Attribution. Oh, attribution. Even typing the word makes me squirm. And I’m not alone — 40% of CMOs say improving ROI and proving attribution across the marketing mix is a top priority, according to the CMO Council’s “CMO Intentions 2024” report. Maybe the anxiety isn’t so dramatic after all.
Yet, while attribution is on everyone’s mind, there are many misconceptions around it — most notably, how it can be used to improve ROI, which, as the study points out, is why marketing executives care about attribution in the first place.
The fundamental problem with attribution
There is a key problem with traditional attribution thinking. It only tells you if something happened, not if it mattered.
What do I mean by this? Take branded paid search. Someone looks up your brand, sees your paid search ad, clicks and makes a purchase. Did bidding on that branded term cause the conversion? Maybe. Maybe not. Most marketers approach branded search with a healthy dose of skepticism. However, that critical lens must be applied across every channel and tactic.
Why? Because most performance marketing campaigns are conversion-optimized. They’re designed to drive a specific action — sales, leads, key page visits. When the platform sees it delivered that action, it optimizes for more people like the ones who converted. And the cycle continues.
Sounds good in theory, but what does “generate” mean to an ad platform? It means “gets attributed” — a tracked signal that gives the platform credit. And who’s most likely to convert? Back to our branded search example: The people who already know your brand and are most likely to buy, with or without the ad.
Dig deeper: The smarter approach to marketing measurement
The attribution paradox
Let’s zoom out. Ad platforms optimize for — and most marketing teams report on and budget based on — attributed outcomes. The hope is that all of this will lead to greater ROI. But what if this approach has the exact opposite effect?
This is precisely the case for some tactics, including those that command significant budgets for brands. Consider the following:
- Branded search.
- Retargeting.
- Performance Max campaigns in Google.
- Advantage+ campaigns in Meta.
Despite showing great attributed metrics, these tactics have the highest risk of wasting money.
The usual suspects: Tactics that mislead through attribution
Branded search
I’ve mentioned this frequently already because I trust most brands have had conversations about the incrementality of branded search (even if they haven’t used the word “incrementality”), so it’s a great place to illustrate the fallacy of attribution.
For many brands (except in highly competitive industries), most consumers will simply click the first link in the SERP for the brand they’re searching for. Due to Google’s efforts to make ads harder to distinguish, most consumers don’t know which result is an ad. Most consumers will simply click the organic listing without paid brand search ads.
While paid brand search is running, it will get a lot of attributed credit. But when removed, organic search increases by the amount paid search used to generate. Any gap would represent the volume incremental to paid brand search, and now we’re halfway to a great starting point for incrementality measurement (but that’s an article for another time).
Retargeting
What’s the best way to juice your numbers? Run 12 retargeting vendors! Suppose you only report on attributed numbers from ad platforms, each with a siloed view of the world. You’ll see a phenomenon that converts many attribution purists to seek a better solution.
For pure-play DTC brands or the online portion of midsized omnichannel retailers, you’ll often see the total number of attributed conversions surpass the total number of actual conversions! If you don’t immediately change course, this is a great way to lose credibility with your CFO.
Unfortunately, the typical response is, “Oh, we just need to deduplicate — let’s use Google Analytics data instead.” This approach is equally dangerous because you’ll rely on click-based data alone, opening another can of worms and reinforcing the “only valuing click-based activity” mentality.
Teams start running paid social campaigns with one-day click attribution and never invest in channels that generate demand, only those that capture demand. The problem? Eventually, there’s no demand left to capture!
Dig deeper: Smarter attribution strategies to help B2B marketers prove campaign value
Performance Max and Advantage+ campaigns
But wait, what about tactics that “collapse the funnel” and drive “full-funnel performance”? Enter Performance Max (PMax) and Advantage+ shopping campaigns (ASC). These tactics exploit advertisers who don’t know how to align them in their favor instead of in favor of Google and Meta’s market cap.
Take PMax, which by default includes branded queries and retargeting audiences. Tell Google you want a ROAS of three, and it will simply balance how much spend goes to branded search and retargeting, with the rest going to non-branded Search/Shopping, YouTube and Display to achieve your blended ROAS.
Want a higher ROAS? Google will allocate more budget to brand and retargeting to get your blended number where you want it while monetizing remnant inventory elsewhere in the Google ecosystem. You get non-incremental spend on one end and wasted impressions on the other — the worst of both worlds neatly wrapped in an attributed ROAS of your choosing. Advantage+ is similar by including existing customers by default (which it is now making harder to exclude).
While I am more cynical about PMax and ASC than most, specifically for their default setting that most small and large brands fail to investigate appropriately, they are not all bad. When paired with incrementality-based measurement, they can drive real value for brands. When you strip away Google’s ability to serve PMax on branded search, for example, it can’t go buy garbage YouTube/Display inventory. It has to find inventory that converts because it isn’t getting conversions from branded search.
The problem is that the PMax or Advantage+ shopping campaign with the best-attributed performance often has the worst incremental impact. That’s why it’s essential for brands to look beyond attribution — especially when evaluating campaigns that blend brand and non-brand traffic or combine prospecting with retargeting.
The way forward: From attribution to incrementality
Many brands focus on attribution due to a misunderstanding of attribution itself. A misunderstanding of what it does and doesn’t tell you. As I often repeat:
- Attribution tells you what happened.
- Incrementality tells you if it mattered.
The simplest way to explain incrementality-based measurement is to compare a treatment group to a control group — a counterfactual that shows what would have happened without the campaign. That’s what establishes causality, and it’s precisely where attribution falls short.
When the goal is profitability, incrementality matters more than ever
Instead of saying, “I drove 100 orders through my ad,” you say, “I drove 100 orders through my ad, but 50 of them were going to happen anyway, so really, my ad generated 50 orders.” Say you spend $10,000 on this campaign at an AOV of $300 and your product margins are 50%.
In an attribution-only world, you think you drive 100 orders, $30,000 in revenue ([orders x AOV]) and $5,000 in contribution margin dollars ([revenue x product margin] – ad spend). You would keep spending.
But in a world where you account for incrementality, you’d know your ads drove only 50 orders, $15,000 in revenue and had a negative contribution margin to the tune of ($2,500). Every dollar spent here on advertising loses you/your company money!
Brands should put far less stock in what happened and shift more energy into understanding if it mattered. Ads that get attribution credit but only drive sales, revenue or leads that were going to happen anyway aren’t worth running. Relying solely on attribution misses opportunities to make investments that drive real, meaningful growth in an increasingly competitive digital marketing ecosystem.
Dig deeper: Measuring marketing incrementality: Best of the MarTechBot
The post How attribution masks what’s actually driving growth appeared first on MarTech.