SEO & AI search insights

SEO has moved beyond your website. Welcome to the discovery ecosystem.

Your website can be fast, well structured and ranking nicely, and a potential customer can still form their opinion of you somewhere else entirely: in a Reddit thread, a LinkedIn post or an AI-generated answer. Here is what has genuinely changed, what has not, and how to build visibility that holds up.

By Sue Chea 23 October 2026 17 min read

Start reading

The short version

Picture two businesses with almost identical websites.

Same quality of content. Same technical health. Similar rankings for the terms that matter. One of them keeps getting named when people ask around, in forums, in professional circles and increasingly in AI-generated answers. The other barely exists outside its own domain.

Ask which business is more discoverable and the answer is obvious. Yet most SEO reporting would score them the same, because it only looks at what happens on and around the website.

That gap is the subject of this article. Not the death of SEO, which has been announced many times and has never arrived, but the widening of the places where discovery actually happens.

In brief

Traditional SEO fundamentals still matter. Google states that no special optimisation is required for AI Overviews or AI Mode beyond them.
What has changed is the journey. People now encounter, evaluate and discuss businesses across communities, professional networks and AI-generated answers, often before (or instead of) visiting a website.
Being mentioned by an AI, being cited as a source, receiving a click and ranking in search are four different outcomes. They need to be measured separately.
The answer is not to post everywhere. It is to strengthen your website as the hub, then build credible, consistent signals in the few places your audience genuinely spends time.

The classic model

What SEO used to mean, and why it worked so well.

For most of the last two decades, SEO had a clear and fairly closed loop. Someone typed a query. A search engine returned a ranked list of pages. The person clicked one. If it was yours, you had a chance to win them over.

Doing well inside that loop came down to a handful of disciplines:

Understanding search intent and the words people actually use.
Building pages that answer those needs better than the alternatives.
Making the site technically accessible so it can be crawled, rendered and indexed.
Earning backlinks as signals that other sites trusted you.
Turning rankings into clicks, and clicks into enquiries or sales.

This model worked because the website really was where the decision happened. The search results page was a doorway, and almost everything worth measuring happened after someone walked through it.

None of these disciplines has become obsolete. Google's own SEO Starter Guide still reads as a sound foundation, and its guidance on helpful, people-first content is arguably more relevant now than when it was written. The fundamentals have not moved. The room around them has grown.

The expansion

What changed: the journey no longer starts, or ends, on your website.

Three shifts have happened at once, and together they change the shape of discovery.

Search itself now answers. Google's AI Overviews and AI Mode synthesise information and present it with supporting links. Google describes how both features may use a "query fan-out" technique, running multiple related searches across subtopics and data sources to build a response. In practice, one question can draw on many pages, from many kinds of sites. The scale is no longer experimental: at its I/O conference in May 2026, Google said AI Mode had passed one billion monthly users, that the average AI Mode query is about three times longer than a traditional search, and that follow-up questions were rising quickly. People are asking fuller, more conversational questions, and expecting a considered answer back.

Community content has become highly visible. Forum threads, Reddit discussions and first-hand reviews now appear prominently in search results, and people actively seek them out because they want unpolished, independent perspectives.

Search is starting to act, not just answer. Google has introduced agentic features in AI Mode that can find live availability for things like restaurant tables, event tickets and some local appointments, then hand people off to a booking partner to finish. These are still limited by country and account type, but the direction is clear.

A realistic journey Hypothetical example

Imagine Priya, who runs a busy café in Sydney's inner west and needs a new bookkeeper. Her path might look like this:

She notices a LinkedIn post from a fellow hospitality owner describing how their bookkeeper untangled a messy payroll setup.
She asks a search engine a long, specific question about bookkeeping for small cafés with casual staff, and reads an AI-generated summary that points to a couple of sources.
She searches the name of one firm she saw mentioned and reads a Reddit thread and several Google reviews.
She finally visits two websites, checks pricing and credentials, and books a call with one of them.

The website did the converting. But by the time Priya arrived, most of her opinion had already formed elsewhere. A firm with a better website but no presence in steps one to three might never have made her shortlist.

Journeys like this are not universal or linear. Plenty of people still type a query, click a result and buy. The point is that the alternative paths now carry real weight.

The new discovery ecosystem

Eight connected surfaces where people find, check and trust you.

Your website

Still the owned foundation. It is where your full expertise, services, pricing, credentials and contact paths live, and it is the one channel whose rules you control. Every other surface ideally points back here.

User-generated content

Reviews, testimonials, customer questions and first-hand experiences. People trust what other customers say, and this content often answers the specific, awkward questions your marketing copy avoids.

Reddit and niche communities

Context-rich discussions where people compare options candidly. These threads frequently appear in search results, which means a conversation about your category can outrank your own page for the same question.

More than an engagement feed. It is where professional expertise becomes visible and attributable to real people, which makes it a meaningful reputation layer for consultants, B2B firms and anyone selling knowledge.

AI search and generated answers

AI Overviews, AI Mode and standalone assistants summarise information and may link to sources. Your content can inform an answer, be cited as a source, or neither, depending on how useful and accessible it is.

Brand mentions and reputation

What independent sources say about you, and whether it is consistent. Clear, matching information about who you are and what you do helps people (and systems that read the web) understand you accurately.

Autonomous agents

An emerging layer. Today's agentic search features mostly find options and hand off to a partner site. More capable research and comparison agents are developing, but how they will choose between businesses is still largely unknown.

Traditional search results

Still a major discovery channel. Crawlability, relevance, useful content and sound site structure remain the price of entry, and they also underpin eligibility for Google's AI features.

How it fits together

These are not separate channels. They feed each other.

The useful mental model is not a list of platforms to manage but a web of reinforcing signals. A thoughtful LinkedIn post can prompt a branded search. A helpful answer in a community thread can earn a mention when the same question comes up again. A detailed guide on your website gives community members, journalists and AI systems something specific and credible to reference.

The connections also run in the other direction. If your website says one thing, your Google Business Profile says another and customers describe a third experience in reviews, that inconsistency travels too.

It is also why it pays to own your content rather than renting visibility on social platforms, while still taking part in them. This is why strategic branding and SEO are converging. The question is no longer just "does our page rank?" but "is the story about us clear, consistent and credible wherever someone encounters it?"

Old thinking, new thinking

The assumptions worth reconsidering.

Ten assumptions still shape how many businesses think about SEO. Most deserve a second look, though not all of them deserve to be thrown out. Here is where I land on each, based on what we know in 2026.

Swipe sideways on mobile to see the full table.
Old assumption	What is still true	What has changed
Only websites matter	Your website is the only discovery asset you fully own and control.	Opinions are increasingly formed before the visit, on surfaces you do not own.
Rank higher = more visibility	Higher positions still tend to earn more attention.	When an AI summary sits above the results, fewer people click through. A 2026 randomised field experiment by university researchers, following 1,065 desktop Chrome users, found AI Overviews reduced clicks to organic results by about 40% when they appeared. The paper is still a working draft awaiting peer review, but it is the strongest causal evidence so far, and it is consistent with earlier observational studies.
Backlinks drive everything	Links remain a meaningful signal of trust and a way people navigate the web.	Unlinked mentions, reviews and discussions also shape how people judge you, even when they pass no link value.
Social is separate from SEO	Google has not said that social engagement is a ranking factor.	Social conversations build awareness and branded demand, and some public posts are themselves surfaced in search and AI answers.
Reddit is just a forum	It is a community first, and treating it as an ad channel backfires.	It is also highly visible in search. In one March 2026 analysis of 8,566 B2B software keywords where Reddit competes with vendors, Reddit held a top-three spot for roughly 40–45% of queries in three of four verticals, but only about 10.5% in the fourth (unified communications and contact-centre software). Highly visible, but far from universal, so check your own category rather than assuming.
LinkedIn is just engagement	Engagement metrics alone say little about business impact.	Its public content can be cited by AI systems. A 2026 analysis of 9.5 million AI citations across six AI platforms, including Google AI Overviews and AI Mode, ranked LinkedIn second behind YouTube. That still amounted to about 0.5% of all citations, and roughly three-quarters of LinkedIn citations pointed to individual member profiles rather than company pages. Text posts made up most of the cited LinkedIn content, well ahead of long-form articles and video.
AI only uses websites	AI features in Google Search link to web pages that are indexed and eligible to show a snippet.	The "websites" AI draws on now include forums, review platforms, video and professional networks. In the 2026 citation analysis above, YouTube and LinkedIn were the two most-cited domains across the AI platforms studied, and community platforms such as Reddit feature heavily in AI answers too.
Traffic = success	Qualified visits still matter enormously.	Some visibility now happens without a click. Better questions: are the right people finding us, and are they acting?
Owned content is enough	Original content on your own site is the foundation.	Content that is never discussed, referenced or found elsewhere has a narrower reach than it once did.
AI search needs secret hacks	Nothing magical is required.	Google states there are no additional requirements and no special schema needed for AI Overviews or AI Mode. Useful, accessible, accurate content is the strategy.

One more nuance worth naming: the infographic frames these as a ladder from old to new. In reality they are layered. The "old" foundations sit underneath the new surfaces and hold them up.

Getting the language right

Mentioned, cited, visited, ranked: four outcomes that are not the same.

A lot of confusion about "AI visibility" comes from lumping different results together. They behave differently and should be tracked differently.

Mentioned

The AI names you

Your brand appears in a generated answer. Good for awareness, but there may be no link and nothing to measure on your site.

Cited

The AI links to you

Your page is shown as a supporting source. This builds credibility, though clicks on these links can be low.

Visited

Someone clicks through

A referral visit reaches your site. This is where you can see behaviour, measure enquiries and convert.

Ranked

You appear in the results

A conventional position in search results. Still valuable, but no longer a full proxy for how visible you are.

You can be mentioned without being cited, cited without being clicked, and ranked without appearing in an AI answer at all. A healthy strategy aims for a mix, and a good report shows which ones are actually moving.

A practical framework

The Hub & Signal framework.

Think of your website as the hub: the authoritative, complete version of who you are. Everything else is a signal: evidence elsewhere on the web that you are credible, relevant and worth choosing. Strong hubs with no signals go unnoticed. Strong signals with a weak hub leak trust at the final step.

The framework has three phases and eight steps.

Phase 1: Build the hub

1. Establish a technically sound, genuinely useful website

What to do: Fix crawl and indexing issues, make pages fast and accessible, and ensure every core service has a clear, specific page. An SEO audit is the quickest way to find what is holding a site back.
Why it matters: To appear as a supporting link in AI Overviews or AI Mode, Google says a page must be indexed and eligible to show a snippet. No hub, no citation.
Example: A physio clinic replaces one vague "Services" page with individual pages for each treatment, each answering who it suits, what happens and what it costs.
Measure: Indexed pages, Core Web Vitals and page experience issues in Search Console, and errors resolved.

2. Make important content accessible to people and search engines

What to do: Put key information in crawlable text, not only images or PDFs. Use clear headings, descriptive alt text and structured data that accurately reflects visible content.
Why it matters: Content that cannot be read cannot be summarised, quoted or recommended, by a person or a system.
Example: A builder moves pricing guidance from a downloadable brochure into an HTML page with a clear table.
Measure: Rich result eligibility, snippet appearance and impressions for previously invisible pages.

3. Connect related content with thoughtful internal links

What to do: Group articles into topic clusters and link between them with descriptive anchor text, not "click here".
Why it matters: Google's AI features guidance specifically lists making content easily findable through internal links as a worthwhile fundamental.
Example: This article links to related pieces on strategic branding and our SEO audit service, so readers and crawlers can follow the thread.
Measure: Orphan pages reduced, and pages per session within a topic cluster.

Phase 2: Earn the signals

4. Identify the platforms and communities your audience genuinely uses

What to do: Ask recent customers where they looked before choosing you. Search your category plus "reddit", "review" or "forum" and see what surfaces.
Why it matters: Effort spread across ten platforms is usually wasted. Two well-chosen ones can do more.
Example: A B2B software consultancy discovers its buyers discuss tools in two specific subreddits and an industry Slack group, not on Instagram.
Measure: A short, evidence-based shortlist of priority platforms, revisited quarterly.

5. Publish original expertise that answers meaningful questions

What to do: Write from first-hand experience: real processes, real numbers you are able to share, real lessons. Put the direct answer near the top, then the depth.
Why it matters: Generic content is easy to summarise and replace. Distinctive insight is what people quote, share and cite.
Example: An accountant publishes a plain-English breakdown of a common BAS mistake they see every quarter, then shares a shorter version as a LinkedIn post.
Measure: Impressions for question-style queries, saves and shares, and mentions of the piece elsewhere.

6. Participate constructively in communities without spamming

What to do: Answer questions where you have real expertise, disclose your affiliation, follow each community's rules and link only when it truly helps.
Why it matters: Communities can tell the difference between help and promotion, and they remember. Manipulative tactics also risk breaching Google's spam policies and platform rules.
Example: A web designer spends twenty minutes a week answering hosting questions in a small-business forum, rarely mentioning their own services.
Measure: Helpful replies posted, unprompted mentions of your brand, and referral visits from those communities.

7. Build consistent brand information and a credible reputation

What to do: Align your name, description, services, location and contact details across your website, Google Business Profile, LinkedIn and directories. Ask happy customers for honest reviews and respond to all of them.
Why it matters: Consistency makes you easier to understand and trust. Contradictions create doubt for people and ambiguity for any system summarising the web.
Example: A café discovers three different opening hours online and fixes them in one afternoon.
Measure: Review volume, rating and response rate, plus the share of key listings that are accurate.

Phase 3: Read the ecosystem

8. Measure outcomes across channels and adjust based on evidence

What to do: Set a baseline, choose a balanced set of metrics (see below) and review them monthly. Drop what is not working; double down on what is.
Why it matters: Without measurement, ecosystem thinking turns into busywork across too many platforms.
Example: A consultant notices most new enquiries mention LinkedIn, so shifts time from a low-performing blog series into fewer, deeper articles shared there.
Measure: Enquiries and conversions by source, trends over three to six months.

Measure what matters

Measurement beyond rankings and clicks.

When discovery happens across many surfaces, no single number tells the story. Track a balanced set across four layers.

Search visibility

Are you showing up?

Impressions and clicks: Search Console's Performance report, which Google confirms includes traffic from AI features.

Generative AI impressions: In June 2026 Google launched dedicated generative AI performance reports in Search Console showing how often your URLs appear in AI Overviews and AI Mode. Google says they reached all websites worldwide by 31 August 2026, though sites with few AI impressions may not see data. They show impressions only, not clicks.

Your AI visibility setting: Search Console now also includes a control that lets site owners opt out of AI Overviews and AI Mode. Google says opting out does not affect regular search rankings, but your pages will no longer appear or earn traffic in those features. For most businesses that want to be discovered, leaving it on is the sensible default.

Brand demand

Are more people looking for you specifically?

Branded search volume: growth in searches for your name is one of the clearest signs that visibility elsewhere is working.

Direct visits: people who arrive by typing your address.

Ecosystem presence

Are you part of the conversation?

Mentions and citations: periodically ask AI assistants the questions your customers ask, and record whether and how you appear. Treat it as a spot check, since answers vary.

Referral traffic: visits from communities, LinkedIn and AI platforms in your analytics.

Review profile: volume, rating and recency.

Business outcomes

Is it turning into revenue?

Qualified enquiries and conversions: the numbers that actually pay the bills.

"How did you hear about us?": a simple form field that captures journeys analytics cannot see, such as a Reddit thread read on a phone weeks earlier.

A note on expectations

Fewer clicks are not automatically a failure. If impressions, branded searches and enquiries are rising while clicks hold steady or dip, more of your discovery may simply be happening before the visit. Read the metrics together, compare against a baseline and look for trends over months, not days.

What this means for you

The same shift lands differently depending on who you are.

Small and local businesses

Your biggest levers are an accurate Google Business Profile, a steady stream of genuine reviews and a website that answers the questions customers ask on the phone. Pick one community and show up there consistently.

Consultants and creators

Your expertise is attached to your name, so personal platforms like LinkedIn carry unusual weight. Publish the deep version on your site and the conversational version where your peers are.

B2B and professional services

Buyers research quietly and in groups. Expect them to look for independent opinions before talking to you, and make sure there is something credible for them to find.

Established brands

You already have mentions. The work is consistency: making sure what the web says about you is current, accurate and aligned with how you want to be understood.

Looking ahead

AI agents: what exists today, and what is still a prediction.

What exists now. Google has introduced agentic capabilities in AI Mode that can search across booking platforms for restaurant availability and, in some markets and account tiers, event tickets and beauty or wellness appointments. These features largely surface options and send people to a partner site to complete the booking. Restaurant booking began in the US and in April 2026 expanded to more markets, including the UK, with a local rollout reported in Australia around the same time. Ticket and appointment booking have been more restricted, so check what is available to your customers before planning around it.

What is reasonable to expect. Agents will likely take on more of the research and comparison work: shortlisting suppliers, checking availability, summarising reviews. Businesses with clear, structured, accurate information will be easier for these systems to work with.

What nobody knows yet. How agents will weigh one business against another, how much they will rely on reviews versus official sites, and whether they will send meaningful traffic or complete tasks without a visit. Anyone selling certainty here is guessing.

The sensible response is the same as everything else in this article: keep your information accurate and accessible, make it easy to transact with you, and build a reputation that holds up when someone, or something, checks.

Common questions

Quick answers.

Is traditional SEO dead?

No. Google's own documentation says SEO best practices remain relevant for its AI features. What has changed is that SEO alone no longer describes your whole visibility.

Do I need to optimise separately for AI Overviews and AI Mode?

According to Google, no special optimisation or schema is required. Pages need to be indexed and eligible to show a snippet, and the usual fundamentals apply.

Are Reddit or LinkedIn ranking factors for my website?

Google has not said that activity on either platform boosts your website's rankings, and there is no reliable evidence that it does. Their value is that they are discovery surfaces in their own right, and that they shape reputation and branded demand.

Should my business be on every platform?

No. Choose the one or two places your customers genuinely use and do them well. Consistency beats coverage.

How do I know whether I appear in AI answers?

Check the generative AI performance report in Google Search Console, available to all sites since August 2026, and periodically test the questions your customers ask in AI tools. Record the results so you can see trends.

Can I stop my site appearing in AI Overviews and AI Mode?

Yes. Since mid-2026, Search Console has included a setting that opts your site out of Google's generative AI search features. Google says it does not affect your regular rankings, but you also give up any visibility and traffic from those features, so think carefully before switching it off.

Looking for more practical guides? Browse the full Insightful Sue resources library.