People Inc. CEO holds off blocking Google crawlers despite AI traffic loss

People Inc. CEO Neil Vogel told analysts what every open-web publisher already knows but won't say plainly: Google has them locked in, and they can't break free. Search traffic is down 40% year over year as AI Overviews absorb the clicks. Blocking Google's crawlers would stop the AI scraping, but it would also kill the search referrals still driving 21% of People Inc.'s traffic. So the crawlers stay on. Vogel calls it a lever. It isn't one he can pull.

The non-session revenue number ($125 million, growing 16%) sounds like a turnaround story. It isn't, quite. Non-session went from 39% to 43% of digital revenue, but part of that mix shift is the session denominator collapsing underneath it. A percentage climbs while the dollars fall. The earnings call didn't headline the absolute-dollar decline on the session side, which is where the real damage is.

The margin question is the one nobody answered. Events cost money to run. Licensing deals like the Apple News arrangement are lumpy and negotiated one at a time. Replacing high-margin programmatic revenue with lower-margin, higher-effort revenue can keep the top line flat and still gut the bottom line. That math matters more than the mix percentage.

Who takes the hit

SSPs (the automated selling platforms) like Magnite and PubMatic sit directly in the path of this. Their pitch is open-web scale. The scale is leaking. Fewer real sessions on premium publishers means fewer quality impressions in the open auction, which compresses CPMs (the price per thousand ad views) and pushes buy-side dollars toward walled gardens.

The beneficiaries are the walled gardens that own their audiences outright, and the first-party data infrastructure, companies like LiveRamp, that let publishers sell what they know about their readers rather than raw session impressions. Advertisers aren't asking publishers to pivot to events. They're asking where the quality reach went. Increasingly the answer is Google, Meta, and Amazon.

The crawler bluff

Vogel frames the crawler block as negotiating leverage for better licensing terms from Google. But if no publisher can afford to pull it individually, and none can, it produces no leverage. It's a threat Google can see through. The only version that works is collective action, which runs into antitrust and coordination problems that don't resolve quickly.

Our call: Through the end of 2026, no major US general-interest open-web publisher (People Inc., Dotdash Meredith, Condé Nast, Hearst) will fully block Google's AI crawlers, and each will keep reporting double-digit year-over-year search-traffic declines on their next earnings prints. The trap Vogel described is shared and structural. AI Overviews answer queries without sending the click, and Google has every incentive to lean further into that, not pull back. Blocking requires either a licensing deal rich enough to replace a fifth of your traffic, or industry-wide coordination. Neither exists today.

People Inc. CEO Neil Vogel just told the market something every open-web publisher already knows in their gut but won't say on an earnings call: Google has them by the throat, and they can't let go. The company is watching Google search traffic fall 40% year over year while AI Overviews eat the clicks, but blocking Google's crawlers would also kill the search referrals that still feed 21% of its traffic. So the crawlers stay on. Vogel calls it a lever. Right now it's a lever nobody can pull.

What's actually being decided: not "should People Inc. block crawlers" (they won't, yet), but whether the open-web publishing model can pivot off session-based ad revenue fast enough to outrun the search collapse. This is a Type 1 problem for the industry: once search referral traffic evaporates, it does not come back. The forcing function is the quarterly print. Every earnings call now carries a search-decline number, and each one gets worse.

The Market Analyst. In plain terms: the machine that sent free readers to publishers for 15 years is winding down, and everyone downstream of that traffic is exposed. Watch where the pain lands. Session-based inventory shrinks in both volume and quality, so open-market CPMs (the price for a thousand ad views bought in the automated auction) keep sliding. The SSPs that aggregate open-web publisher supply, Magnite and PubMatic, sit directly in that current. Their pitch is scale; the scale is leaking. The winners are the walled gardens that own the audience outright and the first-party data plumbing, LiveRamp and clean-room players, that let publishers sell what they know about readers instead of raw impressions. Consensus still treats search decline as cyclical. It isn't.

The Skeptic. Vogel says People Inc. is "nearly on the other side" of search mattering. With search traffic down 40%? That's a man narrating his escape while still in the room. Non-session revenue at $125 million growing 16% sounds like a turnaround until you ask what the session side is bleeding in absolute dollars, a figure the call conveniently didn't headline. And the crawler lever is the tell. If it were real optionality, they'd have modeled pulling it. They can't, because search referrals still pay the light bill. That's not a negotiating position. That's a hostage explaining, calmly, that the situation is under control.

The Operator. Tuesday morning at Dotdash Meredith, Condé Nast, Hearst, the revenue ops teams are running the same math and it's ugly. Open-market yield compresses first, before direct sales even flinch, because programmatic is where shrinking, lower-signal impressions get dumped. The org chart tells the story of who believes what: audience development and first-party data teams get headcount, SEO teams get quietly reorganized into "search and AI visibility" and told to do more with less. The email list and a tool like D/Cipher stop being upsell line items and become the thing keeping the P&L alive. What breaks at 90 days is any budget premised on flat session volume.

The CFO. Follow the mix, not the growth rate. Non-session revenue went from 39% to 43% of digital. That shift is partly People Inc. executing, and partly the denominator shrinking underneath them. A percentage can climb while the dollars beneath it fall. The real question for any publisher CFO copying this playbook: does licensing and events revenue carry anything close to the margin that programmatic did at scale? Events cost money to run. Licensing deals like the Apple News arrangement are lumpy and negotiated one at a time. Replacing high-margin automated ad revenue with lower-margin, higher-effort revenue can keep the top line flat and still gut the bottom line. Nobody put that number on the call.

The Customer / End User (the advertiser). From the buy side, this is a supply-quality story before it's a publisher story. Fewer sessions on premium open-web sites means fewer chances to reach real humans in a brand-safe context, and more budget sloshing toward made-for-advertising junk and the walled gardens. A tool like D/Cipher matters to advertisers only if it delivers targeting that survives without the cookie and without the search-referral audience. That's the actual test. Advertisers aren't asking publishers to pivot to events. They're asking where the quality reach went, and increasingly the answer is Google, Meta, and Amazon, which is exactly the concentration they've spent a decade complaining about.

Where the council splits. Three real disagreements. First, the Strategist read in the briefing says People Inc. is becoming an IP holder that sells structured data, while the CFO and Skeptic say the mix shift is flattered by a collapsing denominator and unproven margins. Second, the Market Analyst treats the search decline as structural and permanent; the sunk-cost instinct across the industry treats it as a rough patch. Third, Vogel frames the crawler as a lever for better licensing terms, but if no publisher can afford to pull it, it isn't leverage at all, it's a bluff Google can see.

What it hinges on. Two beliefs. One: whether non-session revenue carries enough margin to replace programmatic dollars, not just percentage of mix. Two: whether the crawler block is a credible threat or an empty one. If publishers can't pull the lever individually, the only version that works is collective, and that's an antitrust and coordination problem, not a strategy. The council leans toward the Skeptic and Analyst: the search decline is permanent, the pivot is real but the economics are unproven, and the crawler "lever" is mostly rhetoric until someone can actually afford to use it.

What to verify before copying the playbook: the absolute-dollar decline of session revenue, not the mix percentage; the gross margin on events, licensing, and D/Cipher versus programmatic; and whether any AI licensing deal is recurring or a one-time check.

Prediction: Through the end of 2026, no major US general-interest open-web publisher (People Inc., Dotdash Meredith, Condé Nast, Hearst) will fully block Google's AI crawlers, and each will keep reporting double-digit year-over-year search-traffic declines on their next earnings prints.

Confidence: High. The search dependency Vogel described is shared and structural, not company-specific.

Why: Vogel spelled out the trap on the call: blocking Google's AI crawler also cuts search referrals, and search still drives 21% of People Inc.'s traffic and real ad dollars. Every scaled open-web publisher runs on the same wiring, so the same math holds for all of them, and none can afford to unilaterally cut off a fifth of their traffic to make a licensing point. The declines keep coming because AI Overviews answer queries without sending the click, a mechanism that gets worse as Google leans further into it, not better. The opposite outcome, a publisher actually pulling the lever, would require either a licensing deal rich enough to replace search or industry-wide coordination, and neither exists today.

Revisit by 2026-12-31: We're right if no major US open-web publisher has blocked Google's AI crawler and Q3/Q4 earnings show continued double-digit search-traffic declines. We're wrong if one of these publishers blocks the crawler outright, or if reported search traffic stabilizes.