Transcript Slide 1
Consumer Survey Datasets What are they? How do you interpret the data? A brief look at Mediamark Reporter Online … What are they? They are the tabulated, numeric results of huge surveys of many consumers. The major producers are: • Simmons Market Research Bureau (SMRB) • Mediamark Research Inc. (MRI) Simmons Mediamark Research & Intelligence Types of Data Collected: • • • • • • Demographics (lots of them!) Product usage Brand preferences Media usage Media preferences Lifestyle (psychographics, mostly in Simmons) Methodology: • Twice-a-year in-depth surveys of 20-30 thousand adult consumers. Includes: – ~10 days of keeping a ‘diary’ of all activities, purchases, consumption, media, etc. – In-person fieldworker interview of household • Data from surveys is ‘projected’ onto annual Census Bureau Population Estimates Limitations: • Separate studies for minors. • Best for consumer products, not so much for B2B products. • Academic libraries are only allowed to buy outof-date data (1-year-old for Mediamark, 2-yearsold for Simmons) • Mediamark is online, but only provides preformatted, ‘canned’ crosstab reports • Simmons is cd-rom data (no online access), but has a powerful interface that allows lots of customization Typical questions: • Demographically, who are the best buyers or customers of product x or brand y? • Demographically, who should the manufacturer target in order to increase market share? • How do I find psychographics for this target group? (Simmons only) Let’s briefly look at how to retrieve a crosstab in Mediamark Reporter, which is the easiest to use. Starting from the Libraries’ homepage, here’s how to get to it … Be sure to use the ‘online version’ of Mediamark Reporter. (The cd-rom version has much older data.) You’ll need this information when you first register for an account … It’s also repeated on a screen later in this Powerpoint file. If you’ve already registered, just go directly to www.mriplus.com/. Registration: Use your [email protected] email to register for an account – ‘marquette’ must be spelled out in full. Both email address and the password you create are case sensitive!! After registering, you’ll receive an email message with an activation link from Mediamark. Login at left or Register for Free to begin using MRI+. This account is free to you: the Libraries pay for the subscription to the database … By the way, the database works best with Internet Explorer. Mediamark registration • Use your Marquette email address & create a password – Use the format [email protected]. ‘marquette’ must be spelled out in full in the domain name (@marquette.edu) – Password does NOT need to be your eMarq password • Use the MU Libraries’ address & phone number: 1355 W. Wisconsin Ave. Milwaukee, WI 53233 414.288.7556 • Go to your eMarq account to complete registration Mediamark Reporter Once you’ve logged in, click on this link to launch the Reporter … A new browser window will be launched, so be sure that your browser allows pop-ups. Volume Types: Spring 2008 Product Product: profiles who uses different product types, brands, etc. Also includes profiles of behaviors (e.g. leisure activities / hobbies, sports participation, voting, lottery, etc.) Media: profiles the media used by different demographic groups (e.g. by age, occupational group, educational level, etc.). Includes the media quintiles. Magazine Qualitative: provides a qualitative measure of how a magazine is used by readers (e.g. where do they read it, with how much attention and thoroughness, do they take any action after reading it). Magazine Cumulative: provides 4-issue reach and frequency data for each title. Click in the Report Volume selection box to choose a report. This will trigger the appearance of another selection box … ►In general, it’s better to start with the Spring datasets because they are usually larger. Scroll through the report categories to choose a product type, media user type (demographic group), etc. You can also search by keyword. ► But beware! Keyword searching works best with brand names and other distinctive single words. Until you are familiar with the database’s terminology, it can be less useful. The Submenus/Search Results AND Crosstab display… The left half of the screen shows the category submenus (or keyword search results). This is an awkward screen that tries to combine two different things … The Submenus/Search Results AND Crosstab display… The right half automatically shows a crosstab for the first selection on the left. The Crosstab on the right always reflects the variables highlighted on the left. “Beware the Asterisk, er, the Jabberwock, my son!” Whenever you see an asterisk (*), be aware that the sample size for this question in the survey was too small for statistical reliability. Use these data with caution. Educ: did not graduate HS * 33,991 When you change the selection in the Category box, the variables in the Target box will also change … And so will the crosstab. Athletic Shoes – Brands Bought Here you see a new Crosstab, reflecting the selections on the left. So now the question is: how do you interpret these numbers?! Athletic Shoes, Brands bought last 12 months: Nike Total Proj Percent Percent ‘000 ‘000 Across Down Index Total 222,210 40,091 18.0 100.0 100 Age 18-24 28,312 7,558 26.7 18.9 148 Age 25-34 39,835 10,868 27.1 151 Let’s enlarge this a bit 27.3 … Age 35-44 43,118 10,661 24.7 26.6 137 Age 45-54 43,296 15.2 variables 16.4 Then we have6,589 three calculated …84 Age 55-64 31,707 2,800 8.8 7.0 49 Age 65+ 35,941 1,615 4.5 4.0 25 Men 18-34 34,257 9,716 28.4 24.2 157 Men 18-49 66,661 16,999 25.5 42.4 141 Men 25-54 62,432 14,231 22.8 35.5 126 Women 18-34 33,891 8,710 25.7 21.7 142 Women 18-49 67,280 15,806 23.5 39.4 130 Women 25-54 13,887 21.8 34.6 121 These data are 63,818 from the These data are projected, Census Bureau’s Annual Population Estimate Survey or extrapolated, from the Mediamark survey results. What is the Index? Athletic last 12ofmonths: TheShoes, Index is Brands a measurebought of the probability finding a Nike Total group Proj Percent user in a specific small, relative to the Percent ‘000larger, Across probability of finding‘000 a user in the general Down Index Total 222,210 100 population. Also called the40,091 ‘propensity to18.0 use’ … 100.0 Age 18-24 28,312 7,558 26.7 18.9 Age 25-34 39,835 10,868 27.3 27.1 Age 35-44 43,118 10,661 24.7 26.6 Age 45-54 43,296 6,589 15.2 16.4 Age 55-64 31,707 2,800 8.8 7.0 Age 65+ 35,941 1,615 4.5 4.0 Men 18-34 34,257 9,716 28.4 24.2 English Men 18-49 66,661 ‘translation’: 16,999 25.5 42.4 Men 25-54 62,432 14,231 22.8 35.5 People who bought Nike athletic past Women 18-34 33,891 8,710 shoes 25.7in the21.7 Women 18-49 15,806 23.5 12 months are67,280 51% more likely to be in the 39.4 25–34 Women 25-54 63,818 13,887 21.8 34.6 age range than the general population. 148 151 137 84 49 25 157 141 126 142 130 121 Athletic Shoes, Brands bought‘translation’? last 12 months: Nike How do you get that English Total Proj Percent Percent You subtract 100 from the value. ‘000 ‘000 Across Down If the answer is positive, you say ‘more18.0 likely’.100.0 Total 222,210 40,091 Age 18-24 28,312 7,558 26.7 18.9 Age 25-34 39,835 10,868 27.3 27.1 Age 35-44 43,118 10,661 24.7 26.6 Age 45-54 43,296 6,589 15.2 16.4 Age 55-64 31,707 2,800 8.8 7.0 Age 65+ 35,941 1,615 4.5 4.0 Men 18-34 34,257 9,716 28.4 24.2 English Men 18-49 66,661 ‘translation’: 16,999 25.5 42.4 Men 25-54 62,432 14,231 22.8 35.5 People who bought Nike athletic past Women 18-34 33,891 8,710 shoes 25.7in the21.7 Women 18-49 15,806 23.5 12 months are67,280 51% more likely to be in the 39.4 25–34 Women 25-54 63,818 13,887 21.8 34.6 151 – 100 = + 51% age range than the general population. Index 100 148 151 137 84 49 25 157 141 126 142 130 121 Athletic Shoes, Brands bought last 12 months: Nike Total Proj Percent Percent ‘000 ‘000 Across Down But if the answer222,210 is negative ... Total 40,091 18.0 100.0 Age 18-24 28,312 7,558 26.7 18.9 Age 25-34 39,835 10,868 27.3 27.1 Age 35-44 43,118 10,661 24.7 26.6 Age 45-54 43,296 6,589 15.2 16.4 Age 55-64 31,707 2,800 8.8 7.0 Age 65+ 35,941 1,615 4.5 4.0 Men 18-34 34,257 9,716 28.4 24.2 English Men 18-49 66,661 ‘translation’: 16,999 25.5 42.4 Men 25-54 62,432 14,231 22.8 35.5 People who bought Nike athletic past Women 18-34 33,891 8,710 shoes 25.7in the21.7 Women 18-49 39.4 12 months are67,280 51% less15,806 likely to be23.5 in the 55–64 Women 25-54 63,818 13,887 21.8 34.6 49 – 100 = – 51% age range than the general population. Index 100 148 151 137 84 49 25 157 141 126 142 130 121 Index numbers • Index value 100 = dead average for total general population • However, there is really an average ‘range’: 90 – 110, or 85 – 115 • Therefore: index values ≥ 110 are significant index values ≤ 90 are significant The average range for index values: Index = 100 Index = 90 Index = 110 On this bell curve, all the average probability people fall between the 90 and 110 index lines. The people with greater and lesser probability, statistically speaking, are in the outside triangular areas. Use high and low index #’s to determine: • Demographics of ‘best’ or core customers • Demographics of low-use customers • Demographics of group to which you might aim an ad campaign. Index numbers are NOT enough! Index numbers are always and only comparative, NOT quantitative: they compare data for a small demographic group to the same data for the general population. Classic example: Asian-Americans have high index values (180-200) for the purchase of Hondas and Toyotas. However, they only constitute ~4.3% of the total population of the US (in 2005). Therefore, they cannot constitute a significant portion of the total number of the buyers of imported Japanese cars. Vertical % or % down numbers are quantitative! Use % down/vertical % to see how many of your target group (column variable) also fall into a specific, smaller demographic (row variable). Mediamark uses the label ‘% down’, and Simmons uses the label ‘vertical %’ – different names for the exact same value. Athletic Shoes, Brands bought last 12 months: Nike Total Proj Percent Percent ‘000 ‘000 Across Down Index Total 222,210 40,091 18.0 100.0 100 Age 18-24 28,312 7,558 26.7 18.9 148 Good English ‘translations’: Age 25-34 39,835 10,868 27.3 27.1 151 Among buyers of Nike athletic shoes in the last 12 26.6 months, 137 Age 35-44 43,118 10,661 24.7 Age 45-54 43,296 84 42.9% were men aged 18–49.6,589 15.2 16.4 Age 55-64 31,707 2,800 8.8 7.0 49 Men aged 18–49 constitute 42.4% of all buyers of Nike Age 65+ 35,941 1,615 4.5 4.0 25 athletic shoes in the past 12 months. Men 18-34 34,257 9,716 28.4 24.2 157 Men 18-49 66,661 16,999 25.5 42.4 141 Men 25-54 62,432 14,231 22.8 35.5 126 Women 18-34 33,891 8,710 25.7 21.7 142 Women 18-49 16,999 67,280 15,806 23.5 39.4 130 100 = 42.4% Women 25-54 63,818X 13,887 21.8 34.6 121 40,091 Index values + vertical %’s = • A good, reliable profile of best customers • A good profile of customer groups to target with an ad campaign The following slide will show a selection of rows on Nike buyers, mostly with higher index values and higher %’s down. Total ‘000 Proj ‘000 % Across % Down Index Men age 18-49 66,661 16,999 25.5 42.4 141 Women age 18-49 67,280 15,806 23.5 39.4 130 Child Age: 6-11 yrs 42,307 10,491 24.8 26.2 137 Child Age: 12-17 yrs 44,788 10,803 24.1 26.9 134 Marital status: never 56,008 13,807 24.7 34.4 137 Marital status: married 124,341 20,856 16.8 52.0 93 Race: Black/Afr. Am. 25,584 7,812 30.5 19.5 169 170,909 26,034 15.2 64.9 84 Race: White What is missing here? and explanations CensusReasoning Region: South 81,150 for these 15,080data …Context 18.6 … 37.6 103 Census Region: West 49,854 8,348 16.7 20.8 93 TV (total) Quintile I 44,469 6,412 14.4 16 80 Comedy Central 48,028 9,218 19.2 28.1 127 ESPN 66,038 14,700 22.3 36.7 123 MTV 35,077 10,114 28.8 25.2 160 Football, Pro – weekend 54,037 12,544 23.2 31.3 129 Total ‘000 Proj ‘000 % Across % Down Index Men age 18-49 66,661 16,999 25.5 42.4 You have to figure out the reasons for the data ... Women age 18-49 67,280 15,806 23.5 39.4 141 Child Age: 6-11 yrs 42,307 10,491 24.8 26.2 137 Race: White 170,909 26,034 15.2 64.9 84 Census Region: South 81,150 15,080 18.6 37.6 103 Census Region: West 49,854 8,348 16.7 20.8 93 TV (total) Quintile I 44,469 6,412 14.4 16 80 Comedy Central 48,028 9,218 19.2 28.1 127 ESPN 66,038 14,700 22.3 36.7 123 MTV 35,077 10,114 28.8 25.2 160 Football, Pro – weekend 54,037 12,544 23.2 31.3 129 130 For example, look at the numbers for 10,803 people in the 24.1 two Census26.9 Child Age: 12-17 yrs 44,788 134 Regions, South and West. Nike buyers are less likely to live in the West Marital status: never 56,008 13,807 24.7 34.4 137 (index=93; %down=20.8) , than in the South (index=103; %down=37.6). Marital status: married 124,341 20,856 16.8 When you check the crosstabs for competitor brands, you find 52.0 that in 93 Race: Black/Afr. Am. shoe brand 25,584is much7,812 30.5... 19.5 169 the West, the Vans more popular Total ‘000 Proj ‘000 % Across % Down Index Men age 18-49 66,661 16,999 25.5 42.4 You have to figure out the reasons for the data ... Women age 18-49 67,280 15,806 23.5 39.4 141 Child Age: 6-11 yrs 42,307 10,491 24.8 26.2 137 Child Age: 12-17 yrs 44,788 10,803 24.1 26.9 134 Marital status: never 56,008 13,807 24.7 34.4 137 Marital status: married 124,341 20,856 16.8 52.0 93 Race: Black/Afr. Am. 25,584 7,812 30.5 19.5 169 Census Region: West 49,854 8,348 16.7 20.8 93 130 If you look at the data about people who’ve never married Race: White 170,909 (index=137, %down=34.4) and focus26,034 especially on15.2 the index64.9 value, 84 Census South 37.6 103 you Region: might think that Nike81,150 buyers are15,080 mostly single.18.6 However, when you look at the numbers for people who are married, TV (total) Quintile I 44,469 6,412 14.4 16 you see that married people do buy Nike’s. And in fact, they Comedyjust Central 48,028 9,218(%down=52.0) 19.2 28.1 represent over half of all Nike buyers even 14,700 22.3 36.7how though their ESPN likelihood of66,038 buying is low-average (index=93). So do you explainMTV this? Look at the numbers 35,077 10,114 for people 28.8with children 25.2 in the household … Who do you think they’re likely buying for?! Football, Pro – weekend 54,037 12,544 23.2 31.3 80 127 123 160 129 Sex and Race: For these two categories of variable, be aware of the underlying demographic composition of the country. Only when quantitative values (%’s down) diverge from the underlying demographics are they truly of interest. • For sex (gender), the US is close to 50% men and 50% women (except among the elderly). • For race or ethnic group, we are roughly* as follows: White African American Asian/Pacific Islander Native American Hispanic ~ 74 % ~ 12 % ~5% ~1% ~ 15 % * This is an extract from the American Community Survey data for 2006-08. See the Census Bureau’s website for more complete data and explanations. Total ‘000 Proj ‘000 % Across % Down Index Men age 18-49 66,661 16,999 25.5 42.4 141 Women age 18-49 67,280 15,806 23.5 39.4 130 Child Age: 6-11 yrs 42,307 10,491 24.8 26.2 137 44,788 and10,803 24.1 (gender) Nike buyers: 26.9 134 34.4 137 Child Age: 12-17 yrs Sex Marital status: never 56,008 13,807 24.7 There is little difference in the Nike buying patterns for men (42.4%) 124,341 20,856 16.8 52.0 and women (39.4%) in the age range 18-49. Marital status: married 93 Race: Black/Afr. Am. 25,584 7,812 30.5 19.5 169 Race: White 170,909 26,034 15.2 64.9 84 Census Region: South 81,150 15,080 18.6 37.6 103 Census Region: West 49,854 8,348 16.7 20.8 93 TV (total) Quintile I 44,469 6,412 14.4 16 80 Comedy Central 48,028 9,218 19.2 28.1 127 ESPN 66,038 14,700 22.3 36.7 123 MTV 35,077 10,114 28.8 25.2 160 Football, Pro – weekend 54,037 12,544 23.2 31.3 129 Total ‘000 Proj ‘000 % Across % Down Index Men age 18-49 66,661 16,999 25.5 42.4 141 Women age 18-49 67,280 15,806 23.5 39.4 130 Child Age: 6-11 yrs 42,307 10,491 24.8 26.2 137 Child Age: 12-17 yrs 44,788 10,803 24.1 26.9 134 Marital status: never 56,008 13,807 24.7 34.4 137 Marital status: married 124,341 20,856 16.8 52.0 93 Race: Black/Afr. Am. 25,584 7,812 30.5 19.5 169 Race: White 170,909 26,034 15.2 64.9 84 Census Region: South 81,150 15,080 18.6 37.6 103 16.7 20.8 93 Census Region: West Race49,854 and Nike8,348 buyers: TVrace, (total)there Quintile 44,469 14.4buyers are16 For is aI difference of note: 6,412 19.5% of Nike African-Americans, proportion 9,218 than the underlying Comedy Central a larger 48,028 19.2 28.1 composition would for. And only 64.9% are22.3 white, a smaller ESPNaccount 66,038 14,700 36.7 proportion than accounted for by the underlying demographics. 80 127 123 MTV 35,077 10,114 28.8 25.2 160 Football, Pro – weekend 54,037 12,544 23.2 31.3 129 % across / horizontal % • Horizontal %’s are also quantitative • Use them to see how many of your row group also fall into your target (column) group. • This value is used mostly in media planning (e.g. when buying ads, you’ll need to know what percentage of a magazine’s readership fits into a particular demographic category.) • Know how to read this value, if only so that you can readily distinguish it from the % down / vertical % Athletic Shoes, Brands bought last 12 months: Nike Total Proj Percent Percent ‘000 ‘000 Across Down Total 222,210 40,091 18.0 100.0 Good English ‘translations’: Age 18-24 28,312 7,558 26.7 18.9 Age 25-34 39,835 10,868bought 27.3 27.1 Among all men aged 18–49, 25.5% Nike athletic Age 35-44 43,118 10,661 24.7 26.6 shoes in the past 12 months. Age 45-54 43,296 6,589 15.2 16.4 Buyers of Nike athletic shoes in2,800 the past 128.8 months 7.0 Age 55-64 31,707 Age 65+ 1,615 4.0 constitute 25.5% of35,941 all men aged 18–49. 4.5 Men 18-34 34,257 9,716 28.4 24.2 66,661 16,999 Men 18-49 25.5 42.4 Men 25-54 62,432 14,231 22.8 35.5 Women 18-34 33,891 8,710 25.7 21.7 Women 18-49 16,999 67,280 15,806 23.5 39.4 Women 25-54 63,818 13,887 21.8 34.6 X 100 = 25.5% 66,661 Index 100 148 151 137 84 49 25 157 141 126 142 130 121 The % down and % across values tell you about the relationship (or proportions) between three groups of people: those in the row variable, those in the column variable, and those in both. Let’s try looking at these two calculated variables, % down and % across, another way. Column: Bought Nike’s in last 12 mos. (40,091) X: (16,999) Row: Men aged 18–49 (66,661) Above is an ordinary Venn diagram representing two of the variables, and their cross-section (X). (Apologies: the circles are only very roughly proportional …) Now let’s look at how % down and % across are calculated for these variables. Column: Bought Nike’s in last 12 mos. (40,091) X: (16,999) % Down: 16,999 40,091 X 100 = 42.4 % Here you see that the crosssection of the two variables (X) represents 43% of the total for the column variable (people who bought Nike’s in the last 12 months). X: (16,999) Here you see that the crosssection of the two variables (X) represents 21% of the total for the row variable (men aged 18–49). Row: Men aged 18–49 (66,661) % Across: 16,999 66,661 X 100 = 25.5 % And that’s our introduction to consumer survey datasets! What they are, and how to interpret the numbers … For more help with using consumer survey datasets (Mediamark or Simmons), please contact us at the Information Desk. 414.288.7556 Or, fill out the research consultation request form at: http://www.marquette.edu/library/research/consultations.shtml