A risk model for media

A risk model for media

I have a friend who likes to compare media plans to investment portfolios. The theory is buyers have high-risk assets like sponsorships and seemingly low-risk assets like outcomes and revenue sharing.

It’s an interesting analogy to unpack because you can start to see metrics along a spectrum of risk shared between media buyers and sellers. Here’s an example for digital brands:

On one end of the spectrum is sponsorship, where a brand receives an almost unknown and unmeasurable amount of attention. On the other is LTV, where a brand assumes no risk and just pays a portion of consumer lifetime value to the seller (note that LTV buying is currently largely hypothetical). In-between are various proxies for value with different risk profiles.

It’s useful to break the spectrum down into two parts: opportunities for attention and outcomes from attention.

Opportunities for attention are just that, the potential for a consumer to pay attention to a message. In recent years, brands have pushed for less risky metrics like viewable impressions and time in view, which increase the likelihood that consumers will pay attention to a message. But even viewable impressions or impressions on a page for a minimum amount of time suffer from shortcomings - Lumen has found that only 9% of display ads are looked at for longer than a second. These metrics can easily be gamed by disingenuous sellers who jam dozens of “viewable” ads onto the same page.

Fed up with opaque and gameable metrics, some brands have pushed further down the risk spectrum into outcomes of attention, demanding that their media dollars result in lower funnel activity or even sales. In theory, these outcome-based buys are lower risk.

But we rarely see outcome-based buying in other industries. It would be like an airline demanding a barrel of oil to fly their planes a certain distance - it sounds nice in theory, but impossible in practice. Sellers would need to price in the risk from the performance of many other variables like load, maintenance, pilot, weather, and so on. Plus a risk premium. This requires a tracking infrastructure, exposure of a business’ inner workings and potentially forfeiting control of how the business is run.

Marketing outcomes offer a similar Faustian bargain. By buying outcomes a marketer is de-risking, but at the expense of training the vendor to find new customers. At some point, that brand becomes a cog in the supply chain of the platform or publisher who is best at creating outcomes - outcomes that will probably be made available to the highest bidder.

The sellers of outcomes will constantly be pushing for more control, since they are on the hook for results, wresting consumer touch points away from brands. Sellers will want creative that drives outcomes, rather than builds brands. Over time the buyers of outcomes slowly lose incentive to make great products as they increasingly focus on margin over cost of sale. Not to mention the privacy-unfriendly tracking and attribution that must be put into place to determine the cause of outcomes.

So are media buyers really stuck between high-risk opaque metrics and outcomes that reduce short-term risk but pervert incentives?

We can find inspiration in other digital mediums. Direct response advertisers settled on cost per click (CPC) almost two decades ago for good reason, it’s a near perfect proxy for site visits, allowing them to take as little risk as possible while retaining full control over what happens after the click. Brands should look for a similarly elegant proxy for the raw asset they are after: attention.

In the rush towards outcomes, we skipped over an almost perfect proxy for attention - time-spent - which is neither an opportunity for attention nor an outcome of attention. Time-spent is the middle-ground that reduces risk for advertisers without requiring them to forfeit control of creative and other consumer touch points. It also allows publishers to assess the quality of creative, and potentially build it into pricing, as we see in CPC models.

Delivering time-spent media requires a unique combination of ad format and measurement to work: the format should capture someone’s full attention for an amount of time they control, and measurement technology, like that from Moat, which can detect when an ad is on screen, is needed. Once these criteria are met, we can use time to transact media, delivering guaranteed (very low-risk) attention.

Looking at media metrics along a risk spectrum is a unique way for advertisers to frame their buying decisions. In my opinion, brands are best served by metrics that provide low-risk access to attention while allowing them to retain control of consumer touch points - like time-spent.

Marc Guldimann, CEO, Parsec

@Guldi on Twitter

Search The Drum Jobs

Explore the best jobs in Marketing and Media industries
View all open jobs