Google’s New AI Search Reality: “Make Content People Want To Read” Isn’t Advice—It’s a Survival Rule
Google’s VP of Search says AI visibility comes down to content people actually want to read—and making it accessible. Here’s what that really means for SMEs, ecommerce, local businesses, publishers, and agencies, plus a practical execution plan and how AYSA helps you monitor, prepare, approve, and implement changes safely.
By Marius Dosinescu (AYSA.ai)
Google’s VP of Search, Liz Reid, recently summarized the new AI Search reality in a way that sounds almost too simple: if you want visibility, publish content people actually want to read—and make sure Google can access it.
That sounds like the same old “Helpful content” advice, but it’s not. In AI-mediated search experiences, “content people want to read” becomes less of a brand slogan and more of an economic rule: if your content isn’t genuinely worth consuming, it won’t earn the behaviors that keep it discoverable (Clicks, saves, shares, brand searches, citations, backlinks, mentions). And if your content isn’t accessible, it won’t even get a chance to be considered.
This editorial is not a news rewrite. It’s a practical resource for SMEs, ecommerce teams, local operators, publishers, and agencies who need to navigate AI search without betting their business on vague guidance.
Concise Summary
- AI search shifts the competition from “who ranks a page” to “who becomes the source the system cites or trusts.”
- Google’s message (via Liz Reid) boils down to two buckets: access (Crawlability/discoverability) and desire (content humans actually value). Source: Search Engine Journal.
- Traffic loss isn’t only AI; audience behavior is shifting toward video and social, which changes what “good publishing” means.
- SMEs should stop chasing “the 1,000th copy” and instead produce proof-based, experience-rich, decision-support content that AI systems can cite and humans will choose.
- Execution is now the bottleneck. AYSA helps by Monitoring, preparing changes, requesting approval, and then implementing accepted fixes so your site can move faster without breaking trust.
Table of Contents
- What Changed: From “Ranking Pages” To “Supplying Answers”
- Why This Matters More For SMEs Than Big Brands
- The Two Buckets Google Highlighted: Access + Desire
- Bucket 1 — Access: Can Google (and AI) Reliably Reach and Understand Your Content?
- Bucket 2 — Desire: Would a Real Person Choose Your Page Over an AI Summary?
- AI Isn’t the Only Threat: The Format Shift (Video, Social, Communities)
- What Content Wins in LLMs and AI Answers (Without Guessing Metrics)
- A Concrete SME Scenario: The Local Clinic That Lost Clicks—And Rebuilt Demand
- The Publisher Reality Check: “Google Zero” Anxiety and What You Can Control
- Agency Reset: New Deliverables for AI Search (AEO/GEO) Without Selling Vapor
- The AYSA Execution Model: Monitor → Prepare → Approve → Implement
- 90-Day Action Plan: What to Fix, What to Build, What to Stop Doing
- What To Do Next
- Sources and Further Reading
What Changed: From “Ranking Pages” To “Supplying Answers”
For most of the modern internet, the implicit deal was:
- You publish pages.
- Google indexes them.
- Users click blue links.
- You monetize the visit (ads, leads, sales, subscriptions).
AI search changes the surface area where that deal happens. A user can ask a question and receive a synthesized response without needing to click—especially for basic, informational queries. The “result” becomes an answer, not a list. Visibility becomes less about ranking position and more about whether your brand, entity, data, and content are selected as supporting evidence.
That’s why Liz Reid’s framing matters. In the SEJ coverage of her interview, she points publishers toward the behaviors that still trigger human interest: unique expertise, freshness, relevance, and content that isn’t just another copy of what’s already out there. Source: Search Engine Journal.
In other words, the bar moved:
- Old bar: Create something indexable that matches keywords and looks credible.
- New bar: Create something that a person would choose even when an AI summary is available—and structure it so systems can cite and trust it.
This doesn’t kill SEO. It compresses the easy middle. The “meh” pages that existed primarily because Search demand existed are the first to be replaced by summaries. The winners are either:
- the fastest, clearest, best source for a factual answer, or
- the brand people trust for decisions, nuance, and consequences.
Why This Matters More For SMEs Than Big Brands
Big brands have margin for error. They have direct traffic, email lists, social followings, PR engines, and offline recognition. Many SMEs don’t. If search changes the rules, SMEs feel it immediately—because search often functions as their primary distribution channel.
But SMEs also have an advantage: proximity to reality.
- You talk to customers every day.
- You know what people misunderstand before they buy.
- You know which edge cases cause refunds, bad reviews, or churn.
- You see the questions that never show up in keyword tools but always show up in sales calls.
AI answers are decent at summarizing what’s already common on the web. They are not inherently great at representing the messy details that differentiate one business from another: the “it depends,” the trade-offs, the local constraints, the real costs, the safety and compliance context, the hands-on experience.
If you publish those details well, you create something AI systems can cite and humans can trust. That’s the game.
The Two Buckets Google Highlighted: Access + Desire
Liz Reid’s answer (as reported by SEJ) essentially describes a two-part filter:
- Access: Can Google reach your content? If you block it, hide it behind scripts, or make discovery hard, you’re opting out.
- Desire: If people do click, will they actually read it—and feel it was worth their time?
This matters because a lot of “AI SEO” advice collapses into gimmicks—schema spam, thin programmatic pages, prompt-chasing, and copycat content. Google’s message is the opposite: don’t build for the machine at the expense of the audience. People will learn over time, and the system will reflect that behavior.
Source: Search Engine Journal.
Bucket 1 — Access: Can Google (and AI) Reliably Reach and Understand Your Content?
“Make content accessible” sounds trivial until you look at typical SME stacks: WordPress plugins, Shopify apps, JavaScript-heavy themes, headless CMS, gated content, duplicated faceted URLs, and years of blog posts that no longer match the current offering.
Accessibility is not only about allowing a bot to crawl. It’s about making it predictable for a system to:
- discover your important pages,
- index the canonical version,
- extract the primary claims and evidence,
- understand the relationships between entities (brand, locations, products, authors), and
- trust that the page won’t change into something else tomorrow.
A practical “Access” checklist for SMEs
Use this as a non-technical triage list. If you’re an owner, you can still spot issues and ask the right questions.
- Indexation sanity: Are your money pages indexed? (Core category pages, service pages, location pages, key guides.)
- Robots & gating: Are you blocking important content via robots.txt, meta noindex, paywalls, or aggressive bot protection?
- Canonical discipline: Do you have one primary URL per topic/product, or do you leak duplicates through parameters, tags, or pagination?
- Internal linking: Can a crawler reach your important pages in a few clicks from the homepage?
- Content rendering: Is critical content present in HTML, or does it require heavy client-side rendering that may delay or complicate indexing?
- Structured data (when appropriate): Are you using schema to clarify what the page is (article, product, local business), without trying to “hack” results?
- Freshness and updates: Do you maintain pages, or publish-and-forget?
Google referenced using webmaster tools and controls for publishers in the interview coverage, which implies the continued importance of Search Console as the operational interface. Source: Search Engine Journal.
Common access mistakes that quietly kill AI visibility
These are the issues that don’t always show up as a “site is down” emergency, but they compound over time:
- Accidental “noindex” on templates after a redesign.
- Overzealous blocking of bots because of scraping fears, without a clear allowlist strategy.
- Orphan pages (good content with no internal links).
- Programmatic bloat creating thousands of near-duplicates (tag pages, internal search pages, faceted filters).
- Thin location pages that look like swapped city names.
- JavaScript-first UX that hides main content behind tabs or “read more” interactions.
In AI search, these errors don’t just reduce classic rankings. They reduce the chance that your content is included in the “candidate set” AI systems might summarize or cite.
Where AYSA helps on access
AYSA is designed as an execution system, not a slide deck. In practice, that means:
- Monitoring detects changes that threaten indexation, crawlability, and key page visibility.
- AYSA prepares specific fixes (for example, internal linking suggestions, metadata/template improvements, content consolidation plans) so you’re not stuck translating “audit findings” into dev tasks.
- It asks for approval before changes go live—critical for regulated industries and brand-sensitive teams.
- Then it implements accepted changes so execution doesn’t stall for weeks.
If you want the conceptual overview first, start with AI search visibility and AYSA’s AI SEO tools.
Bucket 2 — Desire: Would a Real Person Choose Your Page Over an AI Summary?
This is the harder bucket, and it’s where most AI-era strategies fail.
When Liz Reid says “make content people want to read,” she’s implicitly pointing at an uncomfortable truth: search has been flooded with content made primarily to be found, not consumed. AI summaries are the natural response to that flood. If 50 pages say the same thing, the system can compress them into a single answer.
So what does “want to read” mean in business terms?
- It reduces risk. The reader feels safer choosing an option after reading you.
- It saves time. It answers the real question behind the query.
- It gives decision tools. Checklists, comparisons, pitfalls, examples, constraints.
- It shows proof. Experience, process, evidence, and what you did when things went wrong.
- It has a point of view. Not hot takes—clear recommendations with context.
Stop being the “1,000th copy”
Reid’s “not the 1000th copy” point (as reported) is a direct critique of the copycat content machine. If your content strategy is “see what ranks, rewrite it, publish it,” AI will eat that for breakfast—because your page adds no unique information gain.
Ask yourself: if your page disappeared tomorrow, would the internet lose anything?
- If the answer is “no,” it’s probably summarizable.
- If the answer is “yes, because we have unique experience/data/process,” it’s defensible.
What “desirable” content looks like in practice
Desirable content isn’t always long. It’s often complete. Here are patterns that consistently outperform generic posts:
- Decision pages that help customers choose a product/service tier, with real trade-offs.
- Cost pages that explain what drives pricing and how to budget (without fake “$X–$Y” ranges if you can’t support them).
- Comparison pages built around customer intents (A vs B, DIY vs hire, tool vs tool).
- Process transparency: what happens after someone buys/books/calls.
- Edge cases: who it’s not for, when it fails, and what to do then.
- Local nuance: permits, seasonal constraints, neighborhood differences, typical timelines.
- Proof assets: before/after galleries, methodology, testing steps, checklists.
Notice how these aren’t “SEO topics.” They’re business topics that remove friction from buying.
The new metric: “Would you bookmark it?”
A simple heuristic I like for AI-era content quality is: would a smart customer bookmark this?
If not, then your content may still rank temporarily, but it’s not building durable demand signals. And in AI search, demand signals and trust signals increasingly matter.
AI Isn’t the Only Threat: The Format Shift (Video, Social, Communities)
Another key point in the SEJ coverage: Liz Reid emphasized that traffic loss isn’t only about AI. People want different formats—especially video—and often go to social platforms for content. Source: Search Engine Journal.
Even without quoting the referenced Reuters study (SEJ mentions it but the details aren’t included in the provided source text), the direction is consistent with what most SMEs already observe: customers ask questions on TikTok, YouTube, Reddit, Instagram, and in private communities before they trust a website.
This doesn’t mean “be everywhere.” It means:
- Assume your website is not the first touch for many prospects.
- Assume AI answers might become a middle touch that filters options.
- Your site must be the conversion and trust asset that closes the loop.
Build a content stack, not a blog
In the old world, a blog could be a traffic engine on its own. In the AI world, your content stack should include:
- Core pages (services/products/locations) that are precise and trustworthy.
- Authority guides that prove expertise and can be cited.
- Conversion assets (FAQs, comparisons, calculators, downloadable checklists).
- Format companions (short videos, explainers) that match how people learn today.
If you only publish blog posts, you’re leaving money on the table—because you’re producing content that’s easier to summarize than to sell from.
What Content Wins in LLMs and AI Answers (Without Guessing Metrics)
Let’s be careful here: it’s tempting to claim “LLMs prefer X” with fake certainty. We shouldn’t. What we can do is reason from what the system needs to generate credible answers and what humans need to make decisions.
AI answer systems generally perform better when sources provide:
- Clear definitions and consistent terminology.
- Concrete steps (procedural guidance) rather than generic advice.
- Explicit constraints: when a rule applies and when it doesn’t.
- Evidence and attribution: who wrote it, why they’re qualified, what experience informs it.
- Structured sections (so extraction is easier): headings, bullets, tables where appropriate.
And humans respond best when your content includes:
- Examples that match their context.
- Trade-offs rather than “one-size-fits-all.”
- Next steps that reduce decision fatigue.
Information gain is the moat
If I had to pick one concept that describes “content people want to read” in AI search, it’s information gain: you add something the reader can’t easily get from the first AI summary or the first three ranking pages.
Information gain can be:
- experience-based: “what we learned after 200 installs,”
- process-based: “our step-by-step checklist,”
- context-based: “what changes in your state/industry,”
- risk-based: “what goes wrong and how to avoid it,”
- product-based: “what we changed in v2 and why.”
That is difficult for copycats to replicate, which is exactly why it survives.
A Concrete SME Scenario: The Local Clinic That Lost Clicks—And Rebuilt Demand
Here’s a realistic scenario (not a case study with claimed results—just a pattern many local businesses face).
Business: A local clinic offering dermatology and cosmetic procedures.
Problem: They notice fewer clicks from informational queries like “how long does recovery take,” “is it safe,” “what’s the difference between X and Y.” AI answers and rich features satisfy many users on the results page. The clinic worries: “If clicks are down, are we doomed?”
Old approach (common): Publish more generic posts targeting keywords: “What is microneedling?” “Benefits of laser resurfacing,” etc. These become the 1,000th copy.
New approach (aligned with Google’s message):
- Create a decision hub for each procedure: who it’s for, who it’s not for, typical timelines, pre-appointment checklist, aftercare, when to call the clinic, contraindications (written carefully and responsibly).
- Add a recovery timeline section that’s practical: what day 1 looks like, day 3, week 2—plus what changes if you have sensitive skin or a specific lifestyle constraint.
- Publish comparison pages that match patient confusion: “X vs Y: which treats scarring better,” “chemical peel vs laser: downtime trade-offs.”
- Upgrade the service page so it is not just marketing copy. Include process, clinician credentials, what equipment is used (if appropriate), and what to expect.
- Offer printable checklists and “what to ask at your consultation” content—assets people genuinely want.
What changes? Even if informational clicks decline, the clinic becomes the page patients prefer when it’s time to make a decision. Brand searches increase. Direct bookings become more likely. The content becomes cite-worthy because it’s specific and experience-rich—not generic.
This is the shift: from “capture traffic” to “capture trust.”
The Publisher Reality Check: “Google Zero” Anxiety and What You Can Control
The SEJ article flags the fear many publishers have: if AI answers reduce referrals, what happens to the economics of publishing? The piece frames Reid’s advice as potentially ignoring “Google Zero” fears, especially for smaller publishers. Source: Search Engine Journal.
I’m sympathetic to publishers here. But I also think many businesses—publishers included—need to separate what’s unfair from what’s actionable.
What you can’t control:
- How aggressively AI answers satisfy certain query types.
- How many clicks Google sends in aggregate.
- Whether new SERP layouts reduce classic organic real estate.
What you can control:
- Whether your content is the “same as everyone else” or uniquely useful.
- Whether your brand becomes a named destination (subscriptions, newsletters, communities).
- Whether you publish in formats your audience prefers (text + video + short-form explainers).
- Whether your pages are accessible, indexable, and structured.
- Whether you build a moat around expertise and experience.
For publishers and content-heavy sites, the new survival pattern often looks like: fewer commodity articles, more flagship pieces, more community, more direct audience relationships. SEO remains part of that—but it can’t be the only pillar.
Agency Reset: New Deliverables for AI Search (AEO/GEO) Without Selling Vapor
Agencies are under pressure to rebrand SEO as “AEO,” “GEO,” “AI SEO,” and “LLM optimization.” Some of that is legitimate evolution. Some of it is just packaging.
Here’s the non-negotiable reality: AI search raises the standards for both strategy and implementation.
What agency deliverables should look like now
- Visibility monitoring across AI surfaces (not just classic rankings). You need a way to see whether your brand is being mentioned/cited, and for what topics.
- Content portfolios designed for decision support and citations, not just keyword matching.
- Technical accessibility governance (crawlability, canonicals, rendering, internal linking, structured data hygiene).
- Execution throughput: a reliable system to push improvements live without months of backlog.
What not to sell
- Guaranteed “AI Overview placement.” No one can promise that safely.
- Schema as a magic lever. Schema helps clarity; it’s not a cheat code.
- Mass AI-generated content as a substitute for expertise. That’s exactly the “1,000th copy” trap.
Agencies that win will be the ones who can connect: content quality → accessibility → measurable visibility signals → business outcomes, and then execute consistently.
Where AYSA fits for agencies
AYSA can function as the execution layer that makes agency strategy real:
- Agencies can use monitoring to detect regressions and opportunities.
- They can produce content plans and optimization requirements, then rely on AYSA to prepare and implement approved site changes.
- That improves turnaround time and reduces the “we recommended it but the client didn’t implement” problem that kills retainers.
If you’re scoping new AI search services, keep the pricing aligned with reality. (You can reference AYSA pricing to understand how execution tooling might fit into your operating model.)
The AYSA Execution Model: Monitor → Prepare → Approve → Implement
Most teams don’t fail because they don’t know what to do. They fail because they can’t execute consistently.
AI search accelerates change cycles. The cost of slow execution goes up, because:
- new SERP layouts roll out quickly,
- competitors can publish faster than ever, and
- small technical mistakes can quietly erase visibility.
AYSA is built around a governance-friendly execution loop:
1) Monitor
Track the signals that matter: indexation, key page health, changes that could impact discoverability, and visibility trends over time. Start here: AYSA Monitoring
2) Prepare
Turn findings into proposed actions: updates to pages, internal link improvements, content consolidation, structured clarity, on-page refinements—packaged as implementable tasks.
3) Approve
Nothing goes live without human approval. That matters when your brand, compliance, or medical/legal risk is real.
4) Implement
Accepted changes get shipped. This is where most SEO programs die—so this is where we focus.
For a broader view of how we think about AI-era visibility, see AI Search Visibility, and for tooling, see AYSA AI SEO Tools. For ongoing perspectives and playbooks, browse the AYSA blog.
90-Day Action Plan: What to Fix, What to Build, What to Stop Doing
Below is a 90-day plan built for SMEs and lean teams. It assumes you have limited time, limited writers, and limited dev support. Adjust the tempo, but keep the sequence.
Days 1–15: Get “Access” under control
- Inventory your money pages: list the 10–50 pages that directly drive leads/sales.
- Confirm indexation: ensure those pages are indexable and not duplicated.
- Fix internal discoverability: add navigation and internal links so critical pages are not orphaned.
- Reduce duplication: consolidate thin variants and kill tag/filter bloat where it creates near-duplicates.
- Set monitoring: use a system (like AYSA) so regressions get caught quickly.
Days 16–45: Upgrade 5–10 pages to “desire-worthy” assets
Pick a small set of pages that map to high-intent decisions. Then rewrite for humans.
- Replace fluff with clarity: who it’s for, who it’s not for, what it costs (drivers), timeline, risks, alternatives.
- Add comparison sections that reflect real buyer confusion.
- Add checklists, FAQs, and next steps that reduce friction.
- Make the page scannable: strong headings, bullets, short paragraphs, real examples.
Days 46–75: Build your “information gain” moat
- Publish 2–4 flagship guides that only you can write (experience, process, local nuance).
- Create a proof asset: methodology, sourcing standards, editorial policy, author bio depth, or a glossary that standardizes terms for your niche.
- Add format companions: short videos or visual explainers that match how users want to learn.
Days 76–90: Consolidate, prune, and improve conversions
- Consolidate overlapping posts into fewer, stronger resources.
- Remove or noindex true dead weight (thin pages that harm trust and add no value).
- Strengthen conversion paths: CTAs, lead forms, booking flows, product UX, and trust signals.
- Document your system: how topics are chosen, how updates happen, and who approves changes.
What to stop doing immediately
- Stop publishing the obvious. If a competitor’s top page and an AI answer already say the same thing, you need a different angle—or don’t publish.
- Stop measuring only traffic. In AI search, fewer visits can still mean better leads. (You need broader visibility and conversion measurement.)
- Stop treating implementation as optional. Unshipped SEO is fiction.
What To Do Next
- Audit “access” today: confirm your key pages are crawlable, indexable, and internally discoverable.
- Pick 5 pages to upgrade for desire: rewrite for decision support, not keywords.
- Commit to information gain: publish something only your business could credibly produce.
- Set up monitoring and execution: if you want a system that monitors, prepares changes, asks for approval, and implements accepted fixes, explore AYSA:
Sources and Further Reading
- Search Engine Journal — Google Says AI Visibility Hinges On Content People Actually Want To Read (primary source for the interview summary and quotes referenced in this editorial)
- Search Engine Journal — Latest News (context on ongoing Google/AI search changes)
- Search Engine Journal — SEO section (broader SEO coverage and analysis)
Note: The SEJ source mentions a Reuters study about audience format shifts, but that study is not included in the provided research text. I’m intentionally not adding claims or numbers from it here. If you want, we can update this editorial with direct Reuters links and specific findings once you provide them.
Continue the AI search topic inside AYSA.
Use these pages to connect the article with AI SEO tools, AI visibility monitoring, AI Overviews and approved website execution.
Turn this topic into a website action plan.
Use these AYSA hubs to move from reading to technical fixes, AI visibility monitoring, research, glossary context and approval-first SEO execution.
