ClarifySEO · 14 May 2026
GEO is changing SEO — and readability is caught in the middle
Optimising for search and writing for people are increasingly in conflict. Here's how to navigate the tradeoff — and why ignoring it will cost you rankings either way.
Most SEO advice treats content as a technical problem. Add the keyword. Hit the word count. Structure for featured snippets. It works — but only until your content starts reading like it was written for an algorithm, not a person.
The tradeoff nobody talks about is this: the more you optimise for search, the harder it becomes to write something worth reading. And as AI search changes what gets surfaced, that tension is getting worse, not better.
Even if you want to heavily prioritise SEO at the expense of readability, your bounce rate starts to climb, users spend less time on the page. Google takes notice, and then your page starts to fall off.
This article is about how to strike that balance.
The framework for making content decisions
Before making any content decision, two questions matter:
Will this help search or the reader? Not both — which one does it actually serve? Adding a keyword, restructuring a heading, padding a section for word count. Each of these has a potential search benefit and a potential readability cost.
What happens if you get it wrong? A page that ranks but doesn't engage sends negative signals back to Google. A page written purely for readers that nobody finds helps nobody either.
The balance shifts depending on where your site is. A new site needs to be found first — so leaning toward search signals early makes sense. A site with real traffic starts paying a real cost for every piece of content that reads like it was optimised rather than written.
How user experience affects rankings
Google doesn't publish a readability score, but the signals it measures tell the same story. Bounce rate, dwell time, pages per session — these are behavioural signals that reflect whether users found what they were looking for.
Research consistently supports the correlation. Backlinko's analysis of ranking factors found that mean dwell time for a first-page result is significantly higher than for results on page two. SEMrush's studies on ranking factors also find direct website visits and pages per session among the strongest correlators with higher rankings.
When a user clicks your result, reads for ten seconds, hits back and clicks a competitor, that's a signal. When they stay, scroll, and don't return to the results page, that's a different signal. The correlation between genuine engagement and ranking is strong enough that ignoring it is a real risk.
The practical implication: SEO and UX aren't competing priorities. They're the same priority measured differently.
Source: Backlinko, Search Engine Ranking Factors
Google's helpful content system
In 2022, Google announced the Helpful Content Update — a system designed to ensure search results surface original, useful content written by people, for people, rather than content created primarily to rank.
In March 2024, it stopped being a separate update and became part of Google's core ranking systems. That's significant. The question "is this content genuinely helpful or is it just optimised?" is now baked into every ranking decision Google makes, not applied periodically.
This means the ceiling for purely optimised content has lowered, and the floor for genuinely useful content has raised. Chasing rankings at the expense of real usefulness isn't just bad for your audience — it's now directly contrary to how Google's core systems evaluate your pages.
Sources: Google, Helpful Content Update 2022 · Google, Core Update March 2024
E-E-A-T and why it matters more now
Alongside the helpful content system, Google's E-E-A-T framework — Experience, Expertise, Authoritativeness, and Trustworthiness — has become increasingly central to how content is evaluated.
In an era where AI can generate technically correct, well-structured content at scale, E-E-A-T is Google's mechanism for surfacing content backed by real knowledge and real experience. It's why authorship, sourcing, and demonstrated expertise matter more now than keyword density ever did.
For smaller sites this is actually an advantage. A site owner writing from direct experience about their niche has something AI-generated content can't replicate — genuine first-hand knowledge. That's worth leaning into.
Source: Google, E-E-A-T
The traditional ranking factors still apply
The established SEO ranking factors — quality content, backlinks, technical SEO, keyword optimisation, and user experience — remain the foundation. But AI search is shifting how each of these is weighted.
Quality content and user experience are moving up. Keyword optimisation as a standalone activity is moving down. Backlinks still matter but topical authority — covering a subject comprehensively across multiple pieces — is increasingly what AI overviews and search rankings reward.
The sites that will do well in search over the next few years are the ones that build genuine authority on specific topics, write for readers first, and let the SEO signals follow from that.
Source: Backlinko, Top SEO Ranking Factors
The tradeoffs that still catch sites out
Publishing frequency vs content quality
More content signals activity. But thin, rushed content actively damages topical authority and wastes crawl budget — and with the helpful content system now part of core ranking, low-quality volume is a liability, not just a missed opportunity.
Fewer, better pieces tend to outperform high-volume output. Quality compounds over time in a way that rushed content rarely does.
Sources: Google, Crawl Budget · Backlinko, Content Marketing
Keyword optimisation vs natural writing
There's a version of keyword optimisation that improves content — using the language your audience actually searches for makes your writing clearer and more relevant. There's another version that damages it — forcing keywords into sentences where they don't belong, repeating phrases unnaturally, restructuring paragraphs around search terms rather than ideas.
The first version is worth doing. The second is what triggers the readability cost.
Source: Google, Write for people, not search engines
Technical SEO vs content investment
A site that can't be crawled can't rank. But beyond the fundamentals — indexability, page speed, mobile usability — technical SEO has diminishing returns for most sites.
Get the fundamentals right once, then leave it. Spending ongoing time on technical SEO when your content is thin is optimising the wrong variable.
Getting your technical fundamentals right is worth prioritising early. But once covered, every additional hour spent on technical work is an hour not spent on content. And it's worth asking: even if a page ranks, if it doesn't drive conversions or meaningful action, how much does the traffic actually matter?
What this means in practice
The question isn't "what should I do for SEO?" It's "what should I do given that Google is now explicitly rewarding content that serves readers, not algorithms?"
The answer is the same for most sites: write clearly, write from experience, cite your sources, and make sure people who land on your pages actually find what they were looking for. The SEO signals follow from that — not the other way around.
My main recommendation for any site owner currently prioritising SEO: start with what you want to be discovered for. Build content around that topic that users actually want to read. Then look for natural ways to include the language your audience searches for — without forcing it at the expense of readability.
For AI search specifically, FAQs and FAQ schema are a practical starting point. Answering specific questions directly, and including clear descriptions of what your site does and who it's for, serves both traditional SEO and the way AI systems surface content. The overlap is bigger than most people think.
Want to see how your current content is performing?
Run a free audit and see exactly where you stand.