
  • Tier 2: Provincial capital cities
  • Tier 3: Cities at the level between provincial capital cities and local townships

Within each city, our sampling structure is presented below:

   

Gender & Age

Monthly household income (RMB)

Tier 1 city

Total

Male

20-29

Male

30-39

Male

40-49

Female

20-29

Female

30-39

Female

40-49

4K-10K

10K-16K

16K+

Beijing

300

50

50

50

50

50

50

100

100

100

Shanghai

300

50

50

50

50

50

50

100

100

100

Guangzhou

300

50

50

50

50

50

50

100

100

100

Chengdu

300

50

50

50

50

50

50

100

100

100

Tier 2/3 city

Total

Male

20-29

Male

30-39

Male

40-49

Female

20-29

Female

30-39

Female

40-49

4K-7K

7K-14K

14K+

City 1

300

50

50

50

50

50

50

100

100

100

: : : : : : : : : : :
: : : : : : : : : : :

Our research partner - QQSurvey

 
  • Founded in 2006, headquartered in Shanghai, with an office in Beijing
  • Online panel size - 2,600,000 (by Jan 2015)
  • Recruited over 300,000 new panellists in 2014
  • Owns the interactive panel website: www.1diaocha.com
  • Member of the China Market Research Association (CMRA)
  • Completed more than 1,900 projects and interviewed 750,000 samples in 2014
  • Provides panel data for major multinational research companies, including IPSOS, GMI, TNS, Lightspeed, and have done projects for the likes of Nielsen, Kantar, Intage, Pulse, SSI, ResearchNow (further details in the Appendix)

QQSurvey's sampling and quality control

 

1.Screening

Exclude those who

  • have participated in any survey project in the past three months
  • are working in any sensitive or related industries
  • have participated in previous Mintel surveys
  • have not met sample criteria

2.Sampling

  • System will output all qualified panelists
  • Random sample 30,000 panelists and send out first wave of invitations via SMS or email
  • Random sample another 30,000 panelists and send out second wave of invitations
  • Random sample final batch of 30,000 panelists and send out third wave of invitations

3. Quality control

  • Each panelist has provided QQSurvey with his or her own IP address together with all personal information
  • Each panelist needs to use the same IP address and cookie as registered IP address and cookie to participate in any survey project, and he or she can participate in the same survey project only once
  • Each respondent can use the weblink they have received once only

A sample will be deleted if the respondent

  • has given an answer to any open-ended question that is judged as of poor quality
  • has failed in any trap questions
  • has given any inconsistent answers, or contradicted his or her registered information
  • has given answers following certain patterns
  • has taken an extraordinary length of time to complete the questionnaire
  • is considered an outlier

Meet the Mintropolitans

 

Why Mintropolitans?

There has long been unresolved debate as to how to define the “middle class” in China, using various formulae about income levels, educational attainment and ownership of certain key items. However, the notion of “middle class” is very much an invention of 19th century North America and Western Europe, and does not comfortably translate that well into 21st century China.

So, rather than continue to struggle to make the round peg of China fit into the square hole of the “middle class”, Mintel decided to create our own definition for the significant consuming class of Chinese consumers, to better reflect the rapidly evolving trends among those consumers, and their diversity across the country.

In order to identify this new group, and as part of consumer surveys conducted for this, and other reports, Mintel added a new set of requirements so as to define this group of significant consumers. To this end, Mintel set four requirements they had to meet:

  • They need to have a higher level of income, depending on which city they live in: this being a household income of RMB 16,000 per month or over in tier one cities; or RMB 14,000 per month or over in tier two and three cities.
  • They need to have a higher level of education: undergraduate or above.
  • They need to own a property – either outright or on a mortgage.
  • Lastly, they also need to meet several key lifestyle criteria, chosen by Mintel to represent a significant measure of consuming power and aspiration, as well as reflecting a certain level of cultural and lifestyle sophistication.

Broadly, they should represent a significant, sophisticated consuming group who pursue quality of life rather than just wealth, are well educated and are the potential trendsetters. While they certainly belong to the higher income group, they are, however, different from “土豪” (tuhao: meaning those who have money but no taste). They are also higher in income, and spending power, than those who would be described as “小康” (xiaokang: meaning those who are living just comfortably). Their opinions are likely to represent the future of consuming trends in China, and therefore represent the key opportunity market, because they are considered to be a more “evolved” group.

Who are they?

Based on demographic data from the consumer research conducted across multiple Mintel Reports, Mintropolitans account for about 15% of total surveyed households – representing a population of 27 million households who live in tier one to three cities in China.

Mintropolitans are placed firmly in the higher income brackets. They are also much more likely to be aged in their 30s and 40s, and to be married, with children, and will have at least an undergraduate education, but more likely a postgraduate degree.

FIGURE 7: DEMOGRAPHIC PROFILE OF MINTROPOLITANS VS NON-MINTROPOLITANS, MARCH 2016

Base: 3,000 internet users aged 20-49

 

 

 SOURCE: QQSURVEY/MINTEL

While they are more likely to live in tier one cities and the east of China, there are still a significant, and likely growing, proportion of consumers in lower tier (tier two and three) cities and across different regions.

What is perhaps most interesting from the results of the survey conducted for this report is that they are slightly more likely to be those working in non-China owned or Sino-foreign business. This is less to do with income, and more to do with consumer aspiration, educational achievement and cultural sophistication.

FIGURE 8: DEMOGRAPHIC PROFILE OF MINTROPOLITANS VS NON-MINTROPOLITANS, MARCH 2016

Base: 3,000 internet users aged 20-49

 

 

 Source: QQSurvey/Mintel 


Further Analysis

 

Mintel employs a variety of quantitative data analysis techniques to enhance the value of our consumer research.  The techniques used vary form one report to another.  Below show some of the techniques more commonly used.

Repertoire Analysis

This technique is used to create consumer groups based on reported behaviour or attitudes. Consumer responses of the same value (or list of values) across a list of survey items are tallied into a single variable. The repertoire variable summarizes the number of occurrences in which the value or values appear among a list of survey items.  For example, a repertoire of brand purchasing might produce groups of those that purchase 1-2 brands, 3-4 brands and 5 or more brands.  Each subgroup should be large enough (ie N=75+) to analyze. 

Cluster Analysis

This technique used to assign a set of individual people to groups called clusters on the basis of one or more question responses, so that respondents within the same cluster are in some sense closer or more similar to one another than to respondents that were grouped into another cluster

Correspondence Analysis

This is a statistical visualisation method for picturing the associations between rows (image, attitudes) and columns (brands, products, segments, etc.) of a two-way contingency table.  It allows us to display brand images (and/ or consumer attitudes towards brands) related to each brand covered in this survey in a joint space that is easy to understand.  The significance of the relationship between a brand and its associated image is measured using the Chi-square test.  If two brands have similar response patterns regarding their perceived images, they are assigned similar scores on underlying dimensions and will then be displayed close to each other in the perceptual map

CHAID analysis

CHAID (Chi-squared Automatic Interaction Detection), a type of decision tree analysis, is a target group identification method that is used to highlight key target groups in a sample by identifying which sub-groups are more likely to show a particular characteristic. This analysis subdivides the sample into a series of subgroups that share similar characteristics towards a specific response variable and allows us to identify which combinations have the highest response rates for the target variable.

It is commonly used to understand and visualize the relationship between a variable of interest such as “interest in trying a new product” and other characteristics of the sample such as demographic composition.

Key Driver Analysis

Key driver analysis can be a useful tool in helping to prioritise focus between different factors which may impact key performance indicators (eg satisfaction, likelihood to switch providers, likelihood to recommend a brand, etc).   Using correlations analysis or regression analysis we can get an understanding of which factors or attributes of a market have the strongest association or “link” with a positive performance on key performance indicators (KPIs).  Hence we are able to identify which factors or attributes are relatively more critical in a market category compared to others and ensures that often limited resources can be allocated to focusing on the main market drivers.


Appendix

 
Gender  China Census  QQSurvey panel 
Male  52%  58% 
Female  48%  42% 
Education level  China Census  QQSurvey panel 
Elementary School  32%  1% 
Junior High School  14%  2% 
Senior High School  21%  8% 
Technical School  12%  9% 
: : :
: : :

Statistical forecasting

 

Statistical modelling

For the majority of reports, Mintel produces five-year forecasts based on an advanced statistical technique known as ‘multivariate time series auto-regression’ using the statistical software package SPSS.

The model is based on historical market size data taken from Mintel’s own market size database and supplemented by published macroeconomic and demographic data from various private and public sources such as the NBS (National Bureau of Statistics of China) and the EIU (The Economist Intelligence Unit).

The model searches for relationships between actual market sizes and a selection of relevant and significant macroeconomic and demographic determinants (independent variables) to identify those predictors having the most influence on the market.

Factors used in a forecast are stated in the relevant report section alongside an interpretation of their role in explaining the development in demand for the product or market in question.

Qualitative insight

At Mintel we understand that historic data is limited in its capacity to act as the only force behind the future state of markets. Thus, rich qualitative insights from industry experts regarding future events that might impact upon various markets play an invaluable role in our post statistical modeling evaluation process.

As a result, the Mintel forecast complements a rigorous statistical process with in-depth market knowledge and expertise to allow for additional factors or market conditions outside of the capacity of the statistical forecast.

 

 

  

The Mintel fan chart

Forecasts of future economic outcomes are always subject to uncertainty. In order to raise awareness amongst our clients and to illustrate this uncertainty, Mintel has introduced a new way of displaying market size forecasts in the form of a fan-chart.

Next to historical market sizes and a current year estimate, the fan chart illustrates the probability of various outcomes for the market value/volume over the next five years.

At a 95% confidence interval, we are saying that 95 out of 100 times the forecast will fall within these outer limits, which we call the best and worst case forecasts. These, based on the statistically driven forecast, are the highest (best case) and lowest (worst case) market sizes the market is expected to achieve.

Over the next five years, the widening bands successively show the developments that occur within 95%, 90%, 70% and 50% probability intervals. Statistical processes predict the central forecast to fall within the darker shaded area which illustrates 50% probability, i.e. a 5 in 10 chance.

A general conclusion: Based on our current knowledge of given historic market size data as well as projections for key macro- and socio-economic measures that were used to create the forecast, we can assume that in 95% of the time the actual market size will fall within the purple shaded fan. In 5% of all cases this model might not be correct due to random errors and the actual market size will fall out of these boundaries.

Weather analogy

To illustrate uncertainty in forecasting in an everyday example, let us assume the following weather forecast was produced based on the meteorologists’ current knowledge of the previous weather condition during the last few days, atmospheric observations, incoming weather fronts etc.

 

 

  

Now, how accurate is this forecast and how certain can we be that the temperature on Saturday will indeed be 15°C?

To state that the temperature in Beijing on Saturday will rise to exactly 15°C is possible but one can’t be 100% certain about that fact.

To say the temperature on Saturday will be between 13°C and 17°C is a broader statement and much more probable.

In general, we can say that based on the existing statistical model, one can be 95% certain that the temperature on Saturday will be between 13°C and 17°C, and respectively 50% certain it will be between about 14.5°C and 15.5°C. Again, only in 5% of all cases this model might not be correct due to random errors and the actual temperature on Saturday will fall out of these boundaries and thus will be below 13°C or above 17°C.

( To learn more about uncertainty in weather forecasts visit: http://research.metoffice.gov.uk/research/nwp/ensemble/uncertainty.html )

Brazil (English)


Brazil Consumer Research

 

For its Brazilian Report series, Mintel uses either online research or face-to-face research to interview consumers representative of the five regions of Brazil, covering all socio-economic groups and ages from 16 to 55+.

  • Our consumer research is based on a random sample of Brazilian consumers from the country’s 5 regions (see details below).
  • Mintel applies a quota-sampling approach with quotas on age, sex, region, income and education in these regions.
  • Our sample data can only be considered as indicative of urban consumers rather than representative of Brazil as a whole. While it does give a proxy of the nation’s attitudes it has been designed to provide comparable, statistically robust results of the most relevant demographic and geographic segments.  
  • For category-focused reports (eg chocolate confectionery), consumers will typically be asked about usage, frequency and location of purchase, consumption occasion, brand usage and a series of attitudinal statements about the category.
  • Lifestyle Reports cover a broader range of attitudinal and behavioural topics.

Our research partners – Lightspeed GMI

 

Mintel partners with Lightspeed GMI to complete online research.  Founded in 1999, and re-branded as Lightspeed GMI in September 2014 (after its acquisition by Kantar in 2011), it delivers online respondents via extensive use of fraud detection and location-verification technology at multiple points in the research cycle, from initial registration through survey fielding and incentive redemption. Lightspeed GMI panellists are profiled on a wide variety of attri­butes to deliver the specific hard-to-reach demographics.  To ensure our surveys are indicative of urban internet users, Mintel sets quotas for each age group, split by gender and weights this to represent Brazil’s major’s regions.


Our research partners – Ipsos Observer Brazil

 

Subsidiary of the French research company IPSOS S.A., Ipsos Observer Brazil is a leading consumer research company in the country. The operation in Brazil is the largest among the “survey based” companies in the country, has more than 850 employees, counts with ISO certification and runs more than 2 million interviews per year.

For Mintel Ipsos research, a CAPI (Computer assisted personal interview) methodology is carried out using a Palm Top device, which allows reduced fieldwork time and continuous review of the survey process.

Sampling and Quality control

In order to achieve the required quotas for age, gender and socio economic group, the Mintel-Ipsos survey combines both fixed-point interviews with door-to-door interviews across all the ten selected cities. The proportion of fixed-point and door-to-door interviews is about equal, although it may change from wave to wave in order to secure the quotas required for each survey. The average proportion of the questioners run is:

Fixed-point

46%

Door-to-door

54%

 

To guarantee accuracy and consistency of the data, a four stage quality control process is applied that involves the following stages:

a.       On the ground supervision of the surveys run to avoid adulteration when applied (Ipsos)

b.      Contact call of a proportion of the people sampled to confirm that the survey has actually been run (Ipsos)

c.       In-depth data analysis and socio-economics crossings of the complete database to check inconsistencies or possible typos by the London consumer research team (Mintel)

d.      Analyst check and comparison with secondary sources and international sources to ensure data is coherent for the category analysed (Mintel).

 


Sample sizes by demographics and geographies

 

Mintel runs consumer research in the 5 regions of Brazil.  Rather than mirror exactly Brazil population distribution, region sample sizes have been selected to provide comparable figures between regions.  Specific region quotas and used and weighted to represent a sample of approximately 1,500 respondents are shown below:

Region

%

 

 

South-East

53

Central-West

7

North

7

North-East

20

South

13

 

 

: :

 

Mintel’s consumer research is also planned to cover a representative number of the socio-economic groups (SEGs) ABC1 and C2DE with the purpose of gathering comparable data among the different SEGs in the country. The classification of the socio economic groups is done by the use of the Criterio Brasil (Socio-economic group segmentation defined by ABEP, the Brazilian Association of Research Companies) based on a combination of educational as well as ownership of properties and consumer goods. The number of interviews used by different socio economic group was defined with the intention to get a representative number of responses across all SEGs:

 

Socio-economic group

%

 

 

ABC1

56.6

C2DE

43.4

 

 

Total

100

 

To also receive representative and comparable age data, Mintel has defined five age segments.

Weighted Age

Male N

Female N

%

 

 

 

 

16-24

150

150

20

25-34

150

150

20

35-44

150

150

20

45-54

150

150

20

55-65

150

150

20

 

 

 

 

: : : :

 


Further Analysis

 

Mintel employs numerous quantitative data analysis techniques to enhance the value of our consumer research. The techniques used vary from one report to another.  Below describes some of the more commonly used techniques.

 

Repertoire Analysis

This is used to create consumer groups based on reported behavior or attitudes. Consumer responses of the same value (or list of values) across a list of survey items are tallied into a single variable. The repertoire variable summarizes the number of occurrences in which the value or values appear among a list of survey items.  For example, a repertoire of brand purchasing might produce groups of those that purchase 1-2 brands, 3-4 brands and 5 or more brands. Each subgroup should be large enough (ie N=75+) to analyse.

 

Cluster Analysis

This technique assigns a set of individual people in to groups called clusters on the basis of one or more question responses, so that respondents within the same cluster are in some sense closer or more similar to one another than to respondents that were grouped into a different cluster.

 

Correspondence Analysis

This is a statistical visualization method for picturing the associations between rows (image, attitudes) and columns (brands, products, segments, etc.) of a two-way contingency table. It allows us to display brand images (and/or consumer attitudes towards brands) related to each brand covered in this survey in a joint space that is easy to understand. The significance of the relationship between a brand and its associated image is measured using the Chi-square test. If two brands have similar response patterns regarding their perceived images, they are assigned similar scores on underlying dimensions and will then be displayed close to each other in the perceptual map.

 

CHAID analysis

CHAID (Chi-squared Automatic Interaction Detection), a type of decision tree analysis, is used to highlight key target groups in a sample by identifying which sub-groups are more likely to show a particular characteristic. This analysis subdivides the sample into a series of subgroups that share similar characteristics towards a specific response variable and allows us to identify which combinations have the highest response rates for the target variable. It is commonly used to understand and visualise the relationship between a variable of interest such as “interest in trying a new product” and other characteristics of the sample, such as demographic composition.

 

Key Driver Analysis

Key driver analysis can be a useful tool in helping to prioritise focus between different factors which may impact key performance indicators (eg satisfaction, likelihood to switch providers, likelihood to recommend a brand, etc). Using correlations analysis or regression analysis we can get an understanding of which factors or attributes of a market have the strongest association or “link” with a positive performance on key performance indicators (KPIs). Hence, we are able to identify which factors or attributes are relatively more critical in a market category compared to others and ensures that often limited resources can be allocated to focusing on the main market drivers.


Statistical Forecasting

 

Statistical modelling

For the majority of Reports, Mintel produces five-year forecasts based on an advanced statistical technique known as ‘multivariate time series auto-regression’ using the statistical software package SPSS.

The model is based on historical market size data taken from Mintel’s own market size database and supplemented by published macroeconomic and demographic data from various private and public sources such as the IBGE (Brazilian Institute of Geography and Statistics) and the EIU (Economist Intelligence Unit).

The model searches for relationships between actual market sizes and a selection of relevant and significant macroeconomic and demographic determinants (independent variables) to identify those predictors having the most influence on the market.

Factors used in a forecast are stated in the relevant report section alongside an interpretation of their role in explaining the development in demand for the product or market in question.

Qualitative insight

At Mintel we understand that historic data is limited in its capacity to act as the only force behind the future state of markets. Thus, rich qualitative insights from industry experts regarding future events that might impact upon various markets play an invaluable role in our post statistical modeling evaluation process.

As a result, the Mintel forecast complements a rigorous statistical process with in-depth market knowledge and expertise to allow for additional factors or market conditions outside of the capacity of the statistical forecast.

 

 

The Mintel fan chart

Forecasts of future economic outcomes are always subject to uncertainty. In order to raise awareness amongst our clients and to illustrate this uncertainty, Mintel has introduced a new way of displaying market size forecasts in the form of a fan chart.

Next to historical market sizes and a current year estimate, the fan chart illustrates the probability of various outcomes for the market value/volume over the next five years.

At a 95% confidence interval we are saying that 95 out of 100 times the forecast will fall within these outer limits, which we call the best and worst case forecasts. These, based on the statistically driven forecast, are the highest (best case) and lowest (worst case) market sizes the market is expected to achieve.

Over the next five years, the widening bands successively show the developments that occur within 95%, 90%, 70% and 50% probability intervals. Statistical processes predict the central forecast to fall within the darker shaded area which illustrates 50% probability ie a 5 in 10 chance.

A general conclusion: Based on our current knowledge of given historic market size data as well as projections for key macro- and socio-economic measures that were used to create the forecast, we can assume that in 95% of the time the actual market size will fall within the purple shaded fan. In 5% of all cases this model might not be correct due to random errors and the actual market size will fall out of these boundaries.

Weather analogy

To illustrate uncertainty in forecasting in an everyday example, let us assume the following weather forecast was produced based on the meteorologists’ current knowledge of the previous weather condition during the last few days, atmospheric observations, incoming weather fronts etc.

 

 

Now, how accurate is this forecast and how certain can we be that the temperature on Saturday will indeed be 15°C?

To state that the temperature in central London on Saturday will rise to exactly 15°C is possible but one can’t be 100% certain about that fact.

To say the temperature on Saturday will be between 13°C and 17°C is a broader statement and much more probable.

In general, we can say that based on the existing statistical model, one can be 95% certain that the temperature on Saturday will be between 13°C and 17°C, and respectively 50% certain it will be between about 14.5°C and 15.5°C. Again, only in 5% of all cases this model might not be correct due to random errors and the actual temperature on Saturday will fall out of these boundaries and thus will be below 13°C or above 17°C.

Metodologia de Pesquisa para o Brasil

Para os relatórios sobre o mercado brasileiro, a Mintel fez uma parceria com a Ipsos Observer Brasil para realizar 1.500 entrevistas face a face, abrangendo as cinco regiões do Brasil, todos os grupos socioeconômicos e faixas etárias a partir de 16 anos até as acima de 55 anos. 

  • Nossa pesquisa sobre o consumidor é feita com amostras aleatórias de consumidores brasileiros nas dez maiores cidades do país (ver detalhes abaixo).
  • A pesquisa representa as dez maiores cidades do Brasil. Mintel utiliza uma amostra com base em quotas, incluindo amostras de idade, sexo, região, renda e nível de ensino nessas cidades.
  • Nossos dados de pesquisa só podem ser considerados como sendo representativos de consumidores dos grandes centros urbanos e não do Brasil como um todo. Embora os dados forneçam uma amostra do comportamento da nação, as pesquisas são desenvolvidas para oferecer resultados comparáveis e estatisticamente sólidos, que representam os segmentos demográficos e geográficos de maior relevância.
  • Para relatórios sobre categorias específicas – por exemplo, produtos de chocolate –, os consumidores, na maioria das vezes, respondem perguntadas sobre uso, frequência e local de compra, ocasião de consumo, uso de marca e uma série de afirmações sobre suas atitudes em relação à categoria.
  • Relatórios sobre estilo de vida abrangem uma gama maior de tópicos sobre atitudes e comportamentos do consumidor.

Tamanho de amostras por demografia e geografia

 

A Mintel realiza pesquisas de consumidor nas dez maiores cidades brasileiras, representando as cinco regiões do país. Em vez de refletir exatamente a distribuição da população brasileira, as amostras das cidades foram selecionadas para fornecerem números comparativos entre as cidades/regiões. Abaixo estão as quotas específicas de cada cidade para a amostra de 1.500 participantes:

Sudeste São Paulo  450
Sudeste Rio de Janeiro 250
Sudeste Belo Horizonte 100
Centro-oeste Brasília 100
Norte Belém 100
Nordeste Fortaleza 100
Nordeste Salvador 100
Nordeste Recife 100
: : :
: : :

A pesquisa de consumidor da Mintel também inclui um número representativo dos grupos socio-econômicos (GSEs) AB, C1, C2 e DE, com o propósito de coletar dados comparáveis entre os diferentes GSEs do país. A classificação dos grupos socioeconômicos são feitas usando o Critério Brasil (segmentação de grupo socioeconômico segundo a definição da ABEP, Associação Brasileira de Empresas de Pesquisa) baseado em uma combinação de nível educacional, bem como posse de propriedades e de bens de consumo. O número de entrevistas para cada grupo socioeconômico foi definido com a intenção de obter um número de respostas representativas de todos GSEs:

Groupo socioeconômico % N
A (A1/A2) 5.0 75
B (B1/B2) 27.1 406
C1 24.5 368
C2 23.9 358
D/E 19.5 293
Total 100 1,500

Para obter, também, dados comparáveis de idade, a Mintel definiu cinco faixas etárias.

Idade Homem Mulher Total %
16-24 150 150  300 20
25-34 150  150  300  20
35-44 150  150  300 20
45-54 150  150  300  20
55-65 150  150  300  20
Total 750  750 1,500 100

Nossa parceira de pesquisa – Ipsos Observer Brazil

 

Subsidiária da companhia de pesquisa francesa, a IPSOS S.A., Ipsos Observer Brasil é uma companhia líder de pesquisa de consumidor no país. Com mais de 850 funcionários, ela possui um certificado da ISO e realiza mais de 2 milhões de pesquisas por ano.

Para a Mintel, a Ipsos utilizou seu método de pesquisa CAPI - Entrevista Pessoal Assistida por Computador - omputer Assisted Personal Interview), que utiliza um computador de mão (Palm Top), o que permite a redução de tempo do trabalho de campo e a revisão contínua do processo de pesquisa.


Amostra e Controle de Qualidade

 

Para poder atingir a quota necessária para idade, sexo e grupos socioeconômicos, a pesquisa Mintel-Ipsos combina tanto entrevistas de ponto-fixo como entrevistas de porta em porta em todas as dez cidades selecionadas. A proporção entre entrevistas de ponto-fixo e porta em porta é quase igual, embora às vezes mude um pouco de pesquisa para pesquisa pois é necessário manter o número da quota para cada pesquisa. A proporção média feita pelos pesquisadores é de:

Ponto-fixo 46%
Porta em porta 54%

Um processo de controle de qualidade de quatro etapas é feito para assegurar a precisão e consistência da informação, seguindo as etapas abaixo:

  • Quando necessário é feita uma supervisão em campo durante as pesquisas para evitar adulterações (Ipsos).
  • Contato telefônico com uma parte das pessoas entrevistadas para confirmar que a pesquisa foi de fato conduzida (Ipsos).
  • Análise profunda de dados e revisão dos dados socioeconômico em todo os banco de dados para verificar inconsistências ou erros ortográficos feitos pelo time de pesquisa do consumidor de Londres (Mintel).
  • Verificação e comparação com fontes secundárias e internacionais, conduzida por analistas para assegurar que os dados sejam coerentes com a categoria analisada (Mintel)

Previsões Estatísticas

 

Modelo estatístico

Para a maioria dos relatórios, a Mintel produz uma previsão de cinco anos baseada em um técnica avançada de estatística conhecida como multivariate time series auto-regression, usando o software de estatística SPSS.

O modelo é baseado em dados históricos de tamanho de mercado originados do banco de dados da própria Mintel e complementados por dados publicados de macroeconomia e demografia de várias fontes privadas e públicas como o IBGE (Instituto Brasileiro de Geografia e Estatísticas) e EIU – Economist Intelligence Unit.

O modelo busca por relações entre o tamanho de mercado real e uma seleção relevante e significativa de fatores macroeconômicos e demográficos determinantes (variáveis independentes) para identificar os fatores de previsão que tem a maior influência no mercado.

Fatores usados na previsão estão presentes nas seções relevantes do relatório junto à uma interpretação de como eles explicam o desenvolvimento em demanda para o produto ou mercado em questão.

Insight Qualitativo

A Mintel entende que dados históricos possuem limites na capacidade de agirem como o único fator determinante do futuro do mercado. Portanto, a compreensão qualitativa dos especialistas da indústria sobre os eventos futuros que poderão impactar sob vários mercados, cumprem um papel de valor inestimável no processo de avaliação pós-modelo.

Assim, a previsão da Mintel completa um processo estatístico rigoroso com conhecimento profundo do mercado e competência profissional que abrangem fatores adicionais ou condições do mercado fora da capacidade da previsão estatística.

Gráfico de leque

Projeções sobre o estado econômico futuro são sempre sujeitas a incertezas. Para gerar consciência entre nossos clientes e ilustrar esse incerteza, a Mintel introduziu uma nova maneira de mostrar projeções de mercado na forma de um gráfico de leque. 

Ao lado de dados históricos de tamanho do mercado e estimativas do ano corrente, o gráfico de leque ilustra a probabilidade de vários resultados para o valor/volume do mercado nos próximos cinco anos.

O intervalo de confiança de 95% significa que 95 de cada 100 vezes a projeção irá cair dentro desse quadro, o qual nós chamamos de melhor e pior casos de projeção. Esses, baseado na projeção estatística, são os maiores (melhor caso) e os menores (pior caso) tamanhos de mercado antecipados.

Nos próximos cinco anos, as faixas alargadas mostram sucessivamente os desenvolvimentos que ocorrem dentro dos intervalos de probabilidade de 95%, 90%, 70% e 50%. Métodos estatísticos preveem que a projeção central cairá dentro da área mais escura, que ilustra 50% de probabilidade, ou seja, uma chance de 5 em 10.

Conclusão geral: Baseado no nosso conhecimento dos dados históricos de mercado, bem como das projeções para medidas chave macro- e socioeconômicas que foram utilizadas para criar a projeção, podemos presumir que 95% das vezes o tamanho real do mercado estará entre o leque roxo. Em 5% dos casos esse modelo pode não estar correto devido a erros aleatórios e o tamanho real do mercado cairá fora desses limites.

Canada

Mintel is an independent market analysis company that prides itself on supplying objective information on a whole range of markets and marketing issues.

There are five main sources of research that are used in the compilation of Mintel reports:

  • Consumer research
  • Social media research
  • Desk research
  • Trade research
  • Statistical forecasting

Mintel reports are written and managed by analysts with experience in the relevant markets.


Consumer research

 

Exclusive and original quantitative consumer research is commissioned for  Mintel reports. In addition, qualitative research is also undertaken for reports in the form of online discussion groups.. The research brings an up-to-date and unique insight into topical issues of importance.

Consumer research is conducted among a nationally representative sample of internet users in Canada and is carried out by Lightspeed GMI. The results are only available in Mintel reports. Note that Mintel’s exclusive research is conducted online in both English and French.

Sampling and weighting

 

Online Surveys

 

Lightspeed GMI

 

Founded in 1996, Lightspeed GMI’s double opt-in online consumer panel contains approximately 187,000 Canadian consumers. Lightspeed GMI recruits its panelists through many different sources including web advertising, permission-based databases and partner-recruited panels. 

 

Age groups by gender

%

N

Male, 18-24

8.80%

176

Male, 25-34

9.90%

198

Male, 35-44

9.00%

181

Male, 45-54

9.60%

191

Male, 55-64

7.40%

148

Male, 65+

4.70%

94

 

 

 

: : :
: : :

Mintel also sets quotas on region and household income. Specific quotas for a sample of 2,000 adults aged 18+ are shown below:

Region

%

N

Ontario 

39.00%

781

Quebec

22.20%

443

British Columbia 

13.50%

270

Alberta

10.70%

213

Saskatchewan

3.80%

77

Manitoba

4.40%

88

Atlantic Provinces  (New Brunswick, Newfoundland/ Labrador, Nova Scotia, Prince Edward Island)

6.40%

128

: : :

*Mintel does not include rural regions such as the Yukon or the Northwest Territories (including Nunavut) in its research. Thus the consumer research data does not reflect opinions and behaviours of the population living in those areas.

Household income

%

N

Less than $25,000

10.70%

215

$25,000 - $49,999

15.60%

312

$50,000 - $74,999

20.80%

416

$75,000 - $99,999

24.30%

485

$100,000 and over

28.60%

572

Total

100%

2000

Secondary Data Analysis

In addition to exclusively commissioned surveys, Mintel gathers syndicated data from the most respected consumer research firms. This allows Mintel analysts to form objective and cohesive analyses of consumer attitudes and behaviour.

 

Qualitative Research

Revelation by FocusVision

FocusVision provides Mintel with qualitative bulletin board software. This allows the creation of Internet-based, ‘virtual’ venues where participants recruited from Mintel’s online surveys gather and engage in interactive, text-based discussions led by Mintel moderators

Further Analysis

Mintel employs numerous quantitative data analysis techniques to enhance the value of our consumer research. The techniques used vary form one report to another.  Below describes some of the more commonly used techniques.

Repertoire Analysis

This is used to create consumer groups based on reported behaviour or attitudes. Consumer responses of the same value (or list of values) across a list of survey items are tallied into a single variable. The repertoire variable summarises the number of occurrences in which the value or values appear among a list of survey items.  For example, a repertoire of brand purchasing might produce groups of those that purchase 1-2 brands, 3-4 brands and 5 or more brands. Each subgroup should be large enough (ie N=75+) to analyse.

Cluster Analysis

This technique assigns a set of individual people in to groups called clusters on the basis of one or more question responses, so that respondents within the same cluster are in some sense closer or more similar to one another than to respondents that were grouped into a different cluster.

Correspondence Analysis

This is a statistical visualisation method for picturing the associations between rows (image, attitudes) and columns (brands, products, segments, etc.) of a two-way contingency table. It allows us to display brand images (and/or consumer attitudes towards brands) related to each brand covered in this survey in a joint space that is easy to understand. The significance of the relationship between a brand and its associated image is measured using the Chi-square test. If two brands have similar response patterns regarding their perceived images, they are assigned similar scores on underlying dimensions and will then be displayed close to each other in the perceptual map. 

CHAID analysis

CHAID (Chi-squared Automatic Interaction Detection), a type of decision tree analysis, is used to highlight key target groups in a sample by identifying which sub-groups are more likely to show a particular characteristic. This analysis subdivides the sample into a series of subgroups that share similar characteristics towards a specific response variable and allows us to identify which combinations have the highest response rates for the target variable. It is commonly used to understand and visualise the relationship between a variable of interest such as “interest in trying a new product” and other characteristics of the sample, such as demographic composition.

Key Driver Analysis

Key driver analysis can be a useful tool in helping to prioritise focus between different factors which may impact key performance indicators (eg satisfaction, likelihood to switch providers, likelihood to recommend a brand, etc). Using correlations analysis or regression analysis provides an understanding of which factors or attributes of a market have the strongest association or “link” with a positive performance on key performance indicators (KPIs). Hence, factors or attributes are identified which are relatively more critical in a market category compared to others and ensures that often limited resources can be allocated to focusing on the main market drivers.

 

TURF Analysis

TURF (Total Unduplicated Reach & Frequency) analysis identifies the mix of features, attributes, or messages that will attract the largest number of unique respondents.  It is typically used when the number of features or attributes must be or should be limited, but the goal is still to reach the widest possible audience. By identifying the Total Unduplicated Reach, it is possible to maximize the number of people who find one or more of their preferred features or attributes in the product line. The resulting output from TURF is additive, with each additional feature increasing total reach. The chart is read from left to right, with each arrow indicating the incremental change in total reach when adding a new feature. The final bar represents the maximum reach of the total population when all shown features are offered. 


Social Media Research

 

To complement its exclusive consumer research, Mintel tracks and analyses social media data for inclusion in selected reports. Using Infegy’s Atlas software, Mintel 'listens in' on online conversations across a range of social platforms including Facebook, Twitter, consumer forums and the wider web.

Atlas provides rich consumer insight via the analysis of commentary posted publicly on the internet. The system performs comprehensive and broad collection of data from millions of internet sources, working to ensure a faithful and extensive sampling of feedback from the widest range of individuals. The dataset contains commentary posted in real time, as well as a substantial archive dating back through 2007.


Trade Research

 

Informal

Trade research is undertaken for all reports. This involves contacting relevant players in the trade, not only to gain information concerning their own operations, but also to obtain explanations and views of the strategic issues pertinent to the market being researched. Such is Mintel’s concern with accuracy that draft copies of reports are sent to industry representatives, to get their feedback and avoid any misrepresentation of the market. These comments are incorporated into reports prior to final publication.

Formal

Internally, Mintel’s analysts undertake extensive trade interviews with selected key experts in the field for the majority of reports. The purpose of these interviews is to assess key issues in the market place in order to ensure that any research undertaken takes these into account.

In addition, using experienced external researchers, trade research is undertaken for some reports. This takes the form of full trade interview questionnaires and direct quotes are included in the report and analysed by experts in the field. This gives a valuable insight into a range of trade views of topical issues.


Desk Research

 

Mintel has an internal team of market analysts who monitor: government statistics, consumer and trade association statistics, manufacturer sponsored reports, annual company reports and accounts, directories, press articles from around the world and online databases. The latter are extracted from hundreds of publications and websites, both Canada and overseas. All information is cross-referenced for immediate access.

Data from other published sources are the latest available at the time of writing the report.

This information is supplemented by an extensive library of Mintel’s reports produced since 1972 globally and added to each year by the 500+ reports which are produced annually.

In addition to in-house sources, researchers also occasionally use outside libraries such as Statistics Canada and the Canadian Grocer. Other information is also gathered from store and exhibition visits across Canada, as well as using other databases within the Mintel Group, such as the Global New Product Database (GNPD), which monitors FMCG sales promotions.

All analysts have access to Mintel’s Market Size and Macroeconomic Databases – a database containing many areas of consumer spending and retail sales as well as macroeconomic and demographic factors which impinge on consumer spending patterns..

The database is used in conjunction with an SPSS forecasting program which uses weighted historical correlations of market dynamics, with independent variables, to produce future spending scenarios.


Statistical Forecasting

 

Statistical modelling

For the majority of reports, Mintel produces five-year forecasts based on an advanced statistical technique known as ‘multivariate time series auto-regression’ using the statistical software package SPSS.

Historical market size data feeding into each forecast are collated in Mintel’s own market size database and supplemented by macro- and socio-economic data sourced from organisations such as Statistics Canada, The Bank of Canada, The Conference Board of Canada and the Economist Intelligence Unit. 

Within the forecasting process, the model searches for, and analyses relationships between, actual market sizes and a selection of key economic and demographic factors (independent variables) in order to identify those predictors having the most influence on the market.

Factors used in a forecast are stated in the relevant report section alongside an interpretation of their role in explaining the development in demand for the product or market in question.

Qualitative insight

At Mintel we understand that historic data is limited in its capacity to act as the only force behind the future state of markets. Thus, rich qualitative insights from industry experts regarding past and future events that may impact the market play a crucial role in our post statistical modeling evaluation process.

As a result, the Mintel forecast allows for additional factors or market conditions outside of the capacity of the data analysis to impact the market forecast model, using a rigorous statistical process complemented by in-depth market knowledge and expertise.

 

 

The Mintel fan chart

Forecasts of future economic outcomes are always subject to uncertainty. In order to raise awareness amongst our clients and to illustrate this uncertainty, Mintel has introduced a new way of displaying market size forecasts in the form of a fan-chart.

Next to historical market sizes and a current year estimate, the fan chart illustrates the probability of various outcomes for the market value/volume over the next five years.

At a 95% confidence interval, we are saying that 95 out of 100 times the forecast will fall within these outer limits, which we call the best and worst case forecasts. These, based on the statistically driven forecast, are the highest (best case) and lowest (worst case) market sizes the market is expected to achieve.

Over the next five years, the widening bands successively show the developments that occur within 95%, 90%, 70% and 50% probability intervals. Statistical processes predict the central forecast to fall within the darker shaded area which illustrates 50% probability, i.e. a 5 in 10 chance.

In general, based on our current knowledge of given historic market size data as well as projections for key macro- and socio-economic measures that were used to create the forecast, we can assume that in 95% of the time the actual market size will fall within the purple shaded fan. In 5% of all cases this model might not be correct due to random errors and the actual market size will fall out of these boundaries.

Weather analogy

To illustrate uncertainty in forecasting in an everyday example, let us assume the following weather forecast was produced based on the meteorologists’ current knowledge of the previous weather condition during the last few days, atmospheric observations, incoming weather fronts etc.

 

 

Now, how accurate is this forecast and how certain can we be that the temperature on Saturday will indeed be 15°C?

To state that the temperature in central London on Saturday will rise to exactly 15°C is possible but one can’t be 100% certain about that fact.

To say the temperature on Saturday will be between 13°C and 17°C is a broader statement and much more probable.

In general, we can say that based on the existing statistical model, one can be 95% certain that the temperature on Saturday will be between 13°C and 17°C, and respectively 50% certain it will be between about 14.5°C and 15.5°C. Again, only in 5% of all cases this model might not be correct due to random errors and the actual temperature on Saturday will fall out of these boundaries and thus will be below 13°C or above 17°C.

(To learn more about uncertainty in weather forecasts visit: http://research.metoffice.gov.uk/research/nwp/ensemble/uncertainty.html)