I’ve been ruminating lately on green shoots that may finally be breaking through. Not spring flowers, mind you, but original and promising new metrics that could loosen ROAS’s iron grip on marketer mindshare.
There are structural and practically metaphysical problems with return on ad spend as a metric.
For one, the whole point of ROAS is to track the ad spend itself. Duh, right? But bear with me because, the thing is, advertising isn’t the point. Take the Amazon-native ACOS metric (“advertising as a cost of sale”). A metric like ACOS is getting warmer for sure in terms of the purpose of advertising.
Because online ad budgets aren’t like a busted arcade coin machine where you put in a dollar and it spits out five quarters. Yet that’s how it’s generally treated by online platforms that have great first-party data and attribution reporting.
Retail media has catalyzed this change, because retailers are first-party data players with different incentives and their own data to back up their claims.
The Home Depot’s retail media business created a new metric last year that it dubbed ROMO (“return on marketing objectives”), which includes measuring everything from category share among Home Depot shoppers to brand awareness among recent homebuyers.
Meanwhile, ecommerce advertisers have led a revitalization of the “Marketing Efficiency Ratio” (MER) metric, mostly in pushback to ROAS.
Which leads us to the announcement on Thursday by the retail media business of the grocery chain Albertsons, which goes by “the Collective” (aside: it was formerly AMC, for Albertsons Media Collective, but the cinema chain objected), which will now begin reporting lifetime value (LTV) in its ad platform analytics.
“If we only are looking at ROAS, we’re being fairly myopic about how we understand the actual relationship between the retailer and the brand,” Liz Roche, VP of media and measurement, told me.
Rather than focus on campaign performance as a “snapshot” of attributable returns in a given lookback window, she said advertisers are asking for more “longitudinal metrics” that quantify value over time.
Eventually, the LTV metric might even be incorporated into ad targeting, Roche said. “That’s honestly the benefit of bringing these teams together,” she added, referring to her remit, which combines both media and measurement.
It takes a lifetime
One challenge in establishing a metric like LTV, however, is that you need access to large graphs of historical person-level data that can be used to model future behavior.
The Collective, for example, already creates cohorts for products it carries, with shoppers falling into “lapsed,” “new-to-brand,” “repeat” or “loyal” customer sets. With historical data going back a year or many years for some shoppers, Roche said the grocer can analyze customer retention, changes in average order value and purchase frequency and see how different marketing tactics impact those outcomes over time.
Using this view, the Collective can project the value of certain cohorts of shoppers – which is to say ad conversions – over the year to come.
“Once we can get [advertisers] into this new category,” Roche said, referring to the mindset of using LTV rather than ROAS, buyers can broaden their focus from new-to-brand attributed customers generated by a given campaign to the customer types and media engagements most likely to drive value over the course of a year.
There are drawbacks, though, which is why it will take a long time to usurp ROAS, if it ever happens.
For example, the new LTV metric is “built really for storytelling,” Roche said, meaning that it’s a way help marketers justify their investments – with the Collective, of course.
The new LTV metric isn’t a measure of user-level or conversion-based success. When a campaign shows high ROAS, it reflects things that have already happened – you can see the revenue, sales or downloads directly. High LTV campaigns, on the other hand, point to the future value those users may generate over time, rather than immediate outcomes.
Also, individual impressions and customer conversions don’t get their own LTV score. The LTV metric applies at an aggregate group level to a cohort of audience targets or previous buyers.
Which is another reason why the new metric fits into analytics but is harder to translate to media buying. And to eventually incorporate LTV into targeting will require being able to more confidently assert the value of lifetime value to the CMO and to the even more skeptical CFO.
“But the bottom line is, you have to start somewhere,” Roche said. “That’s the fun part about research analytics and starting to build this stuff.”
