News & Digital Publishers

Start by separating the causes before fixing anything. A Discover drop is usually one of four things: a core or Discover-specific update, an eligibility problem such as missing large-image markup or a news indexation fault, a headline and thumbnail pattern that the system has started treating as clickbait, or simple topic decay where the stories that were feeding the feed stopped happening. Google ran its first standalone Discover core update from 5 to 27 February 2026 — initially for English-language users in the United States, with wider rollout following — so a trend line crossing that window needs to be read against it, and publishers outside that scope should be slower to blame it.

It can be influenced, but not ranked in the ordinary sense. There is no query and no position, so the levers are eligibility and quality signals: large-image markup on every article, clean news indexation, headlines that describe the story rather than withhold it, demonstrated expertise on a specific topic rather than broad coverage, and publishing consistency. We measure it as a share of sessions over time, not as a position.

Explainer content keeps earning impressions and stops earning sessions, which is why publishers see flat impressions and falling clicks on exactly the pages that used to hold the revenue floor. The response is to shift evergreen output toward what an overview cannot replace — original data, first-hand reporting, tooling, and opinion with a named author — and to stop measuring those pages on click-through alone.

We work with it. The engagement is measurement, technical setup, and editorial process — indexation, schema, headline and image standards, topic authority planning, and a reporting line the newsroom can act on. Commissioning and reporting stay with your editors, who know the beat better than any agency will.

On sessions by surface and on revenue per thousand pageviews, tracked together. Traffic mix matters as much as traffic volume, because Discover sessions and search sessions do not monetise identically, so a rising session count with a falling revenue figure is a real outcome that a traffic-only report hides.

It works the surface your traffic actually arrives on. The clear majority of Google's traffic to news publishers now comes through Discover rather than Search, and most publishers take more from Discover than from Search. Discover has no query report, no rank tracker, and no keyword plan — so a general agency's entire toolkit points at the smaller half of your traffic. Publisher work is indexation velocity, headline and image standards, topic authority, and Discover-specific diagnosis when volatility hits, none of which a standard search retainer covers.

We do not publish a standard figure. Discover exposure, beat competitiveness, publishing capacity, and current technical health all change what is achievable. Before work starts we record the existing split across Discover, Search, and direct along with revenue per thousand pageviews, then report movement against that baseline.