A Google Lighthouse update changes how PageSpeed Insights simulates page loads, which can raise lab scores on pages that have not changed, Search Engine Journal reported on November 2, 2026. The news matters because PageSpeed Insights scores are widely quoted, compared, and chased, and a change in how the simulation works can move those scores without any change to the page itself. A higher score after this update may mean the measurement changed, not the site.
The update changes how PageSpeed Insights simulates page loading in its lab environment. Lab simulations are controlled tests: they load a page under standardised conditions and measure how quickly it becomes usable. Changing the simulation changes the conditions of that test, and different conditions produce different scores, even for identical pages.
The reported effect is that scores can rise on pages that have not changed at all. That is not a bug in the usual sense; it is the expected consequence of updating a measurement tool. But it creates a trap for anyone who reads a score without asking which version of the tool produced it, because a before-and-after comparison across the update is no longer comparing like with like.
PageSpeed Insights lab scores have always been a proxy, not a direct measure of what real visitors experience. Real users arrive on different devices, networks, and locations, and their experience varies in ways a single lab simulation cannot capture. That is why the tool also reports field data from real users, which reflects actual visits rather than a simulated load.
The simulation change is a useful reminder of that distinction. A lab score is a diagnostic aid: it helps developers find bottlenecks and test fixes under consistent conditions. It is not a ranking in itself, and a lab improvement that does not show up in real-user data should not be celebrated as a user-experience win.
When a simulation becomes more generous, or models loading more efficiently, pages that previously scored poorly under the old model may score better under the new one without a single line of code changing. Teams that track scores over time may therefore see an apparent improvement that corresponds to no engineering effort and no benefit to visitors.
That apparent improvement can mislead in both directions. A team might claim a performance win it did not earn, or a team might stop work on real performance problems because the score now looks acceptable. Both mistakes come from treating the score as the goal, rather than treating faster real loading as the goal and the score as one imperfect measure of it.
Site owners should note the date of the update in their performance records and avoid comparing lab scores from before and after it as if they were measured the same way. New baselines should be set under the updated simulation, and future comparisons should stay within the new measurement regime.
More importantly, site owners should keep watching real-user performance alongside lab scores. If field data is stable while lab scores jump, the jump is a measurement effect, not a user benefit. If both improve together, the improvement is real. That cross-check, lab against field, is the discipline that keeps a simulation change from turning into a false success story.
Developers should also update any internal documentation or client reports that quote lab scores, so the version change is visible to anyone reading them later. A score quoted without its measurement context invites exactly the misreading this update makes possible, while a score quoted with its date and tool version stays honest even as the tool evolves.
Sources: Search Engine Journal, November 2, 2026, reporting the Google Lighthouse update to PageSpeed Insights.
Linkbaits helps you create content that earns citations and backlinks. Start with the free tools and build your first linkable asset this week.
Start free trial