Latest / The CMO Podcast with Fexingo: Marketing Leadership, Budgets, and Executive Strategy / Why CMOs Are Killing Their Own Performance Ads
Transcript
- Lucas: So there's this moment every CMO knows — you pull up the dashboard, you see cost per acquisition is down twenty percent quarter-over-quarter, ROAS is three-point-five, and your CFO is nodding. Everything looks clean. Luna: That's the dream, right? The green light on the spend. Lucas: Right. But here's the thing — some of the smartest CMOs I've talked to lately are doing the opposite of scaling those campaigns. They're deliberately pausing their highest-performing paid search and social ads. Luna: Wait — the ones that look best in the dashboard? Why would you kill the golden goose? Lucas: Because they've started to believe that dashboard is lying to them. Or at least telling an incomplete story. Let me give you a concrete example — a mid-market DTC brand in the home goods space. They had a Google Ads account that was generating what looked like a four-point-two ROAS. Textbook efficiency. The CMO decided to run an experiment: she paused all brand search for two weeks. Luna: Brand search meaning someone types her company name and clicks the ad instead of the organic result? Lucas: Exactly. That's the cannibalization zone. When she paused those ads, the brand's overall revenue from organic and direct traffic actually went up by six percent. Because a chunk of those clicks were people who would have found her site anyway — she was paying for traffic she already owned. Luna: So the ROAS was inflated by counting conversions that would have happened without the ad. That's incrementality — or lack of it. Lucas: Exactly. And that's just brand search. There's a bigger problem with non-brand campaigns too. I spoke with a CPG company that ran a controlled incrementality test — geo holdouts — on their Google Shopping ads. They found that thirty to forty percent of the attributed conversions were actually happening regardless of the ad exposure. Luna: That's huge. That basically means a chunk of your 'efficient' spend is just overhead. Lucas: It gets worse. Because the platforms are also stacking ads — serving multiple impressions to the same user across devices and attributing each one as incremental. So you think you're reaching three distinct people, but it's one person on three screens. Luna: Right — the identity problem. Without a unified ID, you're double-counting. Lucas: Now, this doesn't mean performance advertising is dead. But it does mean the metric that CMOs have been optimizing for — last-click ROAS — is increasingly unreliable. Especially as privacy changes fragment the measurement landscape. I think we're seeing a shift toward incrementality testing as the new gold standard. Luna: That requires a different kind of rigor though. You need holdout groups, geo experiments, proper measurement partners. That's not the agile move-fast culture most performance teams have. Lucas: Right. And that's where the CMO's role becomes really interesting — they have to be willing to slow down spend growth to build better measurement. The CPG company I mentioned, they cut their Google Ads budget by forty percent after their tests. But guess what happened to their overall revenue over the next quarter? Luna: I'm guessing it didn't drop forty percent. Lucas: It dropped less than two percent. So they basically freed up a huge chunk of budget to reinvest into channels that actually drove net-new demand — like retail media networks, influencer partnerships, even direct mail. Luna: Direct mail — the ultimate non-digital channel. But you see that in a lot of DTC brands now. They're realizing physical touchpoints create a different kind of memory structure. Lucas: Absolutely. And this brings us to a broader point — the zero-click economy. More and more searches end without a click — people get their answer in the featured snippet, or they watch a YouTube thumbnail, or they interact with a chatbot. So the old assumption that a search ad leads to a site visit and then a purchase is breaking down. Luna: So where does that leave the traditional performance marketer? I mean, their whole career has been built on optimizing the click. Lucas: Yeah, and I think that's the painful part. The CMOs I'm talking to are having to retrain their teams — or in some cases rebuild them. They're looking for people who think in terms of total business outcomes, not just campaign-level efficiency. It's a shift from 'how do I lower my CPA' to 'how do I grow my total addressable market'. Luna: That's a much harder question. It requires blending brand and performance into one P&L. Lucas: Exactly. And that's why this is a cmo level conversation, not a media buyer conversation. The media buyer is incentivized to spend the budget and report a good ROAS. The CMO has to ask — is that ROAS real? Is it incremental? Is there a better use of that dollar? Luna: Let's talk about one more layer — the role of creative. If you're killing high-volume campaigns, you're also losing data signals. Some CMOs worry that without the volume, they lose the feedback loop for creative testing. Lucas: That's a real tension. But there's an argument that the signal to noise ratio actually improves. When you're running fewer, more targeted campaigns with proper incrementality measurement, the feedback is cleaner. You're not confusing correlation with causation. Luna: So the CMO who pauses those ads isn't anti-performance. They're pro-incrementality. They want to know what actually moves the needle. Lucas: Right. And look, I'm not saying every brand should cut their Google Ads budget by forty percent. But I think every CMO should run the experiment — pause brand search for two weeks, set up a geo holdout, or even just reduce spend on your top campaign by twenty percent and see what happens to total revenue. The answer might surprise you. Luna: It reminds me of that old story about the retailer who cut their catalog mailings by half and sales actually went up — because they were mailing to people who already shopped in store. Same principle. Lucas: Same principle. The problem is we've built entire marketing operations around attribution models that were designed in a simpler digital era. And the platforms have every incentive to keep those models opaque. They want you to keep spending. Luna: If today's episode has sparked something you're considering testing — maybe that pause on brand search or a geo holdout — that's exactly the kind of practical takeaway we hope you get here. And the reason we can keep having these conversations without ad breaks is listener support. If it was useful, you can throw a few dollars in the hat at buy me a coffee dot com slash fexingo. No pressure, just a way to keep the show independent. Lucas: Yeah, and we really appreciate those who do. It lets us keep digging into topics that matter to you — not just what's trending. So thank you for that. Luna: Alright, back to the numbers. So we talked about DTC and CPG. What about B2B? Are CMOs in that space seeing the same issues with performance ads? Lucas: Great question. B2B is interesting because the sales cycle is longer and the attribution problem is even harder. A lot of B2B CMOs I've spoken with are pulling back on LinkedIn lead gen ads — not because they don't work, but because they can't prove incrementality. They're seeing high cost per lead but low conversion to opportunity. So they're shifting toward account-based marketing and direct sales touchpoints. Luna: So the trend is cross-industry, but the tactics differ. For B2C, it's brand search and shopping ads. For B2B, it's LinkedIn and content syndication. Lucas: Exactly. And one thing that's consistent across both is the need for better measurement infrastructure. That means investing in a good attribution partner — something like a Measured or a Rockerbox — or building a data clean room if you're big enough. The CMOs who are ahead of this are the ones who have already started the conversation with their CFO about moving from ROAS to incrementality. Luna: That's a tough conversation though, because the CFO sees a campaign with a 4x ROAS and it's easy to defend. If you say 'I want to cut it to test incrementality,' the CFO might think you're just hedging. Lucas: Absolutely. That's why the test needs to be structured. You don't just cut — you run a controlled experiment and present the results. The CFO loves data. If you show them that the brand search pause led to revenue growth, they'll be on board. But you have to do the work to prove it. Luna: I've also noticed some CMOs are using this as an opportunity to renegotiate with the platforms. If Google sees you pulling spend, they might offer better terms or more transparent data. Lucas: That's a smart tactical move. The platforms are oligopolistic, but they're not immune to share shifts. If enough advertisers demand incrementality, the platforms will adapt. We're already seeing Google's ads data hub and Amazon's marketing cloud offer more granular measurement — but it's still a walled garden. Luna: So the takeaway for CMOs listening: don't trust your dashboard blindly. Run one incrementality test this quarter. Start small — pause brand search or a top campaign — and measure total revenue impact. Lucas: Yeah. And then use that learning to reallocate budget into channels that actually create net-new demand. The future of marketing isn't about optimizing the click — it's about optimizing the business outcome. And that starts with knowing what's incremental and what's just overhead. Luna: Great episode. Thanks, Lucas. Lucas: Thanks, Luna. Next time we'll look at how CMOs are using first-party data to fuel ai driven creative — but that's a whole other conversation.