It’s no secret that video viewership has been exploding online. Or that advertisers are continuing to increase investments in online video advertising as they attempt to follow the eyeballs. Strangely though, it’s also no secret that the space doesn’t seem to be getting its fair share of advertiser dollars.
Is it because of a perception that the Web has yet to prove itself as a legitimate branding medium? Because a certain magical “scale” number isn’t there yet? Could it be that the space is too fragmented? That it’s inefficient to buy when compared to TV? The continued hesitation and sometimes outright avoidance of user-generated content? Lack of compelling video creative that’s specifically suited for the Web?
Clearly, there are a multitude of possible reasons as to why video ad spend isn’t where we, as an industry, had hoped it would be by this time in its evolution. Every one of the questions above plays some part in this growth curve. But the granddaddy of them all is the one I usually avoid: a lack of a universal standard metric that can be used to measure and compare video ad buys across different properties and media. Yes, folks, I feel the need to go there. It’s time to visit the Gross Rating Point (GRP) debate.
This issue has been a thorn in the side of the entire industry pretty much since the first video ad appeared online. It rears its head at nearly every conference that contains any discussion of online video, and is perhaps the second-most-tired dispute still happening in our space.
The core of the issue is this: the TV buyers have 50+ years of econometric modeling history that tells them if they buy X amount of GRPs or TRPs (Target Rating Points), it will generate Y in return. Everyone acknowledges that there are major flaws with this methodology, but are, for the most part, resigned to it; accepting it as the best we’ve got.
As video expands to other platforms, including online, digital out-of-home, and mobile, there’s a natural desire to take that same metric and apply. But doing so fails to account for the unique attributes of these new digital delivery channels — things like interactivity, ratio of ad clutter to content, dynamic ad serving, and so forth. Are all screens created equal? Is it fair to suggest that an impression on TV (which hits a family of four who are all multitasking and talking amongst themselves) is of the same value as an impression that hits a single person watching alone on an iPhone while sitting on a plane? The oft-discussed, but ill-defined concept of “engagement” is inextricably connected to the GRP debate.
For me, the whole thing breaks down here: GRP is a reach metric. Researchers have created models that allow it to function as a proxy for impact — again, if I buy X GRPs, I’ll get Y in return. But it’s just an estimate built on small panels and projected out to the broader population. When it comes to digital, we all know that accountability is one of its most important and powerful differentiators. If you’re selling online, I can tell you how many sales your ads are driving (granted, online tracking has its own flaws, such as cookie death, but it’s getting us further down the purchase funnel). I can count — not from a panel, but from actual server logs — how many people actually viewed and interacted with your spot. I can tell you how many went from that spot to your site and registered to receive your special offers via SMS or e-mail. In other words, digital enables tracking not just of reach, but of what should be a more important business metric: impact. So, shouldn’t we be rallying around a new standard built on actual impact? As TV continues it’s own evolution, it’ll look more like online in terms of census data versus panel data, and actual engagement and exposure versus proxies, so the GRP may ultimately be threatened there as well.
And that brings me to my biggest problem with this whole debate: faced with technology that enables tracking of actual impact, why in the world would we even consider adopting a flawed, panel-based reach metric?
I spoke informally with a lot of industry experts on this, and everyone, even the TV buyers, universally agreed with that point — why revert to reach when impact is available?
But a few of them also helped get me to this conclusion: we may well need the GRP online for a short while. The problem is that while everyone acknowledges that we can and should be looking at impact, no one has universally solved what this new metric should look like. It may be easy for e-tailers who can track directly through to sales, but what about CPG companies and others who have brick-and-mortar sales channels? We’re back to modeling and proxies, for those guys. I believe that ultimately there will be a better proxy than GRP, but it’s going to take two things to get there: time and money. Where will that money come from? Maybe, just maybe, the industry will conduct some research funded by marketers and publishers alike. But the alternative is somewhat sad and ironic: adopting the GRP standard for a period of time. Given the right investment and enough time, we’ll get to something that’s more appropriate.
Jeremy is off today. This column was originally published on November 2, 2009 on ClickZ.