No doubt, it's been a big couple of weeks for audience measurement.
First off, Adobe buys Omniture, and not but a few days later, comScore and Omniture announce a partnership to launch a unified audience measurement system. If you missed it - basically they are combining comScore's method of analyzing data by measuring a panel of users, and Omniture's method of measuring data collected by the browser.
But why is this a big deal?
Certainly, most of the coverage and analysis has been on Adobe writing a big check to expand their business footprint. Heck, I even took a shot at that. But, after thinking about it - if the new partnership gains the traction it promises, it may turn out to to have a big short-term, and even bigger long-term, affect on the entire way internet media is sold, purchased and measured.
But why....
See the challenge has been, of course, that there are inconsistencies in both methodologies. Omniture utilizes a device-centric (or more accurately page-side) technology to generate traffic data. So, issues like one person accessing a Web site from a different device (e.g. iPhone vs. computer vs. work vs. home) or people deleting cookies, or just simple programming errors can result in an overcount of audience. And, purely panel based research is ripe for errors too - because the sample is typically way too small and error-correction separation of work vs. home tends to undercount actual usage.
Or so people much smarter than me have debated and proposed...
But here's the real question: Who cares?
Since when has advertising driven media ever cared about how accurate their numbers *really* are?
The entire television industry has operated for decades, and aligned billions of dollars, against numbers generated by Nielsen Media Research. Nielsen has measurement diaries from a panel, as well as the equivalent of Omniture's "page side" technology in the way of television meters. So far so good right?
Well not so much. This "monopoly" is the defacto standard for television audience measurement for 307 million people, and it's set by a panel of 25,000. That's .008%. Or, look at it this way. It would be as if only 270 people from each state were the only ones who could vote for President. I know I know...bad jokes aside...
Okay - how about this.... Let's say you spend a million dollars per year in your marketing budget... It would be as if you determined your entire annual budget from the information you derived from $80 worth of keyword buys.
So, while this almost sorta kinda sometimes worked when there were 3 big broadcast networks. In a world of hundreds of cable channels it barely operates - and in a world of global internet and cross-channel media, it borders on ridiculous. Having worked at a niche cable television network back in the 90's we in the analytics group would say "geez, I hope Alfred (our made up name for the one guy who we envisioned giving us our sometimes singular rating point) doesn't get sick this week".
And, just to be clear, this isn't just relegated to ol' Nielsen or television either. Arbitron has had it's own "monopoly" on radio for years, and there's always been a question of "accuracy". In fact, just over the last few months, minority advocacy groups have repeatedly complained about Arbitron's new proposed way of measuring audiences. It finally even got the FCC's attention.
See, here's the thing... Media companies hate having one audience measurement solution because they feel like their getting undercounted. Agencies and advertisers hate the defacto audience measurement company because they feel like the measurement isn't accurate. But they both greatly benefit from having one consistent source. One solution is always defacto standard for measurement of the market - and it's just much easier to operate from a consistent baseline. Any of you who have ever switched from an installed, log file based Web analytics solution to a SaaS based "page-level" provider know this all too well.
Now, lately (in television) there's been alot of noise about CIMM (The Coalition for Innovative Media Measurement). This group was formed in September by a loose coalition of broadcasters, media buyers and advertisers looking for a new way to measure TV audiences. They've said they don't want to be a replacement for Nielsen, but rather just “fund the development of a new measurement system”.
But if you've been around the block for a while, you might think this sounds familiar. Anybody remember SMART? That was the solution proposed in the mid-90's. Basically, ABC, CBS and NBC - along with a few big advertisers and agencies commissioned SRI (Statistical Research Inc) to start up a new way to measure audiences on TV. And, they gave them $40 million to get it done.
Well that system was closed in 1999. It never really even got off the ground. As Gale Metzger, SRI President said at the time " no one said 'no', but they didn't say 'yes'". Meaning that basically all the networks, advertisers and agencies said basically "yeah, go for it... we'll do it if they do it..." And then no one did.
Or how about even further back to 1990 when AGB, a British company was invited in by the networks to introduce a new ratings system to challenge Nielsen. Where are they now? Merged with Nielsen Media to provide ratings in Europe and other countries. Funny how things work out.
So, what's the point?
Not to be cynical about it, but it's not about accuracy at the end of the day. It is simply who is the consistent and authoritative source of the numbers. The question that needs to be settled for the Media Buyer and the Publisher is: Who's set of numbers can we agree will serve as the authoritative source of where we will base our marketing decisions?
And this is where I think the Omnniture/comScore deal is a huge step in that direction. I'm already hearing from people who have seen the presentations. They say it's impressive. The bright shiny object that is "analytics accuracy" and "trust in my measurement" will go far. If (and it's a big 'if' mind you) comScore/Omniture can convince both sides of the industry that theirs is the best Rube Goldberg machine for creating the standard way to measure Internet Audiences. They win. It's that simple.
So, this puts others like Nielsen, WebTrends and Coremetrics in an interesting place (it's been a very interesting couple of weeks for those two).
- Are they going to get together - and try to assemble their own "alliance" and try to do their own thing?
- Will any of them step up and publish some "standards" that the world will have to address...
And I think that this too means that the industry as a whole is going to have to start settling on standards and once and for all settle on who is going to drive those standards. Who is going to own the standard for Web audience measurement? Or, maybe more expansively, audience measurement for media as a whole. Is it going to be the IAB - and their metrics group or the Web Analytics Association (focusing on Web metrics)... As Adobe starts to incorporate Omniture measurement technologies into other platforms (PDF, Flash etc..) and as the solutions for a singular methodology for measuring Internet audiences starts to emerge, the speed of innovation in this space is only going to accelerate.
One thing is for sure...It's going to be a bumpy ride - and it's going to be very very interesting!!
Photo Credit: Dav
Comments