Does page speed affect SEO? Yes, but not the score you are looking at
Google says Core Web Vitals are used by its ranking systems. It ranks on real visitors' timings, not a PageSpeed lab score, and relevance still comes first.
Marketers ask whether page speed affects SEO usually because a PageSpeed report has just turned red, and an owner wants to know whether it matters. Google answers the question itself, in its own documentation, and the answer is more specific than yes or no. The Baltic Web Audit, our near-census of the Latvian, Estonian and Lithuanian web in July and August 2026, shows why the specifics matter: most sites in the region fail the speed test, while the same test gives them an SEO score of about 90 that measures something other than ranking.
In short
Yes. Google says plainly that Core Web Vitals are used by its ranking systems, and Largest Contentful Paint (LCP), the time until a page’s main content appears, is one of them. But those vitals measure how fast the page loaded for real visitors, not the score a PageSpeed lab test prints, and it says relevance comes first: a slow page that answers the question still beats a fast one that does not. In the Baltic audit’s throttled mobile test, 71% of Latvian, 68% of Lithuanian and 63% of Estonian sites rate poor on LCP, while the same run gives them average SEO scores of about 90. Neither number is your ranking. The first tells you where to look; Search Console tells you what Google sees.
What Google says, in its own words
All of this is from Google’s page experience documentation, read on 19 November 2026.
- Speed is a ranking input. “Core Web Vitals are used by our ranking systems.” Google recommends site owners achieve good Core Web Vitals “for success with Search”.
- It is one input among several. “There is no single signal.” Beyond Core Web Vitals, Google says, other page experience aspects “don’t directly help your website rank higher in search results”.
- Relevance wins. “Google Search always seeks to show the most relevant content, even if the page experience is sub-par.” Good page experience can contribute to success where “there is lots of helpful content available”, which is to say where pages are otherwise close.
- A perfect score is not the goal. A good result in Search Console’s report or in third-party tools “doesn’t guarantee that your pages will rank at the top”, and “trying to get a perfect score just for SEO reasons may not be the best use of your time.”
So speed works as a tiebreaker you cannot afford to lose, not as a lever that lifts a thin page. That is the honest version, and it is the one we give clients.
The score you see is not the one Google uses
This is where most of the confusion comes from. PageSpeed Insights shows two kinds of data on the same screen. The large performance score is lab data, generated by Lighthouse “simulating page load”. Beside it, when a site has enough traffic, is field data from the Chrome User Experience Report: what real Chrome users experienced over the previous 28-day collection period.
The vitals Google’s ranking systems use are the real-visitor kind: Google’s Core Web Vitals page describes them as “a set of metrics that measure real-world user experience”. Its LCP definition sets the bar at 2.5 seconds or less for good and over 4.0 seconds for poor, and says the threshold to measure is “the 75th percentile of page loads”, split by mobile and desktop. In plain terms: three quarters of your real visits have to see the main content within 2.5 seconds.
Two consequences follow:
- A lab score can be red while real visitors are fine, for example when most of your visitors are on fast office connections. The reverse also happens. The lab test is a diagnostic, not a verdict.
- A small site may have no field data at all. Search Console’s Core Web Vitals report shows “No data available” when there is not enough data in the Chrome UX Report for that device type. When it does have data, it groups similar pages and rates each group by its worst metric.
The Baltic picture: slow in the lab, and the SEO score hides a second gap
The audit ran Google’s PageSpeed test, in its throttled mobile setting, across the region. These are lab figures, so they describe how the sites behave under a fixed test, not the field figure Google ranks on, but the pattern is too strong to be a test artefact. Every figure here is from the report’s speed chapter and its basics chapter.
| Latvia | Estonia | Lithuania | |
|---|---|---|---|
| Median mobile LCP | 6.60 s | 5.73 s | 6.33 s |
| Sites in the good band (2.5 s or less) | 15% | 22% | 18% |
| Sites in the poor band (over 4 s) | 71% | 63% | 68% |
| Average PageSpeed SEO score | 90.1 | 90.0 | 90.7 |
| Homepages with no meta description | 44.0% | 49.9% | 45.1% |
The fourth row is the one that misleads. A PageSpeed SEO score of 90 looks like a pass, and owners read it as “the SEO is done”. The report’s own verdict is that this score “measures whether a machine can get in: is there a title, can a search engine fetch the page, does it say it is mobile-ready. It does not measure whether anyone actually wrote anything.” The fifth row is the proof: close to half the homepages in every country have not written the one sentence a site gets to put in its own search result.
So a typical Baltic site has room to gain in two places at once: the speed layer, where Google’s ranking systems read real-visitor timings, and the written layer, which no automated score checks. Of the two, Google’s own documentation says the written layer, relevance, comes first. Why sites end up at six seconds, cause by cause, is in the six-second web, and what TTFB, FCP and LCP each mean is in our guide to the speed metrics.
Improving LCP: what you can check yourself, and what needs a developer
Google’s guide to optimizing LCP splits the time into four parts: time to first byte, resource load delay, resource load duration and element render delay. That split is useful because it tells you who has to fix what.
What an owner can do alone, in an afternoon:
- Read the right number. Open Search Console, go to Core Web Vitals, and look at the mobile report. If it says “No data available”, use the field section of PageSpeed Insights for your homepage, and if that is empty too, treat the lab result as a guide to what to fix rather than as a grade.
- Find the main element and shrink it. On most sites the largest thing on the first screen is the hero image or a slider. Replace it with a copy sized for a phone screen and saved in a modern format, and remove slider slides nobody clicks.
What needs a developer:
- The image is lazy-loaded or found late. Google is blunt: “Never lazy-load your LCP image.” WordPress has added
fetchpriority="high"to the image it judges most likely to be the main one automatically since version 6.3, but page builders, sliders and background images set in CSS often route around it. Fixing that is a template change. - The first byte is slow. If the server takes seconds to answer, the other three parts cannot make up for it. That is caching, hosting or heavy code running on every request.
- The theme loads everything everywhere. Render delay usually comes from stylesheets and scripts the page does not need. Removing them without breaking the design is the structural half of the job.
That structural half is what our speed optimization service does, and the written half is what SEO work is for. If your report is red and your rankings are fine, the honest advice is still to fix the speed, but not to expect it to carry a page that does not answer the question.