Are You Ready for This Week’s SEO News—“Read More” Deep Links, Robots.txt Changes, and EU Search Data for AI?

If your SEO to-do list already looks like a tab hoarder’s browser… same. And just when you think you can finally close a few tabs, Google drops a fresh batch of “small changes” that somehow turn into big traffic swings. This week: “Read more” deep links that can make (or break) your snippets, robots.txt rules getting the spotlight (and maybe a spellcheck), and the EU asking Google to share search data that could feed rival search engines and AI tools. Let’s turn the news into a simple, do-this-next checklist—without the headache.

“Read more” Deep Links: Stop Making Google Play Hide-and-Seek With Your Best Content

You know that moment when you click a result expecting the answer… and the page drops you at the top with a giant hero image, a cookie banner, and a “tap to expand” accordion that’s guarding the one sentence you came for? Google’s trying to reduce that pain.

Google updated its Search snippet documentation with specific best practices for “Read more” deep links—those extra links in search results that jump people straight to a section of your page. The headline rule is almost insulting in how simple it is: your important content needs to be immediately visible on page load. If the key line is tucked behind tabs, accordions, “click to expand,” or content that only appears after scrolling/interaction, you may be lowering your chances of earning those deep links in the snippet .

Google also calls out structure: use H2/H3 headings to define sections that are easy to jump into . Translation: if your page is one endless wall of text with random bolding, you’re making it harder for Google to confidently send searchers to the exact right spot.

What’s changing (in plain English)

Google is finally putting “Read more” deep links into the “here are the rules” bucket. The guidance reads like a general preference, too: content that renders without user interaction is more likely to show up in enhanced Search displays . So this isn’t just about snippets; it’s about how Google (and anything acting like a crawler/agent) consumes your page.

Quick page update checklist (steal this for your next content refresh)

1. Put the money lines above the fold (and outside UI traps).If your best snippet-worthy answer is inside an accordion or tabbed “Details/Specs/FAQ” area, consider rendering the core 1–3 sentences open by default .
   

2. Write section headings like actual signposts.Use clear, keyword-aligned H2/H3s (think: “Refund policy,” “Shipping times,” “How long does X take,” “Pricing,” “Ingredients”) so Google can map a deep link to a real chunk of intent.
   

3. Make the snippet text match the on-page copy.Google’s guidance is blunt here: the snippet text should match what appears on the page . If your meta description promises “Free returns for 60 days” but your on-page line says “Returns accepted case-by-case,” you’re asking for disappointment (and pogo-sticking).
   

4. Stop relying on “loads after scroll” for core answers.Infinite scroll sections, lazy-loaded FAQs, and “only appears when you interact” content can reduce the likelihood of deep links appearing . Keep the essentials in the initial HTML/render.
   

5. Audit your jump destination like a reader, not an SEO.If a deep link lands someone in the right section, they should instantly see:
   

   • the heading they expected,

   • the exact answer line, and

   • any needed context (one short sentence is usually enough).

     No bait-and-switch. No “the answer is in another accordion.”

If you’ve ever complained that Google “doesn’t send traffic like it used to,” this is one of those small, annoying page-structure things that quietly decides whether you get the click—or you get skipped.

Robots.txt Audit Time: Clean Up the Weird Stuff Before Google Calls It Out

If the “Read more” update was Google saying “stop hiding the good stuff,” this one is Google saying, “stop putting weird notes in the margin and hoping we’ll interpret them.”

Google’s been looking at real-world robots.txt files (at scale) and is preparing to expand documentation around the most common unsupported directives people keep using anyway.

This came up via Gary Illyes and Martin Splitt, who said they analyzed robots.txt rules seen across millions of URLs and plan to document the top 10–15 most-used unsupported rules, beyond the basics Google actually supports (User-agent, Allow, Disallow, Sitemap) .

There’s also a spicy little aside: Google might expand which typos it accepts for Disallow in its robots.txt parser (no timeline, no promise) . Helpful? Maybe. But you really don’t want your crawl strategy depending on Google being “typo tolerant” this week and “lol nope” next week.

What this means in plain English

• If your robots.txt has “creative” directives, Google may already be ignoring them—and now it’s getting more official about calling that out .

• If your robots.txt has sloppy spelling or copy-paste leftovers, you’re playing roulette with crawling and indexing. Google hinted the parser already accepts some misspellings of disallow, and might accept more later . That’s not a plan. That’s a prayer.

Fast robots.txt audit checklist (15–30 minutes, max)

1) Make your robots.txt boring again

• Keep it to what Google supports: User-agent, Allow, Disallow, Sitemap .

• Delete anything that looks like a “rule” from a tool/plugin era you don’t even remember installing.

2) Search for “unsupported directive” red flags

Common offenders SEOs still bump into:

• noindex (robots.txt can’t reliably “noindex” in modern Google-land)

• crawl-delay (popular, but not a Google control knob)

• random vendor-specific directives added by old SEO plugins

Even if other bots read these, Google documenting unsupported rules is your cue to stop assuming Google cares .

3) Check you’re not blocking what makes pages work

• Don’t block CSS/JS directories if they’re needed to render templates properly.

• Don’t block important page types (product pages, category pages, blog posts) because of one overly broad Disallow: / moment.

4) Re-check every User-agent block like it’s code review

• Is Googlebot getting the rules you think it’s getting?

• Any conflicting Allow/Disallow paths?

• Any copy-paste sections that block staging paths… but also block production?

5) Fix typos now (don’t wait for “tolerance”)

• Normalize spelling and capitalization.

• Correct anything that’s a misspelling of Disallow (or any directive at all). Google hinted typo handling could expand, but that’s not something to bank on .

6) Test it in a real robots.txt tester

• Don’t “looks good to me” this.

• Test representative URLs from key templates (homepage, category, product/service, blog article, internal search, faceted pages).

Because when robots.txt goes wrong, it doesn’t go wrong politely. It goes wrong like a light switch.

EU Search Data for Rivals (and AI): What Happens When “The Recipe” Gets Shared?

So you’ve cleaned up your robots.txt and stopped accidentally whisper-blocking your own site. Cool. Now zoom out, because the EU is basically asking Google to share the signals that make Search… Search.

The European Commission sent preliminary findings that would require Google to share Search data with rival search engines across the EU/EEA—and that eligibility could extend to AI chatbots if they qualify as “online search engines” under the DMA . It’s not final yet (consultation open until May 1, final decision due July 27) , but the direction is clear: discovery might stop being “Google-only” data.

What data are we talking about?

The proposal calls out four categories of data that could be shared on fair terms:

• Ranking data

• Query data

• Click data

• View data

Put in human terms: it’s like handing someone the kitchen notes for what people order, what they looked at, what they actually ate, and what they came back for.

Why SEOs (and publishers) should care

If rival search engines—or AI search interfaces—can access anonymized versions of these signals, they can get better at answering the same questions users type into Google. That can change where people discover you, and also how quickly competitors learn what tends to win a click .

Even if you don’t operate in the EU, the more interesting piece is the definition shift: the proposal explicitly pulls AI chatbots into the “search engine” bucket (if they meet the DMA definition) . That’s a signal about how “search” is being framed going forward.

Visibility survival plan (the practical stuff)

Track demand like a hawk (not vibes)

1. Brand demand: set up a weekly view of branded queries and brand+category combos (in GSC and whatever rank tracker you trust).

2. Category demand: monitor non-branded, early-funnel queries that introduce your product/service (“what is…”, “how does…”, “best…”, “alternatives to…”).

If discovery spreads across more interfaces, you want to know whether interest is dropping… or just moving.

Fortify the pages that already win

• Re-check the intent match on top landing pages: is the first screen actually answering the query, or warming up with filler?

• Tighten “last-click” pages (pricing, product, demo, booking, store locator) so they’re useful even if users arrive half-convinced from an AI summary.

Build early-funnel content that’s hard to “copy-paste”

AI tools can repeat facts. They struggle more when the answer needs:

• clear steps,

• specific constraints (“if X, do Y; if not, do Z”),

• definitions + edge cases,

• up-to-date details that are easy to verify on-page.

Make your best answers structured and checkable. If an interface is going to summarize you, you want it summarizing something you actually control.

Watch for traffic shape-shifts, not just traffic drops

If new EU experiences pop up, you may see:

• fewer clicks on simple informational queries,

• steadier clicks on pages with tools, calculators, templates, or “do the thing” actions,

• more volatility in mid-funnel “best X for Y” queries as competitors learn faster what earns attention.

None of this is about panic. It’s about realizing the “SEO game” might be played on more boards at once—and you’ll want your analytics set up to spot that early.

Task-Based Search Is Getting Real: Price-Drop Alerts + AI Agents (and Your Clicks on the Line)

All the EU data-sharing talk is about who gets the signals. This part is about where the action happens.

Google’s rolling out more task-based Search features—stuff that used to start on your site, but now starts (and sometimes finishes) inside Google. Two moves stood out this week: hotel price-drop tracking and the ability to launch AI agents from AI Mode .

What Google launched (and what it means for clicks)

1) Hotel price-drop tracking (now at the individual hotel level)

• There’s a toggle in the Search bar that lets users track price drops .

• It’s available globally .

• When prices drop, Google can send an email alert .

The uncomfortable part: price-drop alerts are a classic “come back and book” trigger that hotels and OTAs love. If Google owns that reminder loop, some of those return visits won’t come back to your site the way they used to .

2) AI agents launched from AI Mode

Google is also adding the ability to kick off AI-handled tasks directly from AI Mode—without leaving Search .

Google hasn’t published deep details yet on what tasks are supported or how sources get cited, so it’s a “watch this space” feature. Still, the direction is clear: Search is moving from “find answers” to “get stuff done” .

Update your SEO audit playbook (so you don’t lose the good clicks)

When Google completes more tasks on-SERP, your goal shifts: win the clicks that still matter.

Prioritize “last-click” pages

Focus QA and improvements on pages where a click has real value:

• booking / checkout

• signup / trial

• demo request / contact sales

• store visit / appointment

If these pages are slow, vague, or cluttered, task-based SERPs won’t be kind.

Put the next step above the fold

Make it painfully obvious what a visitor should do in the first screen:

• a clear primary CTA (one, not five)

• price, availability, or key qualifiers (so users don’t bounce back to the SERP)

• trust basics (policies, reviews, guarantees—whatever matters in your category)

Format for “skimmable wins”

Task-heavy SERPs punish fluffy pages. Tighten:

• short sections with descriptive headings

• bulleted lists for requirements, pricing inclusions, and steps

• quick “who this is for” / “who this isn’t for” blocks

Monitor query groups where Google starts doing the job

Pick a few clusters and track them weekly in GSC:

• “price” / “deals” / “discount” queries (especially travel + retail)

• “best” / “top” queries (comparison-heavy)

• “how to” queries (step-by-step answers that can be summarized)

If impressions hold steady but clicks slide, don’t just blame “seasonality.” It might be Google’s new default behavior.