TL;DR
- Chrome's Annie Sullivan said on web.dev on 13 October 2026 that CrUX now publishes experimental LCP, INP and CLS that include SPA soft navigations.
- Across all CrUX origins, the share passing all three Core Web Vitals rises from 55.9% (stable) to 59.5% when soft navigations are included, per Chrome's release notes.
- On Android, origins with 75-100% soft navigations go from 26% to 87% with good LCP; around 70% of origins have under 5% soft navigations and should see little change.
- Chrome aims to graduate the metrics from experimental in May 2027, subject to feedback; Search Console will then switch to the new data.
Google's Chrome team said on 13 October 2026 that the Chrome UX Report (CrUX) now publishes experimental versions of all three Core Web Vitals that count soft navigations, the in-app route changes that single-page applications (SPAs) make without a full page reload. The stable Core Web Vitals scores are unchanged for now. Chrome aims to move the new metrics out of experimental status in May 2027, subject to feedback, and once that happens Search Console will switch to the new data. For a site that already loads a fresh HTML page on every click, nothing should change. For an SPA where most page views happen after the first load, the reported numbers can change a lot.
What Chrome announced on 13 October 2026
The announcement came in a post on web.dev by Annie Sullivan of the Chrome team, published 13 October 2026. It follows the launch of the Soft Navigations web platform API in Chrome 151. Chrome is now adding three new fields to the CrUX dataset, each one an experimental twin of an existing Core Web Vital:
- experimental_largest_contentful_paint_with_softnavs measures Largest Contentful Paint (LCP) across both hard and soft navigations.
- experimental_interaction_to_next_paint_with_softnavs measures Interaction to Next Paint (INP) per individual soft navigation as well as per hard navigation, instead of charging every interaction in an SPA session to the first page load.
- experimental_cumulative_layout_shift_with_softnavs measures Cumulative Layout Shift (CLS) per view and URL, across hard and soft navigations.
The post is explicit that Google's stable Core Web Vitals scores stay as they are while these experimental versions are evaluated next to them. Chrome also said it expects to add experimental soft-navigation metrics for navigation types and for First Contentful Paint (FCP) in the next CrUX release in November 2026.
What a soft navigation is, and why CrUX missed it
A traditional multi-page site (an MPA, in Chrome's shorthand) asks the server for a new HTML document every time a visitor clicks a link. An SPA works differently. After the first load it uses JavaScript to swap the page content and update the URL in place. Chrome calls that transition a soft navigation. Because every framework handles it in its own way, measuring it has needed custom code, which made like-for-like comparison between sites hard.
The web.dev post describes what CrUX did before this change. When a user browsed an SPA, CrUX recorded only the initial "hard" page load on arrival. Later in-app page changes were not treated as separate loads, and every interaction and layout shift from the whole session was grouped into that first landing page. With soft navigations included, each in-app view change that updates both the content and the URL is measured as its own navigation.
There is a practical upside for debugging. According to the post, in the CrUX API the metrics for each soft navigation are reported under the URL the user navigated to, not the landing page, as long as that URL meets the usual CrUX eligibility criteria. Under the old model a slow product page reached by an in-app click would have its problems blamed on the home page or category page where the session started.
How many sites this touches
Most of the web will see no difference. Chrome said that across all CrUX origins around 70% have little or no soft-navigation traffic, meaning fewer than 5% of their page views are soft (66% to 75% depending on platform). Only about 7% to 11% of origins use soft navigations for more than half of their page views: 7.0% on Android and 10.6% on desktop.
The picture flips among the biggest sites. Among the top 1,000 websites, 35% on Android and 42% on desktop rely on soft navigations for more than half of their page views. So a change that is invisible for most blogs and brochure sites lands on a large share of the most-visited sites.
Sullivan's post puts the boundary plainly: if a site does not use JavaScript to swap content and update the URL without a full browser reload, its metrics should not change.
The before-and-after numbers
Chrome's CrUX release notes, published the same day for the dataset covering September 2026 (18,376,044 origins), give the effect across the whole dataset. The figures below are the share of all origins rated "good" on each metric, first under the stable definition and then with soft navigations included.
| Metric (share of all CrUX origins rated "good") | Stable, hard navigations only | Experimental, soft navigations included |
|---|---|---|
| Largest Contentful Paint (LCP) | 68.3% | 72.1% |
| Cumulative Layout Shift (CLS) | 81.7% | 85.1% |
| Interaction to Next Paint (INP) | 85.4% | 87.4% |
| All three Core Web Vitals passed | 55.9% | 59.5% |
Across the whole web the shift is modest, because most origins have almost no soft navigations. The large movements sit in the small group of SPA-heavy origins, and the web.dev post breaks those down metric by metric.
LCP: the biggest jump, with a caveat
The headline figure needs to be read carefully. In the web.dev post's Figure 2, origins whose page views are 75% to 100% soft navigations go from 26% with "good" LCP under the stable metric to 87% under the experimental metric, on Android. That is a share of origins in that one bucket, on one platform. It is not an all-web number and it is not a desktop number.
Chrome also explains why. Follow-up soft navigations reuse code and assets already downloaded during the slower first load, so they paint faster. The post says this is not because SPAs are faster than multi-page sites overall. Follow-up clicks are faster on any site, and the new metric simply brings SPAs with many follow-up visits in line with MPAs that also get many follow-up visits.
INP: better, but SPAs still trail
INP is now measured per page view, not as the slowest interaction across a whole browsing session. Chrome said this roughly doubles the Android INP pass rate for sites with mostly soft navigations. Even so, SPA-heavy sites still trail MPA sites on "good" INP overall, and the post warns that heavy main-thread JavaScript during route transitions can still cause INP problems.
CLS: closer to parity
Layout shifts are now split into individual page views instead of piling up across a long session. Chrome said this brings SPA-heavy sites in line with MPA sites on desktop and nearly level with them on Android.
Where to see the data, and the timeline
The experimental metrics are in the CrUX BigQuery dataset, the CrUX API and the CrUX History API starting with the October 2026 release. CrUX Vis, Chrome's visualisation tool, has been updated so the new metrics can be viewed for any URL or origin in the dataset.
The timeline Chrome gave is short and conditional:
- 13 October 2026: experimental LCP, INP and CLS with soft navigations published in CrUX.
- November 2026 CrUX release: experimental soft-navigation metrics for navigation types and FCP expected.
- May 2027: Chrome's target for graduating the metrics out of experimental status, "subject to feedback".
- After graduation: Search Console switches to the new data.
Chrome labels every metric "experimental" on purpose and invites feedback through the soft navigations specification repository. The post warns readers to expect some month-to-month shifts while the measurement is refined. Any number pulled today is a preview, not a final score.
What Chrome did not say
Several questions an SEO would ask are not answered in the post. It does not say whether Google Search ranking systems will use the soft-navigation data, or when. The only Search product it names is Search Console, and only for after graduation. It does not say whether PageSpeed Insights field data will change, and it gives no country-level or Thailand-level breakdown. It also does not promise May 2027 as a fixed date. Treat anything beyond those stated points as unknown.
Analysis: what to check now
The following is analysis based on the source, not something Chrome stated.
First, find out whether the site is an SPA in practice. A site built on a JavaScript framework can still do full page loads on most clicks, and a site that looks traditional can route client-side. Open CrUX Vis for the origin and compare the stable and experimental series. If the lines sit on top of each other, the site is among the majority that Chrome expects to see little or no change, and can move on.
Second, if the experimental numbers are much better, resist reading that as a fix. The LCP gain comes from follow-up views being cheaper to paint, which Chrome attributes to reused code and assets. The first-load LCP that a visitor from Google Search actually experiences has not changed. A slow landing page is still slow.
Third, look hardest at INP. Chrome's own conclusion is that SPA-heavy sites still trail on responsiveness even after the boost, and it points at main-thread JavaScript during route transitions. That is where a technical review of an SPA is likely to find work, for example in code that runs when a filter, tab or product route changes.
Fourth, use the per-URL attribution. Because soft-navigation metrics now land under the destination URL, pages that were invisible in field data (because users only ever reached them by in-app clicks) may start to show up with their own numbers once they pass eligibility. That gives a cleaner list of which templates to fix.
Fifth, plan reporting around the Search Console switch. When Search Console moves to the new data after graduation, Core Web Vitals reports for SPA-heavy sites may shift without any code change. Annotating that date in reporting avoids an unexplained jump being credited to, or blamed on, a release. A technical SEO audit is a sensible place to baseline the stable and experimental numbers side by side before then.
What this means for Thai marketers
CrUX includes Thai origins that have enough Chrome traffic, so Thai sites can look up their experimental numbers today in CrUX Vis. Chrome did not publish a Thailand-specific breakdown, so there is no way to say from the source how many Thai sites are in the SPA-heavy group.
For in-house teams and agencies, the useful move is to treat this as an early look at how Search Console will report these sites from mid-2027 if Chrome keeps to its target. If a redesign or a migration to a JavaScript framework is on the roadmap, the route-transition INP warning belongs in the brief for the build team. Teams working on website design and rebuilds can test route changes for main-thread work before launch, and ongoing SEO in Thailand reporting should note the change so that a later shift in Core Web Vitals is read correctly.
FAQ: CrUX soft navigation metrics
Has Google changed Core Web Vitals scores for SPAs?
No, the stable Core Web Vitals scores are unchanged. Chrome said on 13 October 2026 that the soft-navigation versions are experimental and are being evaluated alongside the stable metrics.
When will Search Console use soft navigation data?
Search Console will switch after the metrics graduate from experimental status, which Chrome aims to do in May 2027, subject to feedback. Chrome did not give a firmer date.
Will my site's Core Web Vitals change?
Only if the site uses JavaScript to swap page content and update the URL without a full reload. Chrome said around 70% of CrUX origins have fewer than 5% soft navigations and should see little or no change, while SPA-heavy sites will see their numbers improve or stay steady.
What does the 26% to 87% LCP figure mean?
It is the share of origins with "good" LCP among Android origins whose page views are 75% to 100% soft navigations, under the stable metric versus the experimental one. It does not describe all websites or desktop traffic.
Where can I check my own site?
The quickest view is CrUX Vis, which Chrome updated to show the new metrics for any URL or origin in CrUX. The same data is in the CrUX BigQuery dataset, the CrUX API and the CrUX History API from the October 2026 release.
The stable scores have not moved yet, but the scoring model for SPAs has a published target date. Sites that check their experimental numbers now will know what Search Console is likely to show before it shows it. If you want a second pair of eyes on what the new data says about your site, Relevant Audience can review it as part of a technical SEO audit.
