If you were waiting for the thing that replaces Google traffic, this week’s publisher earnings round-up delivered the answer: there isn’t one. There are six smaller things, and the publishers furthest along have accepted the operational cost of running all of them at once.
The clearest articulation came from People Inc., which has cut Google search to 21% of traffic, down from 25% just last quarter. More importantly, it grew non-session revenue from 39% to 43% of digital revenue. CEO Neil Vogel put the position plainly: “We are nearly on the other side of search being a material driver of value for us. But we’re not there yet.”
The rest of the round-up, reported by Digiday on 13 August, shows the same migration by different routes. Dow Jones now takes 81% of revenue from circulation against just 17% from advertising — a reader-revenue business with an ads sideline, not the reverse. USA Today Co. grew digital-only subscription revenue 6.8% in the same period its digital advertising fell 9.2% — the two lines crossing in real time. And Ziff Davis supplied the number that explains the urgency: roughly 50% of relevant Google queries now trigger AI Overviews, up from 36% a quarter ago.
That last figure deserves a beat. AI Overviews went from a third of relevant queries to half of them in three months. Whatever your diversification timeline was, Google’s is faster.
The numbers in this piece
01The metric that matters: revenue that needs no session
The habit of measuring Google dependency in traffic share is obsolete, and People Inc.’s reporting shows what replaces it. Traffic share tells you where sessions come from; it does not tell you what happens to revenue when they stop coming. The better question — the one these companies now answer for investors — is: what percentage of your digital revenue requires a Google-originated session to exist?
Run that split honestly and the categories sort themselves. Subscriptions, licensing, newsletters, events, commerce and direct-sold sponsorships survive a search collapse, because they are anchored to a relationship or a contract rather than a referral. Open-auction programmatic on search-referred sessions does not — it is the revenue most directly indexed to the traffic that AI Overviews is eroding. A publisher whose monetisation is 80% open-auction on referred traffic and a publisher at 43% non-session-based revenue can have identical traffic profiles today and utterly different survival profiles at the same erosion rate.
People Inc.’s 43% is the benchmark worth stealing — not because 43% is a magic number, but because the metric reframes the goal. The task is not “replace the traffic.” Nobody in this earnings round-up replaced the traffic. The task is “make revenue that doesn’t need the traffic,” and every company showing progress this quarter is doing that instead.
02A portfolio costs more to run than a pipeline
The honest half of the portfolio story is the operational bill. Google search was brutal economics but simple operations: one channel, one discipline, one dashboard. A portfolio of subscriptions, newsletters, licensing, events, commerce and direct sales needs an owner per channel, measurement per channel, and — hardest of all for organisations used to search-scale numbers — a tolerance for individually unimpressive results that only work in aggregate. A newsletter that drives 3% of revenue looks like a failure next to search’s former contribution. Six such channels are the entire strategy.
USA Today Co.’s crossing lines make the same point from the transition’s midpoint: digital subscriptions growing 6.8% while digital advertising falls 9.2% is what the middle of this migration actually looks like — the new engine spinning up while the old one winds down, with total revenue caught between them. Publishers should expect that squeeze and budget through it rather than reading it as failure.
The most useful thing about this earnings round-up is its refusal to offer a hero.
03Why this matters for publishers
| The erosion is the baseline, not a scenario | AI Overviews on half of relevant queries — up from 36% in one quarter — means search decline is no longer a risk you plan against but an environment you operate in. Plans still built on "if Google traffic declines" are a quarter behind the data. |
|---|---|
| Revenue exposure, not traffic share, is the number your board needs | People Inc. cut Google to 21% of traffic, but the figure it leads with is 43% of revenue that needs no session. That is the disclosure standard the market is converging on, and you should measure yourself the way the leaders report. |
| The advertising-only model is the exposed model | Dow Jones at 81% circulation and USA Today's diverging lines both point the same way: publishers with a paying-reader relationship are absorbing the search transition; publishers monetising referred sessions with open-auction ads are absorbing the damage. |
| There is no product to wait for | No AI platform, licensing scheme or discovery channel in this round-up replaces search-scale traffic. Waiting for the one big substitute is a strategy of delay; the working strategy is many small substitutes, run simultaneously. |
04What publishers should do
05The bottom line
The most useful thing about this earnings round-up is its refusal to offer a hero. Nobody found the Google replacement; the companies making progress stopped looking for one and built portfolios instead, accepting higher operational cost as the price of lower existential risk. Vogel’s framing — nearly on the other side, but not there yet — is where the leaders are, and the Ziff Davis figure is the clock running on everyone else. Budget your diversification against that trajectory, not against last year’s referral report.