When I look back at my 2009 predictions, I feel pretty good about my hit rate; you can
judge for
yourself. Of course predictions are always a tricky business, especially ones about the future. So herewith, three trends I think will shape the online metrics business over the next 12
months. Let's call them trends, not predictions.
Web analytics and audience measurement will continue to enrich and embrace one another.When I first joined comScore and began contributing to this space in 2007, I talked about the Internet as the most measurable medium-- meaning, for better or worse, that we would be the
medium with the most measures. In particular, when it came to counting up the audience, we had two very different techniques: person-centric panel measurement, and site-centric server
data. At the time, many suggested that server-based numbers had to be correct - after all, wasn't the Web site's servers the true census of how many people were visiting a particular Web
site? As it turned out, there were many issues -- cookie deletion being the most prominent -- that caused unique visitor counts from server data to grossly overstate the true number.
We've travelled a long way in the almost-three years since then. Panel/server hybrid measurement exists in the market, and it is helping to resolve what once seemed to be irreconcilable
differences between two ways of counting. Removal of uncertainty in measurement can only be good for business.
But this convergence doesn't mean there isn't still room for two
distinct metrics disciplines. Syndicated audience measurement supports competitive analysis and the planning, buying and selling of advertising; whereas the bread and butter of web analytics is
to provide publishers with KPIs for site optimization. The function and purpose of these two data streams remains separate, and both are essential.
I can tell you firsthand that we at
comScore have learned that panel data can make census server data better, and that server data can make panel data better. Moving forward, I expect to see the purveyors of the two respective
online metrics disciplines continuing to benefit from one another. For example, the IAB Audience
Reach Measurement Guidelines, written in conjunction with the MRC, call for syndicated audience measurement organizations to make cookie deletion data available to subscribers for adjustment of
their own internal data. Continued cross-pollination between audience measurement and Web analytics means everyone wins.
New data on the branding value of online display,
especially when campaigns run with premium content publishers. We're seeing a confluence of trends in the digital advertising space that are, collectively, paradoxical. We know
that clickthroughs on display ads have dropped to minimal levels, but we have also learned that a preponderance of the immediate measurable effects of online advertising are latent (occurring
both in subsequent online session, and offline) - both arguing for a shift from direct response to brand-building as a core benefit of online display advertising. We know too that advertisers
are increasingly demanding engagement, a quality that is difficult to define, but which we can all agree has something to do with the quality of the experience the consumer has with the impression.
At the same time, ad distribution strategies and technologies such as ad and data exchanges can get an advertiser's impression in front of a target group of cookies (not, it should be
stressed, persons) with minimal lead time, largely by skipping over the media vehicle that originally aggregated and attracted the audience, instead speaking directly to those cookies through a
variety of (mostly) unknown Web sites over which the advertiser has little control.. By circumventing the original media vehicle, these delivery technologies promise to squeeze much of the cost (and
subsequently profitability) out of online advertising models.
Clearly there is a place for this type of message dissemination; when the goal is quick, low-cost tonnage of impressions, the
model works.
But historically, "quick and cheap" doesn't build brands, and don't deliver engagement.
In 2010, demonstrating the branding and engagement value of
quality, differentiated content is going to be one of the most important challenges facing premium publishers -- by which I mean, the media companies we all know and live with across platforms and
screens every day. And, the long tail notwithstanding, these publishers collectively account for the majority of time spent and page views consumed online. Ultimately, I suspect that some
publishers will align around a model wherein they retain their own high-value inventory, for sale to advertisers who are looking to build brand equity and establish and maintain relationships and
conversations with customers over the long haul; while making the remainder of inventory available for more commoditized sales. Others will adamantly refuse to sell their ad inventory through the
exchanges and networks.
So in 2010, look for an increase in demand for services, studies, and analyses that quantify the effectiveness, ROI, engagement, and branding value of
prominent placement in premium, differentiated content.
By the way, it you ever doubted the importance of context, consider this fascinating experiment conducted by the Washington Post.
Cross-platform
measurement tools and insights
Media are becoming digital. Mobile usage is increasing dramatically -- although the mobile ad market is still in a nascent stage -- and
set-top-box data is turning TV into a digital medium -- with the census data and addressability we have long been accustomed to in the Internet space. As a result, more and more media
measurement will begin to resemble Internet measurement -- with census data and panel data, direct response data and branding/awareness data.
At the same time, buyers and sellers
increasingly want to understand the holistic nature of a campaign. How do TV and the Internet work together? Print, mobile, and the Web? What does online video advertising add to the
reach of an offline TV campaign?
This may well be the year that cross-platform measurement goes digital, and goes live. We know that CIMM has called for a (TV-centric) cross-platform
solution. We know that browser-based mobile traffic to the traditional Web is increasing, and thus almost by default site-centric systems are capturing data from two platforms (computer-
originated and mobile-originated Internet usage.)
In 2010, expect to see increased activity in cross-platform measurement -- moving beyond fusion to incorporate new data collection
and integration strategies.
What these developments mean for all of us is that this year -- like last year, and the year before that -- the online metrics space will continue to evolve at a
rapid pace. And it won't be dull.