Google Discover for UK Publishers: Eligibility, Optimisation & Volatility

What Google Discover is and why it matters for UK publishers

Google Discover is the part of Google Search that shows people a personalised feed of content related to their interests, based on their Web and App Activity — no query required. It appears on mobile surfaces including the Google app and Chrome. Where classic search answers a question someone typed, Discover proactively pushes stories to people Google believes will find them interesting.

That distinction matters commercially. A Discover card can put a story in front of a large mobile audience that never searched for the topic, which is why Discover has become a significant referral channel for many UK publishers — in some newsrooms it is discussed alongside, or ahead of, classic search referrals. Yet it is rarely covered in audience strategy with the same rigour as SEO or social, partly because there is no keyword research to do and no feed to game.

The honest framing for a UK newsroom: Discover is upside, not foundation. Google itself warns that traffic from Discover is “less predictable or dependable” than keyword-driven search visits and advises treating it as supplemental to search traffic. Everything in this guide follows from that: do the things that make Discover more likely, measure it properly, and never build a budget on it.

Eligibility: how content gets into Discover

• 1Content is automatically eligible if it is indexed by Google and meets Discover’s content policies — there is no submission process, no application, and no special feed to register.
• 2No special tags or structured data are required for Discover eligibility, per Google’s documentation. Standard technical SEO (crawlable, indexable pages) is the entry ticket.
• 3Google’s content policies still apply — content that violates them (dangerous, deceptive, manipulated media, and similar categories) is excluded regardless of quality.
• 4Selection is algorithmic and interest-based: Discover matches content to individual users’ interests, so there is no single “Discover audience” to target — only topics and quality signals.
• 5Being eligible is not being included. Most indexed pages never appear in Discover; the sections below cover what Google says increases the likelihood.

What Google says increases your chances: the checklist

These points come from Google's own Discover documentation and helpful-content guidance — not from SEO folklore.

• Use large, high-quality images: at least 1200px wide, high resolution, declared via og:image or schema.org markup. Image-led cards dominate the Discover feed.
• Enable large image previews with max-image-preview:large in your robots meta tag (or use AMP) — Google’s Discover documentation recommends this explicitly, and the setting applies to Discover as well as web search.
• Write page titles and headlines that capture the essence of the content — Google’s wording — rather than headlines engineered purely for curiosity.
• Avoid clickbait and sensationalism tactics that manipulate appeal; Google warns against withholding key information or exaggerating to provoke clicks.
• Publish content that is timely for current interests, tells a story well, or provides unique insights — the three qualities Google names for Discover.
• Meet the people-first content bar: demonstrable experience, expertise, authoritativeness and trust (E-E-A-T), with trust the most important of the four per Google’s guidance.
• Keep author and publication transparency strong: clear bylines, dates, and editorial standards pages support the trust signals Google’s systems look for.
• Do not optimise RSS feeds for the Discover Follow feature — Google removed Follow-feature guidance from its Discover documentation in November 2025 because the feature is no longer shown in Discover.

Headline craft for Discover: specific beats sneaky

Discover rewards headlines differently from search. There is no query to match, so keyword placement matters less; what matters is whether the headline-plus-image card makes a reader stop scrolling. The newsroom temptation is the curiosity gap — “This seaside town has a problem nobody saw coming” — and in the short term such headlines can lift clickthrough. But Google explicitly warns against clickbait and tactics that manipulate appeal, and publishers who lean on them report the channel punishing it over time.

The craft that holds up in practice: front-load the concrete, surprising specific rather than hiding it. “Whitby votes to ban second-home owners from new builds” will outperform a teaser version with UK audiences who see dozens of curiosity-gap cards a day — and it survives algorithm updates aimed at clickbait. Pair it with a strong, relevant 1200px image rather than a generic stock photo; the image is half the card.

For UK journalists there is also a regulatory floor: IPSO Clause 1 accuracy obligations cover headlines, which must be supported by the text beneath them. A Discover headline that overstates the story is not just an algorithmic risk — it is a complaints risk. Write the headline you could defend to a readers' editor.

Red flags: when Discover reliance becomes dangerous

• ⚠Budgeting or hiring against Discover traffic levels — Google explicitly says Discover traffic is less predictable and dependable than search and should be treated as supplemental.
• ⚠Chasing a Discover spike by mass-producing me-too content on the spiking topic — quality signals, not volume, drive inclusion.
• ⚠Shifting the whole headline style towards curiosity-gap teasers because one performed well — the documented guidance warns against exactly this.
• ⚠Treating a Discover traffic drop as proof of a penalty — volatility is normal for this channel; investigate content quality, but do not assume wrongdoing.
• ⚠Ignoring the rest of the funnel — Discover visitors are typically cold, casual mobile readers; without newsletter, follow, or subscription prompts the traffic evaporates without residue.

Measuring Discover in Search Console

Search Console has a dedicated Performance report for Discover showing impressions, clicks, and clickthrough rate for the last 16 months, with data grouped by page, country, and day. One quirk to know before you go looking: the report is only visible once your property has reached a minimum number of impressions in Discover. If you cannot see a Discover tab in Search Console, your site has not yet had meaningful Discover exposure — that absence is itself a data point.

Read the report differently from the search report. There are no queries, so the useful analysis is page-level: which stories earned Discover distribution, what their images and headlines had in common, and how long each ran. Discover distribution typically arrives in bursts per story rather than as steady daily traffic to a page, so compare stories by total Discover clicks over their run, not by daily averages.

In your analytics package, Discover traffic generally surfaces within Google-sourced traffic rather than as a cleanly separated channel, so Search Console's report is the authoritative view. Google's own help documentation notes that Discover traffic is somewhat less predictable than Search traffic precisely because it is proactively served rather than query-driven — set stakeholder expectations with that sentence before showing anyone a Discover chart.

Common mistakes

• ›Publishing strong journalism with weak, small, or generic images — in an image-led feed this forfeits the channel before quality is even assessed.
• ›Missing max-image-preview:large site-wide — one line in a robots meta tag, and one of the few concrete technical levers Google documents for Discover.
• ›Attributing Discover wins to a secret trick and codifying it — sample sizes per story are small and the feed is personalised; look for patterns across many stories.
• ›Reporting Discover and search referrals as one Google number — they behave differently and mask each other’s trends when blended.
• ›Assuming evergreen content is wasted on Discover — well-maintained explainers resurface when interest spikes; stale ones with outdated facts are a trust liability.
• ›Following pre-2026 guides that recommend Follow-feature RSS optimisation — Google removed that guidance in November 2025 as the feature is no longer shown in Discover.
• ›Letting Discover-era headline habits leak into court, health, or election coverage where precision is a legal and regulatory requirement, not just a style choice.
• ›Never testing article-page conversion for cold mobile readers — Discover traffic converts poorly if the page greets it with an interstitial wall and no reason to stay.

Frequently asked questions