SEO and user experience: the six questions Google asks itself, and the one answer it uses for ranking

Google's ranking systems use one measurable part of user experience directly: Core Web Vitals. Everything else in the page experience bundle, which Google lists as HTTPS, mobile display, ad density, intrusive interstitials and a clear separation of main content, makes a site better to use without lifting its position on its own. Google states this plainly in its page experience documentation, read on 20 October 2026: "Beyond Core Web Vitals, other page experience aspects don't directly help your website rank higher in search results. However, they can make your website more satisfying to use, which is generally aligned with what our ranking systems seek to reward." So the useful way to plan UX work for search is two lists: the one short list that feeds a ranking system, and the longer list that decides whether the visit was worth anything.

The six questions Google tells you to ask about a page

The page experience document is written as a self-assessment rather than a score. These are its six questions, in its order:

• Do your pages have good Core Web Vitals?
• Are your pages served in a secure fashion?
• Does your content display well on mobile devices?
• Does your content avoid using an excessive amount of ads that distract from or interfere with the main content?
• Do your pages avoid using intrusive interstitials?
• Is your page designed so visitors can easily distinguish the main content from other content on your page?

Google adds that the list is not exhaustive. Read as a brief rather than a checklist, it is a good one: five of the six are about getting out of the reader's way, and the sixth is about speed.

There is no single page experience signal

The document answers this directly: "There is no single signal. Our core ranking systems look at a variety of signals that align with overall page experience." That matters because the 2021 page experience update left behind a widespread belief in one combined score, and tools still present a single number as if Google held one.

Two more sentences from the same document are worth keeping in front of a design team. On how it is assessed: "Our core ranking systems generally evaluate content on a page-specific basis, including when understanding aspects related to page experience. However, we do have some site-wide assessments." And on chasing a perfect result: "trying to get a perfect score just for SEO reasons may not be the best use of your time."

The part that does feed ranking, and its thresholds

Core Web Vitals are three field measurements taken from real Chrome users: how fast the main content paints, how quickly the page answers an interaction, and how much the layout shifts while it loads. Each has a published threshold, each has a different set of usual causes, and each is fixed by different people on most teams. The metrics, the pass marks and the usual causes of each failure are set out in Core Web Vitals, the three metrics and the thresholds, which is the page to work from when the Search Console report turns amber.

One practical note that saves arguments: the Core Web Vitals report in Search Console is field data from real visits, and a lab tool run on your laptop is a simulation. They disagree for good reasons, and the field data is the one the ranking systems read.

The UX numbers that are not ranking inputs

Metric	What it is good for	Its role in ranking
Bounce rate, engagement rate	Spotting a page whose content does not match the query it wins	Not a ranking input. It is an analytics number computed by your own analytics tool
Time on page, dwell	Comparing two versions of the same page in a test	Not published as a ranking input, and not measurable from outside your site
Accessibility score	Legal obligation, and a larger addressable audience	Indirect. A page a screen reader can parse is usually a page a crawler can parse
Design quality, brand feel	Conversion and repeat visits	Indirect, through the behaviour it produces off the search result

Bounce rate attracts more wrong advice than anything else on that list, partly because the definition changed when Google Analytics 4 arrived. What it measures now, and the honest way to use it, is covered in bounce rate and SEO.

Where UX and search genuinely meet

The overlap is bigger than the ranking input list suggests, because four pieces of UX work change what a crawler and a reader both receive:

• Information architecture. The navigation a visitor uses is the internal link graph a crawler follows. A page three clicks from the home page through clear labels is easy for both.
• One job per page. A page that answers one question well satisfies the searcher who clicked and gives the search engine an unambiguous subject. Pages that try to serve three intents tend to rank for none of them.
• Content order. Putting the answer in the first paragraph is a usability decision and a snippet decision at once.
• Interface rendering. Tabs, accordions, infinite scroll and client-side routing are UX patterns with crawling consequences. Content hidden behind an interaction can still be indexed when it is in the HTML, and content that only exists after a click often is not.

That last one is where most UX and SEO disagreements actually live, and it is settled by testing rather than by opinion. Google's own documentation set is the reference for all of it, and the documentation Google publishes itself maps which document settles which argument.

A working order for a UX and SEO programme

1. Fix anything that blocks access first: a page that fails on mobile, an interstitial that covers the content, a template that is not served over HTTPS.
2. Take the Core Web Vitals reading from field data, by template rather than by URL, and fix the worst template.
3. Check that the main content is visually distinguishable from the furniture on your three highest traffic templates.
4. Audit ad and pop-up density on the pages that earn search clicks, not on the home page.
5. Then, and only then, spend time on design refinements. They pay in conversion rather than in position, which is still worth having.

Questions about SEO and user experience

Is user experience a Google ranking factor?

One part of it is. Core Web Vitals are used by Google's ranking systems, which Google states in its page experience documentation. The other page experience aspects are described there as not directly helping a page rank higher, while still being worth improving.

Did Google remove the page experience signal?

Google retired the separate page experience report in Search Console and folded the guidance into one document, and it stopped describing page experience as a single signal. Core Web Vitals kept their own report and their role in ranking.

Does a better looking site rank higher?

Not for being better looking. It ranks higher when the redesign also made the content clearer, the page faster, the navigation shallower and the main content easier to find, which is what most good redesigns do.

Should I aim for a perfect PageSpeed score?

No. Aim to pass the three Core Web Vitals thresholds on real visits, then stop. Google's own guidance says chasing a perfect score purely for search is probably not the best use of the time.

Who should own this work, design or SEO?

Split it by artefact. Design owns layout, interaction and the ad and interstitial rules. Engineering owns the three Core Web Vitals. SEO owns the brief, the measurement and the question of whether the page still answers the query it wins. One shared document with the six questions above, reviewed per template, keeps it from becoming an argument.

How often should the reading be taken?

Monthly, by template, from field data, so you are always working from current numbers and can tell a release from a seasonal change.

The reader who never sees your layout

Every improvement above assumes somebody arrives on the page. A growing share of the people deciding whether to consider you never do: they ask an assistant to compare the options and read a paragraph it assembled from pages it crawled. For that reader, your typography, your hero image and your interaction design are invisible, and what survives is the plain text, the structure of the headings and the clarity of the facts. A page that reads well as flat text to a machine and well as a layout to a person is the version that now works twice. This is an audience no usability study can recruit, and there is a cheap way to meet it: turn the stylesheet off on your most important page and read what is left, then ask an assistant the question a buyer would and compare the two. AI Knows Us is built for the second half of that exercise, putting the assistant's description of a market beside the pages it drew on, and the gap between those pages and the one a team just redesigned is usually the finding.