Programmatic buying needs surrogate data

Programmatic buying needs surrogate data

Giles Cottle, Wed 27 May 2015

Programmatic buying, or real-time bidding (RTB) is right up there with big data on the hype scale.

For the uninitiated, it enables advertisers to judge what to bid for a piece of inventory at the final moment according to their own business rules.

Media owners, meanwhile, can use it to extract greater value from their inventory, and reduce the amount of unsold inventory that they have. It has been long established in the online world – YouTube, for example, has enabled the bulk of its video inventory for programmatic buying in 2011.

Programmatic buying has great promise by delivering efficiencies on both the buy and sell side, and potentially making ads far more effective. But it’s worth us breaking down what RTB is, and isn’t, and does, and doesn’t, necessarily do. It definitely oils the machinery of ad selling, which admittedly yields cost savings and revenue gains for both buyers and sellers.

Advertising in Time Square

What it doesn’t do – not necessarily – is automatically improve the response to an ad or make it resonate better among its target audience. There’s a surprising – and inaccurate – assumption that these efficiencies automatically carry through to impact, in other words that the price paid reflects the true value of the ad. RTB has become associated almost with a “measurement dividend”, with the idea that it is no longer necessary to invest as much assessing an ad’s effectiveness after it has played.

Several studies have debased this assumption – most notably one from University College London, which found from analysis of conversion rates that the current bidding strategies are far from optimal. The study argued that there were “significant needs for optimisation algorithms incorporating the facts such as the temporal behaviours, the frequency and relevancy of the ad displays, which have not been well considered in the past.”

This study focused on the online market, but this applies just as much to TV now that RTB is gaining ground rapidly for trading ad inventory there. Some major brands have announced plans to divert up to 75% of their digital ad budgets to programmatic buying over the next year. It is clear that RTB will require accurate measurement more than ever to drive analysis in near real time. Armed with accurate information the RTB algorithms can take account of response and feedback in near real time and use this to modify bidding strategy as campaigns unfold.

This does beg the question of how to measure impact in near real time given that the full effect of many ads can only be assessed later after they have rippled across multiple channels and provoked delayed response among their target audience. This will inevitably require surrogate measures of impact that are predictive of the final outcome, based perhaps on engagement and viewing time. There is certainly no shortage of data that can be collected, as all media becomes increasingly measurable at ever deeper levels of granularity. We can now measure not just every media channel and campaign, but drill down across every outlet down to impressions and individual consumers, with increased capability to tune in to how ads are being consumed on the screen.

The critical factor though is the actions that we can enable by analysing the data, which must be responsive to immediate feedback and sufficiently granular to identify specific aspects of a campaign that need adjusting during its course. Waiting to observe a spike in sales, and retrospectively attribute it to TV advertising, may not cut the mustard in the future. On the one hand we need continuous reporting through easily digestible scores and measurements that can be used to make manual adjustments to a campaign over time. On the other hand we also need immediate feedback into the RTB algorithms so that they can bid on the basis of up to date information rather than forward looking assumptions.

Given reliable measurement coupled with rapid and relevant analysis, programmatic buying has the potential to deliver not just trading efficiencies but real impact – as long as we have all the data we need to prove it.

Need help? Get in touch...

Sign up below and one of our data consultants will get right back to you

Other articles about Data Strategy

Dativa is a global consulting firm providing data consulting and engineering services to companies that want to build and implement strategies to put data to work. We work with primary data generators, businesses harvesting their own internal data, data-centric service providers, data brokers, agencies, media buyers and media sellers.

145 Marina Boulevard
San Rafael

Registered in Delaware

Thames Tower
Station Road

Registered in England & Wales, number 10202531