|
|
MRI's
Position on the Pros and Cons of Fusion
Advertisers' need to plan and analyze the best mix of media has spurred an industry-wide discussion on fusion. In this interview,
MRI Executive Vice-President and Chief Research Officer, Dr. Julian Baim, offers his point-of-view and Mediamark's position on the pros and cons of fusion.
|
Q. What exactly is fusion and why is it needed?
Fusion is linking respondent level information from two or more separate surveys into one, fused database. It's become something of a hot topic of late, stemming from advertisers' desire to plan and analyze the best mix of media. Integrated media planning not only requires understanding audience levels for major media—magazine, TV, and radio—but profiling each audience's demographics, brand and product usage, and behaviors, lifestyles and attitudes as well. The ideal way to get this information is with a single, all-encompassing study. Fusion is essentially a "stand-in" for a single-source database…and theoretically allows media researchers and planners to conduct single-source analyses from multiple surveys.
|
|
"Fusion is not magic or alchemy. You have to have
quality studies
to begin with or else you can mistakenly attribute more power to fusion than it actually has."
MRI
Executive
Vice-President and
Chief Research Officer,
Dr. Julian Baim |
|
Q. But surely some studies measure multiple media?
Yes, some do. MRI, for instance, provides the "currency" for evaluating magazine audience levels, but we offer much more than that. We also deliver TV, radio and Internet usage information linked to extensive intelligence on consumers' buying and lifestyle behavior. So, while we're best known for magazine measurement, we are also a single-source multi-media study. The rationale for fusion is negated, to a large degree, by the fact that we collect media, demographic and product data all in one, very high quality study. In fact, we believe that our non-magazine media data are the best source for multi-media duplication analysis and audience product profiling. The dilemma is that unless planners are willing to work with single-source data in which audience levels typically differ from the accepted "currencies", data integration becomes the solution.
Q. What is MRI's view of fusion?
We're not at all convinced that fusion is the answer—and feel that
using single-source media, purchasing and behavior data is by far
the more reliable alternative. However, we do endorse
experimenting with fusion. Fusion offers the possibility of being
able to integrate media "currencies" while reducing both
respondent burden and questionnaire length. Those are two things
we are always trying to improve upon.
Q. What are the major caveats to fusion?
The first is the quality of the research involved. If you take one
high-quality research product and one of poor quality and fuse
them, it doesn't enhance or improve the quality of the originally
collected data. Fusion is not magic or alchemy. You have to have
quality studies to begin with or else you can mistakenly
attribute more power to fusion than it actually has. Another concern is
that, in most cases, fusion comes down to linking two media
currencies solely based on demographics--and our research has
shown that demographics alone do not effectively link variables.
Q. So what kinds of things should researchers study when
considering fusing databases?
First, their evaluation of the studies
involved should be a very thorough process and include everything
from sample design to execution of the study to response rate.
Also, Media Rating Council accreditation of the research provider
would be a good indication that minimum research standards are in
place.
Q. What else?
In the actual implementation of
fusion, a number of complexities can arise. For instance, the
sizes for the two fused samples are often not the same, which
means that some respondent data may need to be used more than
once, not an ideal situation. And, in many cases both samples are
"weighted" adding further complexity to the "reuse" issue. But a
deeper concern is that fusion, by using only demographic variables
to link studies, overlooks the importance of more critical
explanatory variables. A strong relationship between two media is
often explained by commonality of interest and not by
demographics. For instance, if you watch golf TV shows, you'll
probably also read Golf Digest, Golf Magazine or other golfing
publications--but demos themselves don't necessarily explain this
connection. If you use demos to match responses, you are going to
get a random intersection of those two media. But if, on the other
hand, you could use interest in golf as the linkage variable,
you'd retain the relationship to a better degree. You dilute the
relationship when you only use demos, which is commonly done in
fusion.
Q. What's the difference between
ascription and fusion?
They are generally based on very
similar techniques, but ascription usually applies to procedures
employed within a study and fusion takes place across separate
studies. Another difference is that, with ascription, a relatively
small portion of responses are imputed or modeled, while 100% of
responses are imputed with fusion. Consequently, there is a much
greater likelihood of error with fusion than with ascription.
Read more about fusion in
Data Fusion from the Media Sellers
Perspective, as presented at the 10th Worldwide
Readership Symposium by Jane Bailey and Susan Boller of Time Inc.
back
to top
|
|
|
Hachette Filipacchi
Media U.S. Uses MRI Online Media Panels to Build In-depth
Relationships with ReadersWhen Hachette Filipacchi Media U.S.
has questions about their readers' attitudes and behaviors, they
call upon online research for fast, accurate information. HFM U.S.
is the first major publishing company to make a corporate
commitment to online reader research to address editorial, sales,
and advertising issues. Partnering with MRI, HFM U.S. has
conducted 60 surveys so far this year querying 54,000 online
members across 17 magazine panels. HFM U.S. was able to collect
in-depth reader feedback to evaluate editorial direction and
provide their ad sales reps with highly targeted research as a
"door-opener" to agencies and advertisers. "We've only begun to
scratch the surface in finding creative ways to use panel
research," says HFM U.S. Research Director Pat Keane. "Online
panels are a cost effective way to contact subscribers, ask
customized questions and address timely issues."
Online
Research Helps Hachette Meet Evolving Industry Needs
HFM U.S. turned to online surveys to meet increased demands in the
magazine industry for fast, pinpoint-accurate and affordable
research. Traditional research response rates were dropping, while
survey costs were rising. This, combined with the demands of
increased competition for readers and advertising dollars, meant
HFM U.S. could no longer afford the up-to-six-month time frame
required to finish a traditional study. "This route gives us not
instant, but very quick turnaround and it's more cost effective
than phone or mail surveys," says Keane.Online surveys also offer unprecedented advantages to the
company's smaller publications like Premiere, Flying and Sound &
Vision. "Smaller magazines at HFM U.S. are now able to conduct
research at the same level as our big books (ELLE, Car and Driver,
Woman's Day) since online panels are so cost effective. This was
very exciting for the publishers of our smaller magazines," Keane
adds. |
|
|
What MediaPanel Can Do For You
MRI's MediaPanel
provides ongoing intelligence about a magazine audience through
Internet surveys conducted among a panel of readers.
"Decisions impacting ad sales, classified, editorial,
circulation, and the client's Website require current
information" says Paul Gold, director of Mediamark Interactive
"and MediaPanel provides it in time to make the right
choices."
MRI builds and hosts a recruitment Website, emails invitations
to participate to readers and Website visitors, and executes a
qualification questionnaire against all panelists. MRI also
designs, programs and hosts monthly -- or more – surveys.
Results are provided in real time, online tabulations so you
can respond to information quickly and effectively. For more
information, contact
Paul Gold. |
|
MediaPanel, combined with MRI's syndicated research, enabled HFM
U.S. to start broad and drill deep for feedback on each of its
seventeen magazine titles. "Research for our enthusiast magazines,
such as Boating and Cycle World, need to go deep into niche
categories. By using MRI as the base for information, and using
panels to fill in the blanks, we truly get detailed data about
readers' attitudes and behaviors for these kinds of magazines,"
explains Keane.
Reader Feedback Keeps Editorial On the
Right Page
Using individual magazine MediaPanels on an ongoing basis, HFM
U.S.'s editors maintain a continuous dialog with readers--whose
demographics, backgrounds and interests cover a wide spectrum--to
gauge reaction to editorial. Are they providing too much, too
little or just enough information on a particular topic? "We've
used panel information to help decide whether to add new sections
or eliminate regular features," says Keane. Woman's Day has added
a creative twist to MediaPanel's functionality by including a
"Reader-to-Reader" page in each issue, which shares thoughts and
opinions gathered from its reader panel.
For HFM U.S. books with significant newsstand circulation, such as
Home and Metropolitan Home, choosing the most effective cover is
vital. MediaPanel allows editors to quickly test readers'
reactions to proposed covers and cover lines, which helps drive
sales.
Developing New Business With Online
Research
HFM U.S. uses MediaPanel to leverage ad sales by conducting
research on behalf of advertisers and by equipping sales
executives with targeted reader opinions, behaviors and product
use preferences.
"As preparation for a new business call, we can conduct a survey
on a specific segment of the market and then demonstrate to the
potential advertiser how our magazine(s) attract the audience they
want to reach. For example, if we can show that 40 percent of our
readers use the advertiser's brand, they see in quantitative terms
the likelihood of reaching their target through our books. Also, a
lot of RFP's today ask for details on readers' psychographics and
attitudes, some of which we don't have at hand. Our online reader
panels help Sales and Marketing incorporate specific information
advertisers' request into their presentations," says Keane.
| MediaPanel offers advertisers the
opportunity to test creative, to measure ad recall and to evaluate
brand awareness as part of their sales packages. "We have
conducted pre-and post-studies for advertisers that have been
extremely successful says Keane, "and have also offered them as
merchandising."
HFM U.S. advertisers can also use MediaPanel for product testing.
Keane comments, "Two of our magazines are working with an
advertiser to test a new product and we've solicited 50 panelists.
The panelists then tried the product and answered like/dislike
questions." Keane adds that this research was completed in less
than a month, which would not be possible using other
methodologies. |
|
"We've only begun to scratch the surface in finding creative
ways to use panel research. Online panels are a cost effective way to
contact subscribers, ask customized questions and address timely
issues."
Pat Keane
HFM U.S. Research Director
|
|
Multi-panel research can also be used to compare readers'
responses across various combinations of magazine titles. HFM U.S.
conducted an automotive multi-panel study to demonstrate the
influence of magazines in the new car buying process. "We wanted
to show that people in the market for a car use different sources
of information at various points in the buying process," says
Keane. "We also wanted to demonstrate women's influence and prove
their car purchase process is similar to men."
Calling on
the Experts
HFM U.S. chose Mediamark to conduct their online studies because
MRI is a research brand that people know and trust. "The name
brings credibility to online research. It's a new technology
and with MRI there's an implied endorsement that you can do
good, quality research online. That makes the results more
appealing to clients." Also, MediaPanel clients can index the
findings to MRI's national Survey of the American Consumer, "a big plus," says Keane. back
to top
|
|
|
Omnibus Re-contact Study
Adds Value to Syndicated Research
Now you can increase the value of Mediamark's research by adding
proprietary data to MRI's National Survey of the American Consumer.
The 2002 Omnibus Re-contact Study will be mailed in January
2003 to MRI's Fall 2002 base of 26,000 respondents. The
survey will take approximately sixteen weeks to complete.
"MRI is conducting an annual Omnibus based upon the success
of the initial 2001 Re-contact Survey," says EVP Research
Julian Baim. Last year's inaugural Omnibus was a significant
achievement – nearly 10,000 consumers responded to the
mailed survey, a response rate of over 40 percent.
The MRI Advantage
The key advantage to using MRI's Omnibus Re-contact Study rather
than a separate custom study is the ability to customize the
robust Mediamark database. By re-contacting the 26,000 respondents
in the Fall Survey, participants in the Omnibus can match their
own unique sets of questions against the detailed brand and media
usage information already included in the national MRI Survey.
And, because MRI data are derived through in-home, in-person
interviews—an unequalled methodology for collecting reliable
information—participants can be assured of superior, actionable
results.
Participating in a multi-client survey has unique advantages as
well. Not only is MRI's Omnibus Re-contact Study sent to the
largest random sample base in the industry, but by conducting a
survey participants share, MRI re-contacts that base only
once--maximizing potential response. "Because it is a shared
survey, the Omnibus minimizes respondent burden," notes SVP of
Custom Research, Karen Tubridy. "This also has the added benefit
of reducing costs to the client, since the survey expenses are
shared by multiple participants."
|
|
|
Arnold MPG Uses
Omnibus for Segmentation Analyses
Arnold MPG used MRI's
2001 Re-Contact Study to gather information for several clients.
Senior Vice President Director of Research & Modeling Tonya Deniz
says, "One of the greatest benefits was the ability to bridge the
gap between clients' proprietary research and syndicated research.
"The Omnibus enabled us to ask respondents the same questions
asked by clients in their own research. This provided us with a
direct link to the MRI Study. Previously created client clusters,
developed from factor analyses using client data, now could be
replicated using disaggregated MRI Omnibus results. Once
aggregated, the clusters can be crossed with the vast array of
marketing and media information that is provided within the MRI
Study. This led to valuable insights that enabled us to make
better-informed strategic and tactical decisions on behalf of our
clients."
Deniz says that their clients were thrilled with the high response
rate which provided enough data to develop segments. "The segments
have been integrated into media planning for Arnold MPG and Media
Planning, both of which are divisions of Arnold MPG. We look
forward to the next MRI Re-Contact Study so that we can track the
behaviors of these clusters over time."
|
|
Omnibus:
Deeper, Personalized Data
Marketers can use the Re-contact Study to extend the value of MRI
data to address a range of research needs.
- Augment the
category, brand and lifestyle data in MRI's database with
tailored psychographic questions
- Create custom
segments that can be incorporated into media plans
- Link existing
attitude & usage surveys to MRI's syndicated data
- Develop direct
marketing strategies by linking customer databases to MRI for
CRM applications
The closing date
for participation is October 1, 2002. If you would like to take
part in this year's study or would like more information, please contact
your account representative.
back
to top
|
|
|
Ready,
Set…Sweat: Americans
Increase Participation in Sports
More American adults are lacing up their athletic shoes and
burning calories today than 12 years ago. Respondents who
participated in at least one of the 37 sports measured
consistently by MRI increased from 62 percent in 1990 to 69
percent in 2002. And, as baby boomers are getting on in years, the
sporting population is getting grayer and the types of athletic
activities gaining popularity have shifted to reflect
America's need to stay in shape.
|
Older
Americans Boast the Greatest Increase for Participating in
Sports in 2002 Versus 1990 |
|
Age |
Spring
1990 |
Spring
2002 |
% Change |
|
18-24 |
70% |
72% |
+3% |
|
25-34 |
72% |
70% |
-2% |
|
35-44 |
68% |
72% |
+5% |
|
45-54 |
63% |
71% |
+12% |
|
55-64 |
50% |
63% |
+26% |
|
65+
|
40% |
50% |
+28% |
|
Source:
MRI Spring 1990, Spring 2002
"Percent of Age Group that Participated in Any Sport (in
last 12 months)" |
|
|
Even though most age groups increased their participation in
sports, older adults are collectively breaking more of a sweat by
boasting the greatest percentage increases for participating in
sports in 2002 compared to 1990. Conditioning
Replaces Team Sports
When the most popular sporting activities -- those in which
greater than 2.5 percent of the population participated -- are
grouped together, those whose prime benefit is improving fitness
have shown the greatest increase. Team sports decreased 18
percent, while participation in conditioning sports increased 10
percent.
|
Perhaps following
doctors' orders, Adults 45+ are leading all other demos for the
greatest increase in participation in conditioning sports; 45-64
year-olds who participated in conditioning activities 11 or more
times in the last 30 days climbed from 19% in 1990 to 34 % in 2002
while the percentage of 18-34 year-olds plummeted from 49% to 33%.
In fact, across all four mainstream categories, younger adults
show an overall decrease in participation, while 45-64 years old
generally show increases. |
|
|
Conditioning on the Rise |
|
Type
Sport |
Spring 1990 |
Spring 2002 |
% Change |
|
Team |
24% |
19% |
-18% |
|
Nature |
23% |
22% |
-5% |
|
Society |
15% |
14% |
-7% |
|
Conditioning |
41% |
45% |
+10% |
|
Source: MRI Spring 1990, Spring
2002; Participated in Any Sport in Last 30 Days |
|
Sample of Sports in Groups: |
|
Team Sports: Baseball, Football,
Volleyball
Nature Sports: Hunting, Fishing,
Hiking
Society Sports: Golf, Tennis, Downhill
Skiing
Conditioning Sports: Running Swimming,
Weightlifting, Walking, Bicycling |
|
|
|
Despite their decreased participation in mainstream categories,
trendsetting younger adults lead older demos in activities such as
kickboxing, skateboarding and frisbee. Adults aged 18-24 are 40
percent more likely than those in other age groups to have
participated in these less traditional sports. Younger
participants also seek greater variety in their activities; on
average, 18-24 year-olds participated in five different sports on
a per capita basis, compared to an adult per capita of four. |
|
|
Conditioning Gains Popularity
Among Older
Enthusiasts
(Percent change between 1990 and 2000) |
|
Sport
Type |
Age
18-24 |
Age
25-34 |
Age
35-44 |
Age
45-54 |
Age
55-64 |
Age
65+ |
|
Conditioning
Sports 11+ |
-4% |
-13% |
0.25% |
11% |
4% |
2% |
|
Team
Sports 6+ |
-1% |
-9% |
7% |
6% |
-2% |
-.35% |
|
Nature
Sports 5+ |
-3% |
-13% |
7% |
6% |
3% |
-.17% |
|
Society
Sports 5+ |
-6% |
-8% |
1% |
6% |
2% |
5% |
|
Source: MRI Spring 1990 & 2002, Heavy
Participation, Last 30 Days
|
|
 |
Marking
a prime target
As sports participation has increased among the 45-64 year-olds,
so has spending on sports-related equipment. Since spring 1990,
the share of spending on big-ticket sports and recreation
equipment among this group has nearly tripled, from 11 percent in
1990 to 30 percent in Spring 2002. While 18-34 year-olds still
command the greatest share of total spending, their share has
decreased during the same period. |
back to top
|