Definition
User experience (UX) is how people perceive and use a website or product: whether they can find what they came for, how fast and stable the pages feel, whether the content is readable and accessible, and how much friction stands between them and their goal. In search, UX matters in two ways. Some parts of it feed ranking systems directly, and all of it decides what happens after the click.
Google's page experience documentation is explicit about the first part. There is no single page experience signal. Its core ranking systems use Core Web Vitals, and Google recommends meeting the thresholds that still apply in 2026: Largest Contentful Paint within 2.5 seconds, Interaction to Next Paint of 200 ms or less, and Cumulative Layout Shift of 0.1 or less, measured at the 75th percentile of real visits. Other page experience aspects, such as HTTPS, display on mobile, ad density, intrusive interstitials and a clear separation of main content, do not directly raise rankings according to Google, but they align with what its systems try to reward. Google also says relevance comes first: a page with weak UX can still rank when it is the best answer. Metrics like bounce rate and dwell time are not documented ranking signals; treat them as diagnostics.
AI search adds three new angles:
- Machine readers. AI crawlers such as GPTBot, ClaudeBot and PerplexityBot generally read the HTML they receive without running JavaScript, so content that appears only after client-side rendering is missing for them (JavaScript rendering). Clean, server-rendered main content is part of UX for these visitors.
- Fewer, later clicks. Pew Research found in March 2025 browsing data that users clicked a traditional result in 8% of Google visits with an AI summary, against 15% without one. Visitors who do arrive from AI answers often land deep on a specific page in the middle of a decision, so that page has to answer the follow-up question and offer a clear next step. Measure this as AI-referred traffic by landing page, not sitewide.
- Agents. Browsing agents such as ChatGPT's agent fill forms and compare plans on a user's behalf. Clear labels, visible prices and forms that work without fragile scripts help both people and agents; see agent experience optimization.
To work on UX, start with field data: the Core Web Vitals report in Search Console or CrUX, split by mobile and desktop and by URL group. Add analytics segmented by source, so AI referrals can be compared with organic search on the same pages, and run usability tests on the pages that receive the most AI and search visitors. For accessibility, WCAG 2.2 is the current W3C reference. Then fix the issues that block tasks first, such as slow interactions, layout shifts near buttons, hidden pricing and broken forms, before cosmetic changes. Good UX does not replace relevance, but it determines how much value each visit from search or an AI answer produces, which matters more as those visits become scarcer (zero-click search). See also page experience and conversion rate optimization.
Examples of User Experience (UX)
- A B2B site sees visitors from ChatGPT land on its pricing page and leave, because plan details sit behind a 'contact sales' form. It adds visible price ranges and a short comparison table, then compares demo requests from that segment month over month.
- An online store's mobile INP sits at 450 ms because a third-party reviews widget blocks the main thread. Deferring the widget brings the URL group under 200 ms in the Search Console Core Web Vitals report.
- A documentation site fetches article text with client-side JavaScript. Human readers see it, but AI crawlers receive an empty template, so the team switches the docs to static generation.
- A news publisher removes a full-screen newsletter interstitial from article pages after reviewing Google's guidance on intrusive interstitials, and replaces it with an inline signup box below the first section.
