Yes, but it is a small factor, and it is almost never the reason a page sits on the second page of results. Anyone selling you speed optimization on the promise of rankings is overstating what it does.
That is worth saying plainly first, because it is the honest answer and because we sell speed optimization. The argument for making a WordPress site fast is a good one. It is just not primarily a ranking argument, and the sites that buy it for rankings alone are usually disappointed.
Here is what Google actually measures, what it does with the numbers, and where the return on the work really comes from.
What Google measures
Three metrics, together called Core Web Vitals. Each has a threshold, and a page either passes or it does not.
- Largest Contentful Paint — 2.5 seconds or under. How long until the biggest thing on screen has drawn. On most WordPress pages that is the hero image or the headline. This is the one people mean when they say a site feels slow.
- Interaction to Next Paint — 200 milliseconds or under. How long the page takes to visibly respond after someone taps or clicks. It replaced First Input Delay in March 2024 and it is a much harder test, because it measures every interaction rather than only the first one. Plenty of sites that passed the old metric fail this one.
- Cumulative Layout Shift — 0.1 or under. How much the page moves around while it loads. This is the one that makes people tap the wrong thing when an ad or a late-loading banner pushes the content down.
Why your PageSpeed score of 92 means nothing
This is the single most misunderstood part of the subject, and it is where most speed work is aimed at the wrong target.
The score out of 100 at the top of PageSpeed Insights is a lab result: one simulated load, on a simulated device, from a simulated connection, right now. It is a diagnostic tool. It is not what Google uses.
What Google uses is field data — the Chrome User Experience Report, collected from real Chrome users on real devices and real connections. Three things about it matter:
- It is measured at the 75th percentile. Your average visitor is irrelevant; the question is whether three visitors in four had a good experience. One slow quarter of your audience fails the page.
- It is a 28-day rolling window. Deploy a fix today and the reported number moves over the following month, not this afternoon. Anyone who reports an improvement the same day is reading the lab score.
- It is mostly mobile, on hardware slower than the machine you are testing from, on networks worse than your office. A desktop score of 92 and a mobile field failure at the same time is completely ordinary.
Use the Core Web Vitals report in Search Console rather than the score. It shows field data, grouped by URL pattern, split by mobile and desktop. That is the number Google is acting on.
How much of a ranking factor it really is
Google has been consistent about this: page experience signals, including Core Web Vitals, are used in ranking, and relevance wins over them every time. Their own guidance has repeatedly described it as something closer to a tiebreaker — where two pages are similarly useful, the better experience can be preferred.
What follows from that is worth being blunt about. Making a page faster does not make it more relevant. If your page is at position 40 because five better pages answer the question more completely, taking LCP from 4.8 seconds to 1.2 will not move it to position 4. It might move it from 40 to 37.
Where it does change the outcome is in a close race — a competitive commercial search where four or five sites are all credible, all thorough, and all roughly as well known. There, being the only one that loads properly on a phone is a genuine edge, and it is an edge most of your competitors have not bothered to take.
The indirect effects, which are larger
Two things that are not ranking factors matter more than the ranking factor.
Crawling. Google adjusts how much it crawls a site based on how the server responds. A site that answers slowly, or starts returning errors under crawl load, gets crawled less. On a small brochure site this is invisible. On a store with twelve thousand product URLs it is the difference between new stock appearing in search this week and appearing next month. Time to first byte is the number that matters here, and it is a server and database question rather than a front-end one.
What happens after the click. Ranking is only worth what the visitor does next. A page that takes eight seconds on a phone loses a large share of its arrivals before anything has drawn, and those people go back to the results and click the next site. Whether or not Google counts that against you, you have paid for the position and thrown away the visit.
The real argument for speed is money, not rankings
On a store, checkout speed is revenue. A checkout that takes six or seven seconds to become usable loses orders that were already decided — the customer wanted the thing, put it in the basket, and left because the page would not respond. That is not a ranking problem and no amount of ranking fixes it.
It is also the one area where a fix is measurable within days rather than weeks. Rankings move slowly and confound with a dozen other things. Time to first byte, checkout completion time and abandonment are measured before and after, on the same host, with the same traffic. If we take a checkout from 6.8 seconds to 1.9, that number is either there or it is not.
The related mechanics are covered in why your WooCommerce checkout is slow, including why caching plugins do nothing for the checkout page specifically.
What actually makes a WordPress site slow
Not the theme, most of the time, and not the host as often as people assume. The recurring causes, roughly in order of how much they cost:
- Database queries that grow with the site. A product page running 412 queries was not slow when the catalogue had eighty items. Uncached options, autoloaded data and plugins querying on every request are the usual sources.
- Plugins doing work on pages that do not need them. A booking system loading its scripts on every URL on the site, including the ones with no booking form.
- Images at the wrong size. A 3,000-pixel photograph displayed in a 400-pixel column, and often several of them above the fold.
- Two caching plugins fighting. Installed a year apart by two different people, each undoing the other’s work. This is common enough that we check for it first.
- Third-party scripts. Chat widgets, tag managers, heat maps and pixels, each loading its own payload from someone else’s server. They are frequently the whole of an INP failure.
What this cannot fix
- Speed will not rank a page that is not relevant. If the page does not answer the search, it stays where it is at any load time. We say this before quoting, not after.
- It will not undo a drop caused by something else. Rankings that fell after a redesign, a migration or a manual action fell for that reason. Speed work does not reverse it.
- A perfect score is not the goal and is sometimes not reachable. If the business depends on a chat widget and a payment provider’s script, some of the page weight is not ours to remove. We say what is possible before starting.
- We do not sell SEO or content. We fix what a browser and a crawler experience. If the remaining problem is that a competitor has written a better page, that is real, and it is not something we can bill you to solve.
Questions we get asked
Is page speed a Google ranking factor in 2026?
Yes. Core Web Vitals are part of Google’s page experience signals and are used in ranking. They are a light-touch factor that works mainly as a tiebreaker between comparable pages, and relevance outranks them consistently. Treat passing the thresholds as removing a handicap rather than as a way to overtake better pages.
What is a good PageSpeed Insights score?
The score is a lab diagnostic and Google does not rank on it. What matters is whether the field data at the top of the same report passes the three thresholds: LCP 2.5 seconds, INP 200 milliseconds, CLS 0.1, measured at the 75th percentile of real visitors over 28 days. A page can score 95 in the lab and fail in the field, usually because real visitors are on phones and slower connections than the simulation.
How long after fixing speed do rankings improve?
The field data itself takes about 28 days to fully reflect a change, because the window is rolling. Any ranking effect follows that, and it is usually small enough to be hard to separate from everything else happening at the same time. Judge the work on the measurements — time to first byte, LCP, INP, conversion — rather than on the position graph, because those move immediately and unambiguously.
Will a faster host fix Core Web Vitals?
Partly. Better hosting improves time to first byte, which helps LCP and helps crawling. It does nothing for INP or CLS, which are decided by the page’s own JavaScript and layout, and it does nothing about a page running 400 database queries — that page just runs them on a faster machine. Hosting is usually a fraction of the problem on a slow WordPress site.
Do caching plugins fix Core Web Vitals?
They fix one part of it well. Page caching turns a slow server response into a fast one for anonymous visitors, which helps LCP considerably. It does not apply to logged-in users, carts or checkouts, which are excluded by design, and it does not touch INP, because the JavaScript still runs in the browser either way. A cached slow page is a fast-loading page that still stutters when you tap it.
What does WordPress speed optimization cost?
Speed optimization starts at $349. The free written audit turns that floor into one fixed number: what is actually slow, measured, what we would change, what we would leave alone, and the price before any work begins. If the site is already fast enough and the real problem is elsewhere, the audit says so.