Having gone through several years of work in Paid Search and Affiliate Marketing, outcome-based buying always seemed part of the media landscape to me. Outcome-based buying is the ability to compensate a platform (or an individual) for its contribution in making prospective customers perform an action carrying business impact: think clicking on an ad, engaging with an interactive ad, performing a key action on a brand website or even purchasing a product/service.
Media owners control the distribution of an ad, and charge by the volume of distributed impressions, but have limited ways of generating outcomes:
- Clicking on an ad or interacting with it is the result of the combination of an action-enticing message, and its placement inside an interface. Media owners control the allocation of real estate (pixel surface) to an ad, increasing the likelihood to generate engagement. Brands control the quality of the ads.
- Performing an action on a brand website (sign up, configure a car, etc.) or purchasing a product/service depends even more on an efficient conversion funnel, which brands and not Media Owners get control.
In all cases, tying the compensation of media owners to the business outcomes they’ve generated involves two major components:
- The quality of a brand’s creative assets: think about the ad unit, think about the funnel… is there a long form that consumers need to fill in, is the buying process streamlined, etc. All these factors are outside of the media owner’s control, yet they are key in determining the success in generating cost-efficient outcomes. Consider the following:
This symbiotic relationship shows how brands who invest in great creative assets for their conversion funnels can negotiate better outcome-based rates, because they make the media owner’s job easier.
2. Measurement of outcomes is the second major component in making these solutions possible.
Most metrics offer little room for debate and negotiation between buyers and sellers: as measurement technologies got more and more accurate and standard (the IAB playing a great role in it), average discrepancies between two separate measurement solutions became neglectable. Tying an Insertion Order impression volume to a buyer ad server and not a publisher’s has become more and more standard practice. The same goes for clicks, and to the extent where companies are MRC accredited, viewability and fraud, too.
When it comes to business outcomes, purchases, online/offline conversions, the battle trenches are still there with tough problems to solve:
- When comprehensive measurement is feasible: tracking every single purchase on a website is easier than attributing them to a single publisher, when consumers are influenced by many touchpoints in the funnel. Leaving with the options to either under-count (attributing only those where the publisher is the only touchpoint) and not compensating publishers for the totality of what’s been generated or counting duplicate conversions at the expense of price-efficiency (paying for more outcomes than actually generated). Some approaches simplify the problem by paying a higher price for direct conversions vs “aided conversions”, but every attribution solution is always highly customized to each business, making it hard for publishers to aim their efforts in the most efficient direction.
- When panels are needed: tracking every transaction in-store being impossible, panels based on credit card payment data and loyalty programs can be of great help, if you accept that the behavior of a subset of the population may accurately represent the efficiency of your marketing campaigns as a whole. And then compensate the companies responsible accordingly. The issue at stake is the same as with attribution solution: when publishers do not have access to detailed and recent granular data about their efforts, they end up losing efficiency, which in return makes them less inclined to sell on an outcome basis.
The combined trends of brands in-housing their media buying efforts, and increased interest in advanced outcome-based buying, should be strong incentives for the industry to coalesce and deviate from the dilemma between overly complex attributed accuracy and imprecise panels (which bad actors can try to game for their benefit).
Collaborative identity consortiums are running hot these days and have the potential to unlock broad measurement across the digital spectrum. Their unique advantage is their inherent authority accepted by both buy and sell sides. Measurement of outcomes could then be approached like teaching how to count: in a universal fashion.