Privacy-driven short-tail bias leaves marketers at a loss as measurement ‘fades to black’

It’s been a hard reality to fully comprehend, but marketers get it now: a recent Digiday study found 76% of brand and agency executives say they’ll lose data from cookie loss.

Easily missed, however, is that same study shows 71% say they “don’t know what to do next” in preparing measurement strategy for a post-cookie world. That revelation – that industry leaders understand there is a crisis in media measurement – is extremely important and hasn’t yet garnered enough attention.

Industry body the Interactive Advertising Bureau warns we’re approaching a “measurement blackout,” as 66% of marketers haven’t changed their measurement strategies. But it won’t happen immediately. Instead, measurement is currently in a phase Metallica might describe as ‘Fade to Black.’

Let’s consider specifically how the status quo in measurement is doomed. Any measurement strategy that’s entirely dependent on user-level tracking at scale represents a fading light. And that’s almost every major lower-funnel metric – multi-touch attribution (MTA), brand lift, digital conversion tracking, in-store sales and store traffic lift. We also need to remember that measurement isn’t just about after-the-fact reporting; measurement also helps optimize decisions during the course of a campaign and factors heavily into how marketers select media and vendors for incremental budgeting.

So where does that lead marketers who depend on those status quo methods of measurement? Two directions: toward the absolute largest media partners that have the scale to remain measurable, or toward wasting money on inaccurate (or even nonexistent) measurement of all other media partners, due to small and unrepresentative samples.

Short-term bias

One area in which marketers are experiencing the creeping darkness, even if they don’t realize it yet, is the onset of short-tail bias.

Marketers have access to fewer clear options for inventory sources scalable enough for existing measurement methods – an after-effect of a lack of addressability. The short tail is measured, and the rest is lost – because the old mechanism for measuring comprehensively is broken and won’t be fixed.

Short-tail bias damages transparency, erodes marketers’ negotiation leverage and – most dangerously, at this moment – kills innovation when the industry needs it most.

Unrepresentative samples

As a result of identity deprecation, businesses are increasingly experiencing the pitfalls of reliance on the lower-funnel metrics status quo.

Take MTA as an example. MTA is limited in that it only works by uniting a broad range of platforms into one single-source matched data set. And with addressability challenges, match rates are dropping below 5% for some media partners. That means that a centralized matched data set is often already too small for traditional MTA. Measurement companies instead try to extrapolate out from tiny samples – which are likely biased. It’s like counting cars driving past a house and using those figures to guess how much traffic is on the expressway.

What marketers can do to fight back

Marketers who depend on ID-based metrics for measuring the lower funnel need to notice the lights flickering and recognize how measurement has changed over the past couple of years. Look at how badly match rates have fallen, how leniently vendors are extrapolating and how representative and trustworthy the available data is now versus in the past.

There has to be a better way – and marketers need to find one immediately.

Here’s how to start moving toward that goal:

1. Pool insights

Take the longer-tail elements of your media plan and, for scale, pool together insights from various partners related to specific elements. For example, look at a niche segment such as ‘power users.‘ Or assess specific creative formats, such as native or add-to-cart functionality.

While each individual partner may not bring the scale needed to effectively measure using traditional methods, you can get a more accurate and precise view of performance by tactic or segment by unifying these subgroups.

2. Shift to attention

Consider how you can measure lower-funnel metrics without depending on scaled IDs. Attention is one good proxy metric – it’s a predictive metric, meaning it can be predictive of lower-funnel metrics. Attention is an appropriate stand-in for the lower funnel and doesn’t depend on matched IDs for scale.

3. Innovative partners

Look to measurement partners that are committed to innovation. Start by seeking partners that recognize how essential new measurement methodologies are – then start testing early.

Measurement as we know it may be ‘fading to black,‘ but marketers need to keep the lights on for their brands. For those who continue to use traditional measurement tactics and frameworks, as if nothing has changed, there’s not much time left to take action before things go dark. Marketers need to keep from tying themselves to the most familiar and established methods and instead commit to exploring measurement solutions proposed by real innovators in the industry.

Jake Moskowitz​ is head of the Emodo Institute and vice-president of data strategy at Emodo.