Millward Brown has passed a major milestone here in the United States. Through the first five months of 2006, the majority of revenues – 54% - were generated from online research, rather than research conducted over the phone or in the mall. This tipping point has been a long time coming, as we conducted our first online surveys in 1997. For those of us who struggled to introduce the new methodology in those early days, it is satisfying to know that we really were paving the way for the mainstream of today.
I discovered this fact because of an article on AdAge.com by Bradley Johnson, dated July 17, and titled “Online Methods Upend Consumer Survey Business.” The article quotes estimates from the newsletter Inside Research that internet-based questionnaires will account for over 35% of U.S. spending on market research surveys this year. If that is true, then Millward Brown seems to be further ahead of this trend than I would have expected.
However, to me, the estimated growth rate for online research in general looks low. Inside Research estimates that online research revenues will grow by 14% in 2006 compared to 2005. Maybe the 25 research companies that were interviewed chose to err on the side of caution, but our online revenues to date are more than 50% higher than last year’s, and the amount of pre-testing we do online has more than doubled.
The percentage of Millward Brown research conducted online reflects the revenue breakdown of U.S. Millward Brown only, and does not include the revenues of Dynamic Logic, the Millward Brown company solely focused on online ad testing and tracking. Therefore, it is based entirely on work that would previously have been done by phone or in the mall, and reflects a massive shift in the way we conduct research. However, along with the benefits of speed and cost highlighted in the Ad Age article, this rate of expansion also brings problems. In order to meet the expanded demand our sample providers are struggling to keep up. Young males are a particularly scarce commodity, because many of our clients wish to talk to them.
So to close, let me add one word of caution about the “price war” referred to in the article. Price wars typically happen when supply exceeds demand. This may have been the case last year, but seems unlikely to be true of the coming year.
(23 votes, average: 2.87 out of 5)
Loading ...
August 11th, 2006 at 12:37 pm
Hi Nigel,
Millward Brown is ahead of most. However, Nigel you note:
“…along with the benefits of speed and cost…this rate of expansion [online] also brings problems. In order to meet the expanded demand our sample providers are struggling to keep up. Young males are a particularly scarce commodity, because many of our clients wish to talk to them.”
I would argue that young males are not scarce, but their attention is. Just as advertisers are having a tough time dealing with the fact that attention is the new scarcity (not shelf space or distribution pipelines), so are survey and sample comapanies, whose models of interception and respondent-reporting are dependent on that same scarce attention. I’m not discounting the vality of these models, but it’s an issue that will get worse over time.
You spawned further thoughts here.
August 12th, 2006 at 9:09 am
Thanks for the comment Max. I think young males have always had an attention problem, just ask any college professor and I am sure they will agree! Joking apart, the difficulty of getting a share of their time will increase, if only because games and mobile phones provide an ever increasing distraction.
August 14th, 2006 at 3:42 am
In the UK, we are seeing huge growth in the take up of online interviewing, and experiencing some of the same growing pains with demand outstripping supply.
Part of the problem with young men is that their time is massively in demand - it’s not just an attention deficit! There are probably a couple of other key groups where we have similar problems (mum with kids for instance) - particular demographics are important to marketers because of their potential as consumers and so the research follows them too, in order to see whether all the marketing is working, and to identify opportunities.
These key in-demand groups need to be over-represented in web panels in order to sustain the level of work that clients want to do with them. But maybe there’s also an opportunity here for applying generic learnings about key audiences so that as an industry we don’t over-research our most important consumers?
August 21st, 2006 at 6:46 am
It was fun working alongside Nigel in those early, rambunctious days. It may have taken the better part of a decade, but the mechanics of online research have managed to iron themselves out to become a tightly defined, consistent and predicatable process.
It’s also heartening to see so much focus on obtaining projectable and representative samples online. While the first wave of practitioners have established profitable and scalable approaches to providing convenience samples, there are some bright academics who are working on ways to make these processes as accurate as telephone samples.
Finally, Max’s musings on generating survey response in an attention-based economy is a rally cry for us all to think about whether we as survey researchers are doing all that we can to leverage the full interactive power of the web.
We’ve done a marvelous job of replicating paper-based surveys online, but the reality is that the places where people spend their time online are laced with AJAX-based interactivity that create true interaction.
We are boring our respondents to death and need a new round of experimentation in order to make ourselves worthy of the attention that we seek.
August 21st, 2006 at 8:48 am
Trust Dan to know what AJAX-based interactivity stands for. If like me you are not up on the latest terms check out http://en.wikipedia.org/wiki/Ajax_%28programming%29. I agree that a more engaging survey experience would be good. So far our experimentation has suggested that there are two key barriers to overcome when building in more creative question approaches: a) respondent understanding of a new task, and, b) time to complete the survey.
September 5th, 2006 at 7:53 am
Just read your comments (Lisa Parente put me on to your blog) regarding the the accuracy of U.S. MR online growth estimated by INSIDE RESEARCH as low, suggesting the research companies that contribute to this estimate choose to err on the side of caution. Anything is possible of course, but we’ve compared the forecast in each of the last four years to the year end finals and found them to be within a few percentage points of each other. See page 3 of the January 2006 IR report.
Glad to know your doing so well with online research. As you know, Millward Brown as well as the rest of the companies in WPP’s Kantar group have refused to contribute to our estimates even though Sarbanes-Oxley can be circumvented or ignored legaly in cases such as this (see CASRO’s statments regarding this). That’s why I am surprised at your revealing that 54% of your revenue was from online. Perhaps you might argue our case with the Kantar powers that be. INSIDE RESEARCH is anxious to make the estimates as accurate as possible, which benefits everyone.
Unfortunately, the “price war” continues. Your Lightspeed unit can attest to that. The major contributor, aside from intense competition among sample providers and clients wanting it cheaper, is the entry of non-researchers, especially database marketers, who simply offer their e-mail lists as sample sources. Also, anyone can easily compile a list off the internet at practically no cost and offer it to the unwary for next to nothing. Not a good thing. We sorely need standards and rules here.
September 5th, 2006 at 11:42 am
Hi Larry,
Thanks for the comment.
The powers that be may not be happy at my revealing specific proportions but it seemed innocuous enough at the time. I am sure absolute numbers are another matter. Guess I better check whether SOX is applicable here.
I agree 100% with your comment the use of e-mail lists. We do need standards on what is acceptable.