Commentary

'Risk Sharing' With Audience Data: It's Complicated

Audience data for TV has quickly become the defining trend for 2015. Like many other industries in the Big Data era, TV has never had so much data available before. Spending patterns from credit card databases are now matched to TV viewership trends. So are shopper card databases from retailers and auto registrations. With all this new data, it’s possible to now know what Oreo cookie lovers watch on TV and which networks are big hits with United Airlines frequent flyers.

Advertisers and their media agencies as well as media sellers are moving quickly to use these new data sets to plan and buy TV schedules that target audiences with more precision than ever before. They’re also taking the data to the natural next step of measuring the results and business impact of the targeting. If we know that Sephora shoppers over-index on CNBC and Bloomberg, then a heavy CNBC and Bloomberg TV plan should increase the number of shoppers and sales at Sephora.

Such campaigns have been so effective that some media agencies and media sellers are entering into agreements that tie business results for TV brand campaigns to the media contract. This type of risk-sharing model should bring incremental value to both sellers and buyers. If sellers are willing to take on the added business performance risk, then buyers should allocate more of their scarce media budgets to data-driven TV campaigns. In theory, buyers should allocate up to the point that the ROI becomes break-even. Risk-sharing should also more closely align the incentives of the buyers and sellers, so that pricing negotiations are focused on a common goal.

advertisement

advertisement

While the concept is positive, these risk-sharing agreements can become very complex very quickly.  Often, the devil is in the details. Years ago, I worked for a management consulting firm that had become so adept at improving efficiencies in its clients’ businesses that it began to write contracts guaranteeing certain business results. After all, if we were willing to “put our money where our mouth was,” we were answering clients’ skepticism about the value of our consulting projects.

While these projects often started off with a great sense of unity between client and consultant, that unity too often flipped when it came time to take credit for results. Frequently, a major unforeseen change happened in the market that affected the client’s business, so it was nearly impossible to isolate and measure the benefits of a consulting project.

It’s critical to be forward-looking with such arrangements, with a clear understanding of how results will be evaluated. Factors that could affect business results outside the control of the TV brand campaign should be delineated,  such as a competitor's response or a change in spend in other media. Further, a great deal of information on business results must be shared by the advertiser, a process that could be sensitive.

It seems to me the best solution to address these complexities is to have an independent third party conduct the measurement and attribution of results. That third party must also be truly impartial and transparent about any commercial biases — so the buyer and the seller should likely finance the third party together. The good news on this front is that a number of digital measurement and attribution companies are making moves into TV, and existing TV measurement companies are investing in more sophisticated capabilities. We may soon see a new body of standards emerging for measuring data-driven TV campaigns.

Doubtless creating shared incentives between media sellers and buyers for business outcomes will bring new value to the TV advertising ecosystem.  With the abundance of data at hand, it is now possible to credibly measure business results from TV brand campaigns. But we can’t underestimate the complexity involved in getting this right -- not only in the correct logic of the measurement techniques, but also in avoiding any commercial bias from the results.

1 comment about "'Risk Sharing' With Audience Data: It's Complicated".
Check to receive email when comments are posted.
  1. Ed Papazian from Media Dynamics Inc, June 5, 2015 at 1:02 p.m.

    But Walt, we already know which TV shows target Oreo cookie users and frequent flyers. We have known this for decades from services like MRI and Simmons. This is nothing new and has been part of the media planning process for a very long time. Moreover, as I keep pointing out as one article after another heralds the great transformation of TV time buying that is supposedly taking place at a breakneck pace---it isn't.

    As i pointed out in my recent comments about another "opinion" article which made the same claims. The mad rush to "TV audience buying" is actually a slow trickle, consisting, mainly of experiments with portions of the "programmatic" methodology and very small, computer assisted buys, mostly involving some very low rated cable channels and a few non-premium shows on larger channels, where there is plenty of available GRP "inventory"---if anyone wants it.

    I don't propose to repeat all that I have said here---as I take pity on the audience. Besides, I've got to save my ammunition for the next opinion piece which says basically the same thing, and the next and the next. I'm sure we will have another four or five of them within a week.

    I happen to believe that any proposed new system should be properly and objectively investigated to see if it has merit. This certainly applies to the TV "programmatic" idea as well as the various data sources and statistical manipulations that are appendages to it. So, let's investigate----and that is what is being done. If it can be demonstrated that this is really an improvement---for buyer and seller----so be it. But that very definitely remains to be seen---and proven.

Next story loading loading..