I used to walk into client meetings with a mental (and sometimes physical) checklist of Google ranking factors. Title tags, backlinks, page speed, mobile-friendliness – the works. I’d explain which factors we were optimising for and which ones we’d ticked off. It felt structured and credible, but I think it was holding everyone back.
For some time, I’ve stopped using the phrase “ranking factors” with clients, in strategy documents, and in my own thinking about SEO. Here’s why that shift has made me a better consultant, and why I think it matters for anyone making decisions about organic visibility in 2026.
Google doesn’t think in ranking factors – and hasn’t for years
The SEO industry loves ranking factors. Entire businesses have been built on “the complete list of 200+ Google ranking factors.” The problem is that Google’s own people keep telling us the model is wrong.
Google’s Search Advocate, John Mueller said back in 2021:
“It was a bad idea for us to frame this (or anything) as a part of a ranking of ranking factors.”
He went further:
“Where does ranking even start? Understanding a query is massively complex, and determines what will be ranked, but is it a ranking factor? How does treating it as a ranking factor help a site-owner make decisions?”
On Google’s Search Off The Record podcast, Mueller described his preferred mental model as something closer to a neural network – lots of interconnected signals passing through a complex system where any individual signal might be decisive for one query and irrelevant for the next.
“It’s not the case that any particular factor within this big network is the one deciding factor, or that you can say that this factor plays a 10% role.”
So why are ranking signals still revered in 2026?
Look at Google’s own documentation and you’ll find a careful vocabulary. They talk about ranking systems (the algorithms – BERT, RankBrain, the helpful content system) and signals (the inputs those systems use). Their ranking systems guide opens with: “Google uses automated ranking systems that look at many factors and signals.” Their page experience documentation states it plainly: “There is no single signal.” This is language designed for nuance.
The E-E-A-T problem shows exactly why this matters
If you want a case study in how “ranking factor” thinking goes wrong, E-E-A-T is exhibit A.
Google’s Search Quality Evaluator Guidelines describe Experience, Expertise, Authoritativeness, and Trustworthiness as the framework human raters use to evaluate whether search results are any good. Thousands of quality raters around the world use E-E-A-T criteria to assess search results and feed that assessment back to Google, which uses it to refine its algorithms. This is not the same thing as a ranking signal being plugged directly into an algorithm.
Google has been uncharacteristically blunt about this. Their own documentation says: “While E-E-A-T itself isn’t a specific ranking factor, using a mix of factors that can identify content with good E-E-A-T is useful.” Danny Sullivan, Google’s Search Liaison, got more emphatic in February 2024:
“It’s not a ranking factor. Having an expert write things doesn’t magically make you rank better, because anyone could self-declare someone to be an expert, and that means nothing.”
And yet. E-E-A-T clearly influences visibility. Sites that genuinely demonstrate expertise and trustworthiness earn more links, generate longer visits, attract more return users, and get mentioned more frequently across the web. Those behavioural and authority signals absolutely feed into Google’s ranking systems. The causation flows from genuine quality through to measurable signals through to rankings, but you can’t reverse-engineer that by adding an author bio and calling it “E-E-A-T optimisation.”
Lily Ray, one of the sharpest voices on this topic, nailed the nuance:
“A lot of times SEOs get hung up on ‘Google never said that’s a ranking factor.’ Yeah, they’re way beyond that. There’s a reason they haven’t told us, because we’re so simplified in the way we think about SEO.”
She’s right. The ranking factor framing oversimplifies and, frankly, actively misleads.
The meeting where “ranking factors” thinking costs you money
Here’s where this gets practical. I’ve sat in meetings where a well-meaning CMO has asked: “But is accessibility a confirmed ranking factor?” And when the honest answer is no, the budget for accessibility improvements evaporates. Never mind that a Semrush-backed study of 10,000 websites found organic traffic increased by an average of 23% as accessibility compliance improved. Never mind the legal exposure, the brand perception, or the growing user base who simply can’t use an inaccessible site.
The same pattern plays out with structured data, brand building, content depth, and digital PR. The moment someone frames the question as “is this a confirmed ranking factor?” they’ve created a binary gate that kills good strategy. Invest or don’t invest. Yes or no. That mental model turns every nuanced, context-dependent conversation about search into a crude pass/fail test.
AI search has made the whole model obsolete
Whatever residual usefulness the ranking factors model had, generative AI search has finished it off.
Google AI Overviews now appear in 57% of search results. ChatGPT serves 900 million users weekly. Perplexity, Claude, Gemini… these systems don’t rank pages on a list. They retrieve information from across the web, evaluate it for trustworthiness and relevance, then synthesise an answer. Your content either gets cited or it doesn’t. There is no position 3.
The data is stark. Research from Moz found that only about 10% of AI Mode citations match Google’s traditional organic results. Ahrefs found just 12% of URLs cited by large language models rank in Google’s top 10 for the original query. The old playbook of optimise for ranking factors, climb the SERPs, has almost no predictive power for whether your brand shows up in an AI-generated response.
As Mike King, CEO of iPullRank, put it:
“If every search result is personalised in real time based on a user’s entire digital history, there is no ‘Position 1’ anymore – there is only intent and relevance.”
The Princeton GEO research paper found that generative engine optimisation strategies, adding statistics, citing credible sources, improving content depth, can boost visibility in AI responses by up to 40%. Traditional SEO metrics like backlink counts? Minimal impact. The signals that matter in AI search are about whether your content is comprehensive, trustworthy, well-sourced, and genuinely useful. You can’t reduce that to a checklist.
What to say instead
I’ve replaced “ranking factors” with a simpler question when I’m talking to clients: what does a good experience look like for the person searching? That question leads to better content, better technical foundations, better brand building, and, as a consequence, better visibility across both traditional and AI search.
When a client asks me “is X a ranking factor?” I now say: “That’s the wrong question. The right question is whether doing X makes your site more useful, more trustworthy, or more likely to be referenced by the systems – human and AI – that determine what people see when they search.” It takes longer to say, but it leads to better decisions every single time.
The ranking factor framing gave us a shared vocabulary. I get why people are reluctant to let it go. But the vocabulary was always a simplification, and in 2026 it’s become an obstacle. Google’s own people have been telling us this for years. The smartest SEOs I know have already moved on. Now everyone needs to catch up.
If you’re rethinking how your team talks about SEO and organic visibility, I help in-house teams and founders build strategies that work across traditional and AI search. Get in touch below.
