Why the old playbook is dead (and why that's good news for you)

The version of SEO most people still picture, write a tidy "what is X" explainer, rank it, collect the clicks, is structurally finished. By November 2025, AI Overviews appeared on a meaningful share of all keyword searches, having peaked at just under a quarter of queries earlier in the year, which means the answer to the generic question is now often taken before the click ever happens. If your entire play was to be the page that answers "what is lead scoring" in 800 competent words, that play no longer pays, because Google answers it inline and the searcher never arrives.

This is exactly the news a founder should want, even though it arrives dressed as a threat.

The doom narrative is half wrong

The loud version of this story is that AI Overviews killed SEO outright. The data says otherwise. When Semrush analysed ten million keywords, zero-click rates on queries showing an AI Overview actually fell slightly, from 33.75 per cent to 31.53 per cent. Read that carefully, because it inverts the panic: the click did not disappear, it got pickier. The thin explainer lost its click to the inline answer, but the click that survives now flows to the deep, genuinely cited source. AI Overviews ended the generic click. They did not end the considered one.

So the death here is specific. Generic content died. First-hand content got more valuable, because it is the thing the inline answer cannot replace and the thing the surviving searcher is hunting for.

Volume without value decays faster than ever

If you needed proof that churning out generic pages is now actively worthless, a sixteen-month experiment on 2,000 AI-generated articles supplies it. Seventy-one per cent of those pages were indexed within 36 days and they briefly pulled over 526,000 impressions at their peak, which looks like a win for about a month. Then it collapses: by month three, only three per cent of the pages still ranked anywhere in the top 100. Volume with no authority, no uniqueness and no first-hand signal decayed almost completely inside ninety days.

That is the precise fate of the work a content team used to do at scale, and it is why their old advantage has evaporated.

Generic was the team's one trick

Put the two facts together and the opening becomes obvious. The single thing a content team reliably produced, competent generic coverage, is now both commoditised by AI and structurally devalued by AI Overviews. The thing they could never produce, lived operator experience, is the only thing that survives. The death of generic content is not a loss for the lean founder. It is the removal of the one moat a bigger, better-resourced competitor used to have over you.