Soft Navigation and Core Web Vitals: Measuring SPAs in 2026
In-page route changes in single-page apps now enter Core Web Vitals data as an experiment. Find out whose score may change and what to do about it.
rabbitclip teamPublished: 6 min read
Short answer
In single-page apps (SPAs), in-page route changes now enter Core Web Vitals data as an experiment, thanks to Chrome's soft navigation measurement. The September 2026 release of the Chrome UX Report published new experimental variants of LCP, INP and CLS that count these transitions too. If your site changes its address without a full page reload, the browsing that follows the first load has become measurable; if it is a classic site that loads a new page on every click, this change does not directly concern you.
This is not a penalty or a ranking announcement, it is a measurement announcement. Chrome's documentation says the feature is experimental and does not say when it will leave that status. Even so, knowing where the data comes from means nobody can surprise you with it later.
Below you will find what a soft navigation is, what the new data shows, the criteria by which Chrome counts a transition, how to measure it, the mistakes made most often and the situations in which you do not need this measurement.
What is a soft navigation?
A soft navigation is an in-page transition in which the address in the browser bar changes without the page being loaded again. Imagine clicking a product in a list and landing on its product page without a refresh: the address changes, the content changes, and the browser downloads no new document. Single-page apps, meaning sites whose pages are built with JavaScript, work this way.
Until now Core Web Vitals measurements mostly rested on the first load. In an SPA, even if a visitor browsed ten pages after the first one, the speed or layout shift of that browsing did not appear as separate data. Soft navigation measurement aims to fill that gap.
What does the new Chrome UX Report data show?
According to the Chrome UX Report release notes for September 2026, published on 13 October 2026, LCP, INP and CLS appear as new experimental variants that take soft navigation into account, in the BigQuery dataset and in the CrUX APIs. The same notes give the share of 'good' results across all CrUX origins without and with soft navigation.
These shares are totals across all origins, not your own site's score. Even if the total improves, the result for a single site need not move in the same direction. The release notes say the CrUX Vis tool was updated too and is meant to show the effect per site; reading your own case there is the sound approach.
- Good LCP share: 68.3% before, 72.1% after
- Good CLS share: 81.7% before, 85.1% after
- Good INP share: 85.4% before, 87.4% after
- Good on all three metrics: 55.9% before, 59.5% after
By what criteria does Chrome count a soft navigation?
Chrome's platform documentation (last updated 2 September 2026) lists three criteria: the transition must be started by a user action, the user must see a visible address change, and the interaction must result in a visible paint. If all three do not happen together, the transition is not counted as a soft navigation.
According to the documentation the feature has been on by default since Chrome 151, and Edge 151 supports it. Firefox and Safari do not, so this measurement comes only from your Chrome and Edge visitors. That is an honest detail that limits the scope of the data.
- A user action starts the transition
- The user sees a visible address change
- The interaction ends in a visible paint
How do you measure this on your own site?
If you run your own tracking code, the documentation points to the reportSoftNavs: true option in version 6.0.0 and later of the web-vitals library. Each record carries a navigationId and a navigationURL, so you can tell which transition belongs to which address. On soft navigation records TTFB is reported as zero because no new document was requested; do not read that zero as your server being very fast.
During development, according to the Chrome DevTools notes for October 2026, the Performance panel now fully supports soft navigations. For real visitor data in the field look at CrUX Vis, and for lab testing use DevTools. For the general state of your speed the rabbitclip site speed tool can be a starting point, though it gives no soft navigation specific data.
What mistakes are made most often?
The most common mistake is confusing the new experimental figure with the old score. The variants are published separately, and if you compare two numbers without knowing which one you are looking at, you will reach the wrong conclusion.
The second is improving only the first load and never testing the transitions that follow. In a single-page app most of a visitor's time passes after the first load.
- Confusing the experimental variant with the old score
- Testing only the first load
- Taking a zero TTFB for success
- Taking the overall share for your own site's result
When do you not need this measurement?
If your site loads a new page on every click, that is, it serves classic server output, soft navigation produces no separate result for you. A small brochure site is often in this group and can do without watching the new data. Even so, knowing which kind of site you have is the first step; asking your developer how the site handles page transitions is enough to settle the matter with a single question.
The wrong choice is rebuilding a working site as an SPA because of this news. The way of measuring has changed; the question of whether your site works well has stayed the same. See today's position first, then decide.
In short, Chrome now measures transitions that happen without a page reload as an experiment, and that data is visible in CrUX. If your site is a single-page app, look at CrUX Vis, consider the soft navigation option in your tracking code and test the transitions that follow the first load as well. If you would like to read together how your site stands in this measurement, we are glad to have a calm discovery conversation with you at rabbitclip.
FAQ
Does soft navigation affect Google rankings?
The sources we used say nothing of the kind. The data appears in the Chrome UX Report as an experimental variant. We have no information about any effect on rankings, so we do not speak with certainty.
When will the experimental status end?
The release notes give no date. For that reason we do not guess one; following the official release notes is the safest way.
Are my Safari and Firefox visitors included in this measurement?
No. According to Chrome's documentation Firefox and Safari do not support soft navigations; the measurement comes from visitors using Chrome and Edge 151.
How do I get a soft navigation report for my own site?
For field data look at the CrUX Vis tool. In your own tracking code you can use the reportSoftNavs: true option in version 6.0.0 and later of the web-vitals library.
Sources
The specific facts in this article rest on these official pages; check them for the current position.
