Digital buyers need to prepare for this before it happens. They have an opportunity to evaluate digital video advertising through the lens of a TV buyer before it's forced on them. If there's ever a time to be proactive about something, it's now. Here's how a digital buyer can be proactive with respect to GRP measurement:
First, recognize that the only impressions that matter to a TV buyer are those that reach the target demographic. For example, if the on-target demo is men ages 18-34, any impressions that reach anyone outside this demo will be considered wasted impressions. So, evaluating a digital buy on TV standards means considering off-target impressions as waste.
Second, develop an in-depth understanding of how well the digital video impressions bought for a campaign match the campaign's on-target demo. It would be easy here to assume that, thanks to audience buying, a digital video buy would have very low levels of off-target waste. However, when Nielsen OCR is used to evaluate the on-target percent of a digital buy, it's using different data than any third-party demographic data used for audience buying today. Digital buyers will want to understand the discrepancies between the targeting they've been using and the on-target percent evaluation that's built into OCR.
For example, suppose a CPG advertiser wants to reach men ages 18-34 with a 30-second commercial spot run across a TV and digital video buy. In digital video,
they run 50% of impressions through site targeting among the likes of sports sites with a strong appeal to their demographic, and the other 50% of impressions through audience targeting using a DSP
and third-party data provider. The interesting part about this is that the advertiser’s TV buy also has a measure of wasted impressions since not all viewers fall within the desired target.
The CPG advertiser then uses OCR to compare GRPs in television and digital. And the digital buyer is also looking at the percent of on-target impressions delivered via their digital buy to identify potential wasted impressions and optimize the buy for the highest on-target percentage.
Here's where the discrepancies begin to surface. The on-target results for the site and audience targeting may be totally different from what the advertiser, publisher or data-provider expected. Nielsen may report that only 60% of site-targeted impressions were on-target, or that only 50% of audience-targeted impressions were on-target.
To truly be prepared for a conversation with the TV buyer, digital buyers need to judge all the ways they buy digital video with respect to on-target percent. They should ask:
Digital buyers now have an opportunity to partner strongly with advertisers on how to use digital video buying effectively as defined by OCR. This expertise will lead to better on-target demo results.
Digital buyers aren't the only ones affected by OCR. Publishers and third-party data providers will want to know how their data offerings look under the OCR model. Publishers will want to know how their site and site categories look under the OCR model. Data providers will want to increase their data’s accuracy, and publishers will want to know how their audiences index under the OCR model.
Similar to the proactive approach digital buyers should take with respect to GRPs, the programmatic companies that support digital buyers need to be proactive as well. Every player that touches digital video ad targeting in one way or another will want to know how their form of targeting looks under GRP measurement. GRPs may end up showing the ad industry that, even online, content is the best proxy for audience. After all, TV advertisers have bought content to reach target demos for decades. If content comes out on top, programmatic companies and their buyers will need to embrace a programmatic and automated way to transact around high quality content
In short, all the digital buyers, data providers, publishers and programmatic providers need to be proactive about GRP measurement. We should prepare now for the day when the TV buyers with the biggest budgets in the ad industry start coming to our meetings. That's the day when we'll see even bigger ad dollars transition to digital.
One thing I continue to struggle with in this equation is that I can't tell you the last time I executed a digital buy and said I want to reach M18-34. I usually look at my TV buys and ask why can't I buy TV more specifically instead of asking why can't I buy digital more broadly. CBS is starting to take steps in that direction which is great, but where are the cable providers in this equation?
The issue of data validity is a whole other area and if it takes Nielsen's OCR to bring a huge spotlight to that party then maybe it will have some positive effect. It's always pained me that with all the eyes on impression viewability and verification that there are no companies acting as third party data auditors, just take providers at their word. I've seen the results from many campaigns where we have either comScore or other audience tools in the works and hope some folks on the data side start backing up what they are selling soon because I've seen a lot of crap out there. Just look how most data identifies users gender, male or female or both if they can't determine.
Ultimately there's some good that will come from Nielsen OCR shining a light on audiences and helping to start cleanup the junk data out there, but I don't think it should be used to dumb down how digital buying has been progressing. Eventually, I'd like to see TV buying get more intelligent and meet digital somewhere in the middle.
Nice article, Chris.
Abe, I wanted to let you know that there is good news, change is absolutely coming to the traditional TV side. Many operators are starting to put in place infrastructure to provide the kind of addressability and accountability you're looking for. At BlackArrow, we are actively working with operators, data providers, and measurement companies to enable the next phase of TV buying. If you'd like to learn more, feel free to give me a ping. It's not quite the pace of digital, but some very cool things are happening.
Chris and Abraham: Your points are well made regarding targeting concerns but as a blue ribbon Media Post Digital Video panel agreed (Yes, I was a member) all GRPs are not created equal (C3 based vs Viewable impressions based vs Eyes-On based vs. etc.). So the real issue is how to harmonize or equivalize the various GRPs by channel (for any common target) in order to make more meaningful comparisons between channels.
BTW: Once that is done this same illustrious Panel then raised the issue of dramatically different reach/frequency that can be achieved by different channels delivering the "same" GRPs. So perhaps we should not get too excited about GRP comparisons that are specious anyway?
Or, we could only measure the online activity of 25,000 of all internet households in the country, extrapolate whishever numbers from that makes our media producers and distributors look better than they actually are, and then shovel that bs in front of advertisers while ringing the dinner bell.
THEN we'd be just like TV, and, apparently, that's what the advertisers want.
I'm with the other commenters. Neilsen is a joke and advertisers should be coming around to us if they want actual value for their money, not begging for a stone age system that's been broken for decades.