Volume II, April 2009

POV MRI Draws on “Currency” Study to Improve
MRI Starch Metrics


Dr. Mickey Galin
Since acquiring Starch in 2008, improving its relevancy as a metric for use by advertisers to evaluate return-on-investment has been a key MRI focus. In this interview, Dr. Mickey Galin, MRI Starch SVP Research, discusses how MRI is leveraging its national Survey of the American Consumer to strengthen Starch’s ad exposure metrics.

Q. How does MRI plan to use the Survey of the American Consumer to make MRI Starch ad readership metrics more relevant when evaluating a print campaign’s ROI?

MRI’s national Survey of the American Consumer is the gold standard for consumer profiling and average issue audience measurement. Because of our extensive knowledge about the readers of each title we measure, MRI is in a unique position to bring important learning to the MRI Starch ad readership metrics.

Starch data are based on relatively small sample sizes (125 respondents per 25 ads in an issue) gathered through Internet panels; not on a probability sample of each title’s readership. This Internet-based methodology allows Starch to field close to 650 studies across a given year. However, because Internet panels are not representative of the general population there are data implications. MRI’s Survey of the American Consumer, on the other hand, is based on a strict area probability sample; it gives reliable details about the profile of a title’s audience. We are in the process of “conforming”-- or adjusting--MRI Starch ad readership metrics to each title’s total audience profile in order to get a truer indication of how many readers saw a particular ad.

Q. How is this achieved?

It’s a three-step process:

  • First, we statistically determine which variables (both reader and demographic) have the greatest impact on the probability that a reader will see an advertisement.
  • Second, we use audience data from MRI’s Survey of the American Consumer to determine how these variables are distributed among readers of a particular title.
  • Third, we weight the raw MRI Starch ad readership scores according to the distribution of these key variables in each title (using statistical modeling).

Q. Can you give an example?

Sure. Let’s take the Subway ad that ran in the 12/29/08 issue of People. And, let’s assume that one of the core variables that drive ad noting is frequency of reading, i.e. how many issues out of an average of four are read or looked into. Since we know what percentage of People readers are frequent readers--as well as how many People readers correspond with the other variables that drive ad readership--we can conform the Subway “ad effectiveness” scores (noting, associated, read some, read most, actions taken) to reflect this distribution.

12/29/08
magazine
  • Subway “raw” Noted Score: 68%
  • Subway “conformed” Noted Score: 71%

As a result of this conforming procedure, we estimate that an additional three percent of People readers saw the Subway ad in this issue than the raw MRI Starch “Ad Noting” scores suggests.

Q. Why is this important?

It’s all about getting more accurate ROI measures. Historically, a magazine’s total readership was used as a proxy for ad exposure. The greater a publication’s audience, the greater the number of consumers who had the opportunity to see an ad within a given issue. MRI Starch data, which at its most basic level measures how many readers see an ad, can help move the needle from opportunity to see to actually see an ad. But we want to ensure our ad metrics are as accurate as possible, since they are being used by agencies and advertisers to evaluate print campaigns. By leveraging MRI’s knowledge about reader behavior for the titles measured by MRI Starch, we can get closer to a true ROI metric. In the above People magazine example, for instance, a 3% difference between the raw and “conformed” ad noting score might not seem particularly dramatic. But, when you consider that People is read by more than 40 million consumers, the difference represents more than 1 million people. That’s a lot of consumers who, until now, People would not get credit from Subway for reaching!

Q. Will the conformed scores typically be higher or lower than historical Starch data?

It depends on how closely the reader profiles between a given MRI Starch survey and the MRI Survey for a given title match. The scores can be higher, lower or even the same as the unconformed Starch data based on how closely the two reader profiles match, with the MRI Survey acting as the anchor since it is representative of the adult U.S. population.

Q. When will this improvement happen? Will there be an extra cost for the conformed data?

We are working on implementing this conforming procedure now and intend to release the new, conformed data by early summer. The conformed data will be the MRI Starch data; in other words, all our ad readership scores will be balanced to each title’s reader profiles. There is no additional charge to MRI Starch clients.

Questions? Comments? Please email Michal.Galin@mediamark.com.

Creative Client NNN Custom MRI Codes Support Strength of Newspapers in Media Plans

Newspaper National Network
Consumers turn to local newspapers for everything from breaking news and community events to food, sports and shopping information. To help national advertisers reach these readers, the Newspaper National Network LP (NNN) created custom MRI codes that show the strength of newspapers in fully-integrated national advertising campaigns. The codes quantify the value of local papers and make it easy for agencies to identify the best combinations of newspapers to deliver specific targets.

NNN, the sales and marketing arm for nearly all U.S. newspapers, advocates for in-market papers by promoting the efficiency of the medium in national advertiser media plans. They support advertisers’ marketing strategies through multi-market newspaper buys, and MRI data are key to the success of their efforts.

Initially developed in 2003, NNN MRI codes are comprised of virtually all newspapers nationwide within the top 10, 25 or 100 markets. Since each code only includes qualified respondents (reported reading within last publication frequency), it is easy to plan and buy local papers that deliver targeted audiences across multiple markets. The codes help advertiser and agency clients identify local media which best perform on all MRI measures: demographic, psychographic and product usage.

“The NNN MRI codes enable advertisers to better evaluate newspaper buys on an apples-to-apples basis with other national media,” said Jason E. Klein President and CEO, Newspaper National Network. 

MRI subscribers can access these codes to compare the demographics and purchasing behavior of a multi-market newspaper buy in a competitive media schedule. The metrics facilitate cross-media analysis and position newspapers to get their fair share of national advertising. For example, the following chart shows how NNN MRI codes over-index other local media for reaching consumers who own money market accounts:

NNN MRI Codes Show How Local Papers Deliver Money Market Account Owners
NNN MRI Codes Show How Local Papers Deliver Money Market Account Owners
Source: MRI Fall 2008; NNN Codes vs. Heavy 1 Media Quintiles

“Marketers want to use the best intelligence available to develop smart media strategies,” said Klein. “The NNN MRI codes show where newspaper readership ‘pops’; they are simple to use and help advertisers and agencies make informed decisions.”

Because You Asked Q: What Makes a Respondent a “Principal Shopper?”

Because You Asked
MRI interviews one person per household for its Survey of the American Consumer. However, the household respondent and the “principal shopper” (the primary person who shops for the household) are not always the same person.

Helpful hints for running product data:

  • Use a base of Principal Shoppers when cross-tabulating household product usage and media. This will include only those respondents who have told MRI about both their media habits and household product usage.

  • When looking for total product consumption or household product usage, use the Household Weight. This will include all respondents and Principal Shoppers.

Data collection takes place in two steps: the Personal Interview and the self-administered Product Booklet. When a respondent is the household’s principal shopper, he/she participates in the Personal Interview and fills out the entire Product Booklet. Approximately 60% of respondents are also principal shoppers; we include only these people in the “Principal Shopper” punch.

Respondents who are not principal shoppers complete the Personal Interview and fill out all sections of the Product Booklet except for the Principal Shopper section. That section is given to the household’s principal shopper for completion.

The Principal Shopper section starts on page 81 of MRI’s Wave 61 Product Booklet. 

For more information, please contact us.

If you have a question that you would like answered in Because You Asked, please contact us.

New and Noteworthy MRI and PointLogic Fusion Optimize Consumer Profiling

New and Noteworthy
MRI has fused its database with PointLogic's Compose system–widely used for cross-media planning. By linking respondent-level data from both surveys, this hybrid database will help planners better determine the optimal media mix for their clients.

Compose combines data from consumers and media planners, along with media consumption and cost information. MRI's inclusion in Compose will improve marketers’ ability to generate deeper, more insightful consumer profiles. It also will make targeting specific audiences actionable by connecting them back to MRI media selection choices. “This fusion gives marketers insights into the full range of communication channels to build campaigns that will resonate with target consumers,” said MRI SVP Marketing Anne Marie Kelly.

For more information, please contact us.

Net//MRI Teen Ratings Data Fusion: New Insights On U.S. Teens & The Internet

New insights about teenagers’ consumer behaviors and media use are available through a data fusion of MRI’s Teenmark, used for product and media consumption data, and Nielsen Online NetView’s research on Internet use. Updates are released monthly.
 
For more information, please contact us.