Latest / The CMO Podcast with Fexingo: Marketing Leadership, Budgets, and Executive Strategy / Why CMOs Are Abandoning Vanity Metrics for Revenue Attribution
Transcript
- Lucas: There's a number that haunts CMOs more than any other these days, and it's not cost per click or cost per thousand impressions. It's the percentage of their marketing spend that can be tied directly to a sale. Luna: And that number is surprisingly low for most companies. I've seen estimates that less than a third of ad dollars for big consumer brands have clear revenue attribution. Lucas: Exactly. So a quiet revolution has been happening inside the world's biggest marketing departments. CMOs are abandoning what they can't prove — and that means walking away from metrics they've used for decades. Luna: If these marketing conversations have sparked something you've actually used in your own work, a couple of dollars a month is genuinely what keeps these going — buy me a coffee dot com slash fexingo, if you've gotten something out of them. Lucas: Yeah, honestly, it makes a real difference. And speaking of real impact, let's talk about the person who arguably started this shift. Luna: Marc Pritchard. Lucas: Marc Pritchard, Procter & Gamble's former Chief Brand Officer. Back in 2017, he stood on stage at the Interactive Advertising Bureau conference and basically declared war on the murky metrics the digital ad industry was selling. Luna: He called it the 'media transparency' moment. He demanded third-party verification, clean measurement, and he threatened to pull P&G's billions in ad spend if the platforms didn't comply. Lucas: And they didn't just talk. P&G cut over $200 million in digital ad spend in 2018 and actually saw their sales grow. That was the proof point that every other CMO needed. Luna: So what specifically did Pritchard push for? I know the term 'validated gross rating points' got tossed around a lot. Lucas: Right. Validated GRPs. Traditional gross rating points measure reach and frequency based on panel data and surveys. But validated GRPs use set-top box data, point of sale data, and digital signals to confirm that an ad actually reached a human in the intended demographic and that it correlated with a sales lift. Luna: It's basically moving from 'we think this ad reached people' to 'we know this ad reached people and they bought.' But that requires data integration most companies don't have. Lucas: True, but the technology has gotten cheaper. There are now a dozen vendors that can link ad exposure to transaction data at the household level. And the big holding companies — GroupM, Publicis — have built their own attribution engines. Luna: But isn't there pushback? I've heard agency executives argue that full attribution is impossible, especially for brand campaigns that work over months. Lucas: You're right — there's genuine resistance. But what I find interesting is that the resistance often comes from the same agencies that sell the metrics being questioned. There's a conflict of interest. If you can't prove your TV spot sold toothpaste, the client might stop buying TV spots. Luna: So who's actually adopting this? Is it just CPG giants like P&G and Unilever? Lucas: It's spreading. A mid-tier DTC skincare brand — I won't name them, but they do about $50 million in annual revenue — hired a measurement consultancy last year. They found that 40 percent of their digital ad spend was going to placements that never led to a single transaction within 90 days. Luna: Forty percent. That's huge. Lucas: They cut those campaigns, reallocated budget to channels that did correlate with sales, and within six months they had reduced ad spend by 30 percent while growing revenue by 12 percent. That's the kind of story that gets CFOs to smile. Luna: But for B2B companies with long sales cycles — six months, nine months — how do you even start to attribute? The data lag is real. Lucas: It's harder, for sure. But some B2B firms are using matched-market experiments. You pick two similar territories, run ads in one but not the other, and compare pipeline generation six months later. It's not perfect, but it's better than relying on last-click attribution from a CRM. Luna: That's essentially a clinical trial for marketing. I've seen Salesforce do that with their own campaigns. They actually reduced spend in some mature markets and saw no drop in qualified leads. Lucas: Because the brand had already done its job. That's the nuance here — this isn't about killing brand advertising. It's about measuring it properly. Pritchard never said brand ads don't work. He said he wanted to measure brand ads with the same rigor as direct response. Luna: And what did the industry do? Did platforms like Facebook and Google willingly provide more transparent data? Lucas: Not without a fight. Google eventually introduced brand lift measurement tools. Facebook — now Meta — tightened viewability standards. But there's still a lot of opacity. A 2025 study from the Association of National Advertisers found that 70 percent of digital ad impressions never reached a human. Bots, fraud, and misattributed placements. Luna: Seventy percent. That's staggering. So the push for revenue attribution is also a fight against ad fraud. Lucas: Exactly right. If you can show that a certain publisher's impressions don't correlate with any sales, you can pull spend. That's what happened with some programmatic exchanges — they lost billions in CPG ad dollars because they couldn't prove their inventory worked. Luna: So the CMO's job is shifting. It used to be about creativity and brand building. Now it's part data scientist, part forensic accountant. Lucas: And the CMOs who succeed in the next five years will be the ones who can speak both languages. They'll have a creative vision, but they'll also demand the receipts. I think we'll see more chief marketing officers come from analytics backgrounds, not just agency creative directors. Luna: That's already happening. I know of a Fortune 500 company that just hired a CMO whose previous role was chief analytics officer at a retail bank. She's already killed half their trade show budget because the attribution data showed zero correlation with closed deals. Lucas: Trade shows are a sacred cow in B2B. If attribution can slay that, it can slay anything. So the question I have is: will the ad platforms eventually open up enough to satisfy CMOs, or will we see more brands bring measurement in-house? Luna: I think it'll be a hybrid. The platforms have the best data, but they have no incentive to share it truthfully. So brands will build their own attribution stacks and use platforms for execution, not reporting. Lucas: That's a future I can get behind. And it all started with one CMO who refused to accept 'trust us' as an answer. Luna: Marc Pritchard. He really did change the game, quietly, by just demanding better data. Lucas: Quietly, but loudly enough. And I think every marketer listening today should ask themselves: what percentage of my budget can I actually prove works? If the answer is below fifty percent, there's an opportunity.