Latest / The CMO Podcast with Fexingo: Marketing Leadership, Budgets, and Executive Strategy / Why Brand Safety Tools Block Your Own Ads
Transcript
- Lucas: Luna, I want to talk about something that sounds like a first-world marketing problem but is actually burning real budget right now: brand safety tools that block your own ads. Luna: Wait — you mean the filters we put in place to keep ads away from sketchy content are accidentally blocking our own campaigns? Lucas: Exactly. I came across a case recently — a Fortune 500 CPG company. They ran an audit of their programmatic buys in Q1 2026 and found that twelve percent of their own digital ad impressions were blocked by their own brand safety filters. Luna: Twelve percent. That's enormous. How does that even happen? Lucas: It's a classic over-engineering story. The company had built a massive keyword block list over the years — thousands of terms — to avoid anything remotely controversial. But here's the thing: brand safety tools operate on a match logic. If your ad is set to run on a page that mentions 'bomb' in a recipe blog post about a baked potato, the tool says no. Luna: Oh, I see. So neutral or even positive uses of those words get caught in the net. Lucas: Right. And it gets worse. The audit found that the brand's own content — its own website articles, its own sponsored posts, its own corporate blog — was being blocked because the keyword lists flagged words like 'crash' or 'death' or 'crisis.' Words that appear in a perfectly normal earnings report or product recall announcement. Luna: So they're paying to run ads, but the safety tool is blocking the ad from appearing on their own pages. That's a circular waste. Lucas: It's a double waste. You pay for the impression, you pay for the safety verification, and the impression never serves. But the CPM is still charged. Luna: What's driving this? Is it just lazy keyword lists, or is the technology itself to blame? Lucas: It's a bit of both. The industry standard has been keyword-only filters for years because they're cheap and easy. But context matters. A word like 'violence' in a news article about peace negotiations is very different from 'violence' in a graphic crime report. Keyword filters can't tell the difference. Luna: So what's the alternative? Contextual AI targeting has been around for a while now. Lucas: It's gaining real traction in 2026. Instead of just matching words, contextual AI looks at the entire page — the sentiment, the topic, the surrounding text. It can say, 'This article is about conflict resolution, not actual violence. Safe to serve.' The CPG company I mentioned moved to a contextual AI layer and recovered about eight percent of their previously blocked impressions. Luna: Eight percent of twelve percent — so roughly one percent of total impressions. For a Fortune 500 CPG, that's probably millions of dollars in reclaimed media. Luna: And it's not just CPG. Any brand with a big programmatic budget is leaving money on the table if they haven't audited their block lists recently. Lucas: A hundred percent. The real lesson here is that brand safety isn't a set-it-and-forget-it thing. Block lists need to be reviewed quarterly — ideally by a human who understands the nuance of the brand's own content and the media environment. Luna: And yet, a lot of CMOs are worried that if they loosen the filters, they'll end up next to something terrible. It's a trust issue. Lucas: It is. But the alternative — keeping the filters tight and losing twelve percent of your reach — means you're leaving audience on the table. The sweet spot is to use contextual AI to differentiate between truly toxic environments and merely ambiguous ones. Then you can tighten the rules on the toxic stuff and relax them on the ambiguous stuff. Luna: Have you seen any brands publicly address this? Or is it still a behind the scenes issue? Lucas: Mostly behind the scenes, because admitting 'our brand safety tools block our own ads' sounds like a failure. But the smart CMOs I talk to are quietly running audits and retooling. One told me they found their block list was blocking their own job postings because the word 'crisis' appeared in a press release about a new initiative. Luna: Wow. So the filter was basically telling the company, 'We don't want to advertise next to your own hiring page because we don't trust your brand.' That's absurd. Lucas: It is. And it points to a broader issue: the programmatic ecosystem is so complex that even the people running it don't fully know what's happening inside the black box. The only fix is to bring more transparency into the supply path. Luna: Right. And that's part of why we love digging into these topics on the show. If these marketing conversations have sparked something you've actually used or challenged your thinking, it turns out that listener support is what keeps this podcast ad-free and independent. There's a buy me a coffee dot com slash fexingo page where a simple one-time — or even monthly — contribution makes a real difference. Lucas: It genuinely does. And we don't run ads, so the show stays lean and focused on what matters: the actual numbers and decisions. Every bit of support helps us keep doing that. Luna: Okay, back to the block list problem. One thing that surprised me: the CPG company also found that some of its competitor's ads were appearing on its own blocked pages. So the brand safety filter was effectively giving away inventory to rivals. Lucas: That's the hidden competitive cost. If your brand safety tool blocks your ad from running on a page, the auction still clears — but a competitor's ad wins the impression instead. So you're not just losing your own reach; you're handing reach to a rival. Luna: That has to be a hard conversation in the quarterly review. 'Hey, our brand safety strategy is actually helping our competitors.' Lucas: Exactly. And it's not just about keyword lists. Some brands also block entire categories of content — news, politics, even weather — out of an abundance of caution. But those pages often have high engagement and broad audiences. By avoiding them, you're essentially ceding the most desirable inventory to competitors who are more sophisticated about risk. Luna: So what's the practical takeaway for a CMO listening right now? What should they do on Monday morning? Lucas: Three things. First, run a full audit of your current block list. Pull a report of every keyword and category you're blocking, and cross-reference it with the actual pages your ads were eligible for in the last quarter. You'll almost certainly find over-blocking. Luna: Second? Get a contextual AI layer in place. Don't rely on keywords alone. Use a tool that evaluates the entire page context. Lucas: And third, set a recurring quarterly review with your ad ops team. Block lists drift over time as new terms get added without much thought. A regular scrub keeps the list lean and current. Luna: Lean and current — that's a good mantra. It's almost like a diet for your programmatic strategy. Lucas: Exactly. And the payoff is immediate. The CPG company I mentioned recovered eight percent of their blocked impressions within a month of switching to contextual AI. That's eight percent more reach, eight percent more value from their existing media budget. Luna: That's a pretty compelling case for any CMO who's feeling budget pressure right now. And that's every CMO, I think. Lucas: Every single one. The question is whether you're willing to look inside the black box and admit that your own safety tools might be working against you. Luna: Alright, I think we just wrote the agenda for a lot of Monday morning meetings. Lucas: I hope so. Because the fix is straightforward, and the upside is real. Thanks for coming along on that deep dive.