The $0 CPM problem
The advertising-economics crisis for online publishers as AI chat interfaces replace search-driven traffic. When readers get answers from ChatGPT or Perplexity instead of clicking through to the source, the page never loads and the CPM is effectively zero. Simon uses this framing to argue that the advertiser-first publishing model cannot survive the AI transition.
In depth
CPM, cost per mille, is revenue per thousand ad impressions. It has been the universal unit of publisher economics for two decades. Every open-web publication is valued, operated, and hired against CPM at some level of the stack. The $0 CPM problem names what happens when answer engines intercept the read: the impression never fires, so the revenue line the publication was engineered around goes to zero. Traffic doesn't just decline. It stops translating into money at all.
The punchline is that the model can't be patched. A publisher doesn't fix a $0 CPM by improving creative, tightening the funnel, or adding another ad slot. The read itself is the missing event. Which means the surviving model has to monetise something other than the impression: subscriptions, community, events, licensed data, products, operator leverage. "How do we get the traffic back?" is the wrong question. The honest question is "what do we do now that the page view is no longer the unit of commerce?"
What operators are watching, in real time, is the order in which different categories cross the threshold. Recipe content, product review, how-to, and reference were the first to lose impressions to AI Overviews and chat answers. Service journalism and explanatory content followed. Original reporting, source-relationship-led work, and bylined opinion are holding longer because the value isn't in the answer, it's in the source. The pattern reads as a leaderboard for which publishing categories have to rebuild their model first.
Examples
- mOOnshot digital's portfolio of online publications saw revenue decline as readers moved to ChatGPT, Claude, and Perplexity for the queries that used to drive the impression. The portfolio had previously contributed to over $80M in cumulative sales at 90%+ margins. The decline was severe enough to force ending the fifteen-person editorial team and rebuilding the operation around AI infrastructure. Costs dropped sharply. Revenue did not recover with them. The CPM line was the missing event.
- A recipe site that used to earn meaningful display revenue on a cooking query now loses that reader to the AI Overview. The answer engine quoted the recipe; the page was never loaded; the CPM is effectively zero despite the intent being identical.
- A product-review publisher whose sessions used to start on a Google search now sees ChatGPT synthesise the review and hand a reader a decision without a click. Same editorial work, a much smaller monetisable surface.
Usage notes
The phrase is shorthand for the structural problem, not the literal claim that every ad-funded publisher earns nothing. Some pages still load, some CPMs still hold. The point is that the revenue curve is bending toward zero across the category, and planning any publishing operation as if that bend isn't happening is the real mistake.
Also known as
$0 cpmzero cpmthe 0 cpm problem
These aliases are what the site's build-time auto-linker matches against to cross-reference this term across the FAQ and machine-readable endpoints.