Does Link Building Still Work? A 2026 Reality Check

by iajyhooda | Aug 17, 2026 | SEO

Google's own search advocate has said the company needs "very few links" to rank a page. This week, a fresh study of 35,000 real ChatGPT citations found that Reddit and Wikipedia together outrank every news publisher on earth as a source AI search trusts. So why are small businesses still being sold "50 backlinks a month" packages like it's 2015?

If you're a business owner paying for SEO, or being pitched on it, this is worth five minutes of your time. It's not a technical question. It's a budget question: where should your money actually go in 2026?

What Google Actually Said About Links

Backlinks have been central to SEO since Google's founding. PageRank, the original algorithm, was built almost entirely on the idea that a link from one page to another was a vote of trust. For two decades, "get more links" was close to the entire industry's advice.

That's been changing for longer than most business owners realise. In April 2024, Google's Gary Illyes, a longtime member of Google's Search Relations team, told a room of SEOs at SERP Conf in Bulgaria: "We need very few links to rank pages… Over the years, we've made links less important." The remark travelled fast enough through the industry that Illyes later half-joked on X that he "shouldn't have said that."

Read that carefully, because the nuance matters. Illyes didn't say Google ignores links. He said Google needs fewer of them, and weighs each one less than it used to. For a business owner, the practical translation is this: quantity-based link building, the kind sold in bulk monthly packages, has been losing value since at least 2024. That's a Google story, not an AI story. AI search has simply made the gap impossible to keep ignoring.

What's Actually Earning Visibility Now

Search Engine Land published an analysis on August 13, 2026 of 35,000 real ChatGPT citations across industries. The finding: user-generated content platforms, mainly Reddit, Wikipedia, and LinkedIn, account for 17.1% of all cited domains. That's more than four times the share held by news publishers, which came in at 4.0%. Wikipedia, Reddit, and LinkedIn alone make up 99% of those UGC citations.

Why does a Reddit thread outweigh a press mention in an AI model's eyes? Large language models are trained heavily on forum and community text, and they tend to treat it as a proxy for authentic, unpaid opinion, the opposite of a press release or a sponsored post. A mention of your business in a niche forum or an industry community can carry real weight with an AI answer engine, even with no link attached at all.

That's not a reason to abandon publisher coverage or PR entirely. It's a reason to stop assuming publisher backlinks are the only thing that counts, when the data says they're currently the smaller share of what AI models actually cite.

This matters more for small and mid-size businesses than it does for national brands. A large brand already has a press archive and a Wikipedia page working in its favour. A local plumber, a regional accounting firm, or a niche ecommerce store rarely does, but that same business often already has something more valuable sitting unused: real customers talking about it in Facebook groups, local subreddits, and review threads. That's the raw material AI models are reading. Most businesses just haven't thought to treat it as an SEO asset.

The "Ghost Citation" Problem

Before you redirect your entire budget toward Reddit threads, one more data point matters. Semrush's research on what it calls "ghost citations" found that 62% of AI citations don't actually produce a visible brand mention in the answer a user reads. In other words: an AI tool can pull facts from your website or a page describing your business, and never once say your business's name to the person asking the question.

This cuts both ways for a business owner. It means citations alone aren't the finish line, being named matters as much as being sourced. And it means any "AI visibility" report that counts citations without checking whether your brand name actually appears in the answer is overstating how visible you really are. If a tool or an agency shows you a citation count and nothing else, ask to see the actual answers.

So Should You Stop Paying for Backlinks?

No, but you should stop paying for the wrong kind, and this is the part most agencies won't tell you, because it's their product.

If your current spend is a fixed number of links per month from anonymous "guest post" networks, private blog networks, or link farms, that's exactly the product Google has been devaluing since 2024. It's also not the kind of mention that shows up in the AI citation studies above. You're paying twice for something that's worth less than it used to be in either system.

What still earns its budget is narrower: a handful of genuinely relevant links from real publications, earned through digital PR or expert commentary rather than bought in bulk, sitting on top of a technically solid website that's actually worth linking to in the first place. If your site itself is the weak link, that's a website development problem before it's a link-building one.

There's also a trust cost to the old playbook that rarely makes it into the sales pitch. Google has spent years penalising sites caught in obvious link schemes, and the same low-quality networks that no longer move rankings much can still trigger a manual action if Google decides they cross the line. That's a real, if less common, downside risk attached to a tactic that's already delivering less upside than it used to. Weigh that against the near-zero downside of building a genuine review and community presence, and the budget decision gets easier.

A Check You Can Run Yourself

This takes five minutes and costs nothing. Open ChatGPT, Gemini, or Google's AI mode and ask the question your customers would actually ask, something like "best [your service] in [your city]" or "who does [what you do] well." See if your business is named. Then, separately, search "[your business name] reddit" and "[your business name] reviews" in a normal search engine. If neither returns anything meaningful, no amount of paid backlinks is going to fix what's actually missing: a genuine footprint in the places AI models are reading from.

The pattern holds up across the client sites we work on: businesses with an active review profile and some real community presence tend to get named in AI answers even when their backlink count is modest. Businesses with a large but generic link profile and no community footprint often don't, however many links point at their homepage.

What to Do With Your SEO Budget Instead

• Keep a small, high-relevance link-building effort aimed at real publications and partners, not volume vendors selling links by the dozen.
• Put real budget into content worth citing on its own merits: original data, direct answers to the questions your customers actually ask, pages that don't need a backlink to be useful.
• Claim and actively manage your presence on the review sites and forums relevant to your industry. This is where AI models are increasingly pulling from, with or without your permission.
• Get a specialist's read on your current link spend rather than guessing. A proper SEO review will show whether your current spend matches what's actually moving the needle for your business today.

FAQ

Is link building dead in 2026?

No. Google's own team has said links matter less than they used to, not that they don't matter at all. What's changed is that link volume from low-quality sources carries little to no benefit, while a smaller number of relevant, earned links still helps.

Do unlinked brand mentions help SEO?

They help AI visibility more directly than they help traditional SEO. Large language models are trained on raw text, not just hyperlinks, so a mention of your brand in a forum thread or review, even without a link, can influence whether an AI tool names or cites you.

How many backlinks does a small business actually need?

There's no fixed number to chase. Google's Gary Illyes has said the company needs "very few links" to rank a page well, so the goal should be relevance and quality over raw quantity.

What is a "ghost citation" in AI search?

It's when an AI tool pulls information from your site as a source but never actually names your brand in the answer the user sees. Semrush's research found this happens in 62% of AI citations, meaning being a source and being visible to the customer are two different things.

Should I fire my link-building agency?

Only if they can't tell you where your links are actually coming from. If the answer is a fixed monthly quota from unnamed sites, that spend is better redirected toward digital PR, reviews, and genuine community presence.

The Takeaway

Links aren't dead. The link-building industry's old pitch, more links, delivered monthly, from wherever, is. Google has been saying so since 2024, and AI search has just made the gap impossible to ignore. If you're not sure whether your current SEO spend matches where visibility is actually coming from now, get in touch and we'll show you exactly what's working and what isn't.

AI Visibility: What to Track Before You Pay for a Tool

by iajyhooda | Aug 12, 2026 | SEO

Over the past year, a new category of software showed up in every marketing tool roundup: the "AI visibility tracker." It promises to tell you whether ChatGPT, Perplexity, or Google's AI Overviews mention your business when someone asks a relevant question. Prices for these tools typically run $200 to $500 a month, and vendors are moving fast — Semrush, Ahrefs, HubSpot, and a wave of newer startups like Profound and Scrunch have all launched competing products in the last twelve months.

If you own a business and someone — an agency, a consultant, a tool's own marketing page — has told you that you need to start tracking your "AI visibility" right now, it's worth pausing before you buy anything. The data on how AI search actually works, published by the same companies selling these tools, tells a more complicated story than the sales pitch does.

What "AI visibility" actually means

"AI visibility" is shorthand for how often, and how accurately, your business shows up when someone asks an AI assistant a question related to what you do. That could mean ChatGPT recommending you by name, Perplexity citing your website as a source, or Google's AI Overviews summarizing your services alongside competitors.

It's tempting to treat this like a new version of search rankings — track your "position," report on it monthly, optimize toward a number. That comparison breaks down quickly. There's no fixed results page. The same question can get a different answer twice in a row. And a business can be recommended by an AI system without its website ever being cited or clicked.

Tom Capper, a search consultant who does regular AMAs on Moz's blog, put it this way in a recent one: “AI search doesn't send traffic the same way traditional search did… A recommendation can still be a win even when your site isn't the one cited.” If someone asks an AI assistant for a plumber recommendation and it names your business while citing a local directory instead of your homepage, that's still visibility — even though nothing in your analytics will show it.

The data behind the sales pitch

Before spending money on a tracking tool, it's worth understanding what the industry's own research says about how AI mentions actually get distributed. Two recent studies are useful here, and both come from companies that sell AI visibility tools — which makes the caution in their own findings notable.

Semrush, working with SEO analyst Kevin Indig, tracked 1,094 US product and service categories in ChatGPT from January through June 2026. The finding: only 15.2% of categories had a clear brand "owner" — meaning one company that reliably appeared across the majority of related questions. For a business to count as a category owner, it had to appear in at least four of five related prompts with a real lead over whoever came second. The other 85% of categories showed no consistent winner. The study also found that only 21% of the most-cited websites in a category belonged to the most-mentioned brand — being mentioned by name and being the source an AI cites are frequently two different companies entirely.

Moz ran a separate experiment, generating 50,000 AI "fan-out" prompts (the follow-up questions an AI system generates behind the scenes to answer a broader query) across 1,000 subtopics. Researcher Dr. Pete Meyers found that even when a brand and its competitors were deliberately fed into the prompt-generation process as hints, only 12.8% of the resulting prompts ended up mentioning any brand at all. Most of the "conversation" an AI system has with itself while researching an answer never touches brand names — it's about the topic, not the company.

Put together, these numbers suggest that for the large majority of categories and queries, there's no stable "ranking" to track yet, and a fair amount of what a $300/month tool would report back to you is noise from month to month rather than a signal you can act on.

What Google Search Console will (and won't) tell you for free

You don't need a third-party subscription to see some of this. In June 2026, Google added generative AI performance reporting directly to Search Console, at no cost, for sites in the rollout. It's worth knowing exactly what it does and doesn't cover before you assume it replaces a paid tool — or before you assume you need a paid tool to get this data at all.

That's a real limitation. You'll know your pages showed up somewhere in Google's AI features, but not what was asked, what was said about you, or whether a competitor was mentioned instead. Google has also confirmed the rollout is still limited to a subset of sites, starting in the UK, so not every Search Console account has access yet.

A DIY check you can run this week

Before paying for a dedicated platform, there's a manual version of the same exercise that costs nothing but half an hour:

1. Write down the 5–10 questions a real customer would type into ChatGPT or Google's AI Mode when they're close to hiring someone in your category — not generic industry questions, but the actual buying-decision ones ("best [service] in [your industry] for [specific need]").
2. Ask each one in ChatGPT, Perplexity, and Google AI Mode, in a fresh/incognito session so past history doesn't skew the answer.
3. Record three things for each answer: were you mentioned, were you cited as a source, and was the description of your business accurate.
4. Repeat monthly. Consistency matters more than volume — the Semrush data above shows that a single appearance means very little; a business only "owns" a topic once it shows up reliably across related questions.

This won't give you impression counts or trend charts, but it answers the question that actually matters for a small or mid-sized business: not "what's my score," but "if a customer asked, would I show up, and would the answer be right."

When a paid tool is actually worth it

None of this means AI visibility tools are a bad purchase for everyone. They start to earn their cost once a few things are true for your business:

• You compete in a category with real commercial stakes per lead (legal, healthcare, financial services, high-ticket home services) where losing a recommendation to a competitor is expensive.
• You operate across enough locations, product lines, or service categories that manually checking prompts monthly isn't realistic.
• You already have a working SEO and content program in place, and AI visibility tracking would be additive intelligence — not a replacement for the fundamentals.

If none of those apply yet, the DIY check above, combined with Search Console's free reporting, covers most of what a small business actually needs to know for now.

What actually moves the needle

Both studies point to the same underlying lever: AI systems build their answers from sources they consider trustworthy on a given topic, and a citation on a well-regarded third-party site does more for visibility than anything you can change on your own homepage alone. That's a case for two things that predate the "AI visibility" trend by years: getting accurate, consistent coverage from other reputable sites in your industry (what's traditionally called digital PR or earned media), and making sure your own site's content clearly and accurately describes what you do, so that when an AI system — or a person — does land on it, the description holds up.

Our SEO services work is built around exactly this: earning the kind of third-party visibility that both AI systems and traditional search reward, rather than chasing a vanity metric that resets every month. If your own site's content is thin or outdated, that's also worth fixing first — our content creation work focuses on making sure what's on your pages actually describes your business accurately, which matters whether the reader is a person or a language model. None of that helps if AI crawlers can't reach or parse your site in the first place, so it's also worth confirming the technical basics are solid — our website development work covers exactly that groundwork.

Takeaway

The AI visibility tool market grew fast, but the research these same companies have published shows the underlying signal is still noisy: most categories don't have a stable "winner," most AI-generated questions never mention a brand, and even Google's own free reporting only shows impressions, not the full picture. For most small and mid-sized businesses, a monthly manual check plus free Search Console data will tell you what you need to know for now. If you want help figuring out whether your business is showing up accurately — in AI answers or regular search — get in touch and we'll walk through it with you.

FAQ

Do I need to buy an AI visibility tool right now?

Not necessarily. Unless you're in a high-stakes category with real per-lead value, or managing visibility across many locations or product lines, a monthly manual check across ChatGPT, Perplexity, and Google AI Mode plus free Search Console data covers most of what a small business needs.

What does Google Search Console's new AI report actually show?

Impressions, which pages appeared, country, and device — for sites in Google's limited June 2026 rollout. It does not show click-through rate, the specific questions asked, or which AI-generated answer your page appeared in.

Can my business be recommended by AI without my website being cited?

Yes. AI systems can name your business while citing a different source, like a directory or review site, for the underlying information. That's still a form of visibility, even though it won't show up in your website analytics.

How often should I check my AI visibility manually?

Monthly, using the same set of buying-decision questions each time. A single appearance in an AI answer means very little on its own — research from Semrush found that reliable, repeated appearance across related questions is what separates a real "topic owner" from a one-off mention.

Does having good SEO automatically improve AI visibility?

It helps, but it's not the same thing. AI systems weigh third-party coverage and how consistently other trusted sites describe your business, alongside your own site's content. A strong SEO and digital PR foundation supports AI visibility, but ranking well in traditional search doesn't guarantee an AI system will recommend you by name.

Does Your Website Need an llms.txt File? Google Says No

by iajyhooda | Aug 7, 2026 | SEO

Over the past few months, llms.txt has become one of those things that shows up in every SEO newsletter, every LinkedIn post, and eventually in a message from a client asking why their site doesn't have one.

The pitch is appealing. Add a single Markdown file to your root directory, and AI systems like ChatGPT, Claude, and Google's AI Overviews will understand your site better and cite you more often. It sounds like robots.txt for the AI era.

There's now enough evidence to answer the question properly. The short version: for the vast majority of businesses, llms.txt is not worth your time. Google has said so directly, and the crawl data backs it up.

What llms.txt Was Supposed to Do

The llms.txt specification proposes a Markdown file at your domain root — yoursite.com/llms.txt — containing a curated summary of your site with links to your most important pages.

The reasoning behind it is sound. Large language models work within a limited context window. HTML pages are full of navigation, scripts, cookie banners, and markup that waste that budget. A clean Markdown summary would, in theory, let an AI system grasp your site efficiently.

It's a genuinely reasonable idea. The problem is that nobody with a major AI crawler agreed to read it.

What Google Actually Said

This is where the case falls apart. Google's AI optimisation guidance, updated in June 2026, states it plainly:

You don't need to create new machine readable files, AI text files, markup, or Markdown to appear in Google Search (including its generative AI capabilities), as Google Search itself doesn't use them.

That's not ambiguous. Google's own AI-features documentation lists llms.txt among tactics you do not need for AI Overviews or AI Mode.

Gary Illyes said at Search Central Live in July 2025 that Google does not support llms.txt and has no plans to. When John Mueller was asked whether Google publishing its own llms.txt file counted as an endorsement, his answer was about as direct as Google gets: "I'm tempted to say something snarky since this has come up so often, but to be direct, no."

Mueller also drew a comparison worth sitting with — he likened llms.txt to the old keywords meta tag: a file where a site declares what it's about, when the search engine would rather just read the site and decide for itself.

Anyone who was doing SEO in 2009 knows how that story ended. Self-reported relevance signals get ignored, because they're trivially gamed.

The Crawl Data Is Worse Than the Statements

Official positions are one thing. Actual bot behaviour is another, and it's more damning. An analysis of 137,000 websites found:

Read that second row again. Nearly every llms.txt file that exists has never been fetched — not by Google, not by OpenAI, not by Anthropic, not by Perplexity.

These aren't files that got crawled and ignored. They were never opened at all. A quarter of the web built a door that nobody has knocked on.

If you want to check this on your own site, don't take anyone's word for it. Open your raw server access logs and search for requests to /llms.txt. On the sites we've looked at, the count is almost always zero. That's a five-minute check, and it's more persuasive than any blog post — including this one.

So Why Does Everyone Keep Talking About It?

Three reasons, and none of them are "it works."

It's easy to sell. "Add this file" is a concrete, cheap deliverable. It looks like progress on an invoice. Real AI visibility work — earning citations, building genuine subject authority, getting mentioned on sites that models actually train on and retrieve from — is slow and much harder to package.

Tools shipped support for it. Yoast added an llms.txt feature. Various generators appeared. Once tooling exists, adoption follows, and adoption gets mistaken for effectiveness. The file being easy to create says nothing about whether it's read.

Correlation gets misread as causation. Sites that publish llms.txt tend to be run by teams who follow SEO closely and do a lot of other things well. When those sites show up in AI answers, the file gets the credit that belongs to the underlying content quality and authority.

What Actually Influences AI Visibility

The uncomfortable answer is that it looks a lot like good SEO — which is roughly what Google has been saying all along.

Ahrefs recently studied 331,000 pages and reached a conclusion worth repeating: Google doesn't punish AI content, it punishes bad content. The same principle applies here. There's no file that substitutes for being genuinely worth citing.

What does move the needle:

1. Content that answers a question completely and can be quoted. AI systems extract passages. Clear, self-contained, factually accurate sections get pulled; rambling ones don't.
2. Being mentioned elsewhere. Models retrieve from and are trained on the wider web. Getting cited on sites that already have authority matters far more than anything in your root directory. This is digital PR and content strategy work, not a technical checkbox.
3. A site that renders fast and clean. If your key content only appears after heavy JavaScript execution, retrieval systems may never see it. This is ordinary technical site quality, and it was already worth fixing.
4. Structured data. Unlike llms.txt, Schema.org markup is documented, actively consumed, and used by Google. If you want a machine-readable file that genuinely gets read, this is the one.
5. Clear, consistent entity information. Who you are, what you do, and where — stated the same way across your site and off it.

None of that is new. That's rather the point. The teams doing well in AI search are largely the teams that were already doing the fundamentals properly.

The Honest Recommendation

Skip it — for now.

If you have a spare ten minutes and a developer already in the codebase, adding an llms.txt file is harmless. It won't hurt your rankings. Google has said it's "fine," which is a long way from saying it's useful.

But if creating and maintaining one would take real hours, displace other work, or appear as a line item on an agency invoice — that's a bad trade. You'd be maintaining a second copy of your site structure that must be kept in sync, in exchange for no demonstrated benefit.

The exception: if you run developer-facing documentation, there's a real case. Tools like Cursor and Claude Code can be pointed at documentation directly, and some documentation platforms have adopted the format meaningfully. If your users are developers who will actively feed your docs to an AI coding assistant, llms.txt may earn its keep. That's a genuine use case — it's just not the one being marketed to most businesses.

Revisit if this changes: if OpenAI, Anthropic, or Google announce official support, the maths changes overnight. It's worth watching. It's not worth pre-emptively building for.

Frequently Asked Questions

Does Google use llms.txt?

No. Google's AI-features documentation, updated June 2026, states that Google Search does not use machine-readable AI text files, and Gary Illyes confirmed there are no plans to support it. Google publishes its own llms.txt file, but John Mueller has explicitly said this is not an endorsement.

Will adding an llms.txt file hurt my SEO?

No. It's a static file that search engines ignore. There's no ranking penalty. The cost is the time spent creating and maintaining it, not any risk to your rankings.

Do ChatGPT or Claude read llms.txt?

There's no confirmed automatic support from any major AI provider. The crawl data is the clearest signal here — in a study of 137,000 sites, 97% of published llms.txt files were never requested by any bot at all.

What should I do instead to show up in AI answers?

Focus on content that's genuinely quotable and complete, earning mentions on authoritative third-party sites, fast clean rendering, and Schema.org structured data. These are documented, actively consumed signals — unlike llms.txt.

Is llms.txt the same as robots.txt?

No. robots.txt is a long-established standard that crawlers genuinely obey and that controls access. llms.txt is a proposed format that suggests content, and major AI crawlers have not adopted it.

The Takeaway

llms.txt is a sensible idea that hasn't been adopted by the systems it was designed for. Google has said it doesn't use it. The crawl logs show almost nobody else does either.

Being cited by AI systems comes from the same place it always has: being genuinely useful, clearly structured, and referenced by other people. There's no file you can upload that shortcuts that.

If you're unsure where your site actually stands in AI search — as opposed to where a checklist says it should — get in touch. We'll look at your real server logs and tell you honestly what's worth fixing.