Speed
Our measurements match what your visitors actually get
performance.lab-and-field-disagree
Why this matters
A scan loads a site once, on one connection, from one place. Google's figures come from four weeks of real Chrome visitors on whatever devices and connections they happened to have. When the two disagree it is worth knowing which way: a site that tests well and performs badly for real people usually has slow pages beyond the one that was tested, or visitors on worse phones than the test assumes. A site that tests badly and performs well in the field is being judged too harshly by the test, and saying so is more useful than defending the number.
Who fixes it
You can, usually
Roughly how long
1–2 hours
Care needed
Low risk to change
How to fix it
Where real visitors have it worse, the home page is usually not the problem — check the pages people actually arrive on, which for most sites means a few specific product or service pages rather than the front door. Where they have it better, there is nothing to do: the lab number is a deliberately pessimistic simulation and your visitors are on better connections than it assumes.
How we score it
Failing this check takes up to 12 points off your speed score. This one is a measurement rather than a fact about your markup, so two scans of an unchanged site can legitimately differ a little.
Does your site pass this one?
This check runs on every scan, along with the other 106. Free, no account, and you see the evidence for each result.
Check my siteOther speed checks
- Every page gets the files it asks forA file that is missing on one template is missing on every page built from it, which is usually hundreds of pages rather than one. It is also the hardest kind to notice from the inside: a browser that has an older copy of the file will carry on showing it long after the file itself has gone, so the site looks correct to whoever built it and broken to everybody arriving fresh.
- Every page responds promptlyWhen one part of a site responds noticeably slower than the rest, it usually points at one specific thing rather than at the hosting in general — a heavy search query, an unindexed database lookup, or a page rebuilding something on every request. That makes it one of the more fixable speed problems, because there is a single cause to find.
- Everything the page asks for loadsYour page requests these and gets nothing back. Depending on what they are, that means a missing image, a stylesheet that never applies, or a script that never runs — and the page may look fine to you if your browser has an older copy cached.
- First visit after a quiet spell is not slowSome hosting puts a site to sleep when nothing has visited for a while, and the next request has to wake it. Once warm the site is fine, which is why this is easy to miss — the person testing it has usually just loaded the page. The people who get the slow version are the ones arriving overnight, and Google when it crawls at an odd hour.
- Google has enough visitor data to report on your siteGoogle publishes real-visitor speed figures for a site only once enough Chrome users have been there to make the numbers meaningful and anonymous. Below that threshold there is nothing to publish, which is the ordinary situation for a small business site and says nothing about how well the site is built. It does mean every speed figure in a report like this one is a simulation rather than a record of what happened to anybody.
- Images reserve their space before loadingThe browser cannot tell how tall these images will be until they have downloaded, so it draws the page without them and then pushes everything down when they arrive. That is the jump people experience as a page moving under their thumb.