Is my PageSpeed score bad?

The short answer: on its own, a low PageSpeed score does not tell you whether real visitors have a problem. Check the field data before spending anything.

PageSpeed Insights gives you two things and most people only read one. The big coloured number is Lighthouse: one page load, on a datacentre machine, with the processor deliberately slowed and the network simulated as a poor connection. Further down, if your site has enough traffic, is field data — what actual Chrome users experienced over the last 28 days.

They frequently disagree.

We hold both for 121 businesses in the Local Digital Visibility Index. Every figure below is computed from the published data when this page is built.

How far apart they are

Largest Contentful Paint	Median across 121 businesses
Lighthouse (the lab test)	12.6s
Real Chrome visitors	1.7s

The lab figure is 7.4× the real-user figure. That is not Lighthouse malfunctioning — throttling is the whole point of a stress test. It just means the number is not a description of your customers.

Of the 121 businesses:

• 37 are rated poorly by the lab test
• 41 actually fail Core Web Vitals for real visitors
• the two agree for 77 and disagree for 44

Which of the four cases are you in

Lab bad, field fine (20 businesses). The most common reason someone is sold performance work they do not need. Your visitors are not experiencing the problem the screenshot shows.

Lab fine, field failing (24 businesses). The dangerous one, because a clean lab score reads as an all-clear. Real-user failures cluster in things one scripted load barely touches — layout shifting as images and banners arrive, and sluggishness while someone is actually tapping.

Both bad. Real, and worth fixing.

Both fine. Spend the money elsewhere.

How to check yours in two minutes

1. Open PageSpeed Insights and enter your address.
2. Ignore the big number at first. Look for “Discover what your real users are experiencing” at the top. That is the field data.
3. If it says Core Web Vitals passed, your visitors are broadly fine regardless of what the lab score says.
4. If there is no field section at all, your site does not have enough traffic to report — and then the lab score is the only evidence you have, so treat it as a lead rather than a verdict.
5. If it fails, note which metric. That is your actual problem, and it is usually not “the site is slow”.

What to ask whoever showed you the score

• “What does the field data say?” If they have not looked, they have not finished. It is free and takes seconds.
• “Which metric is failing, and on mobile or desktop?” “Performance: 34” is not a diagnosis.
• “What changes for my enquiries if we fix it?” Nobody honest can promise a number, but the question surfaces the truth — speed work is worth doing when real users are affected and is otherwise housekeeping.

The bigger caveat

Speed is not why local businesses cannot be found. In this dataset the Speed pillar correlates −0.13 with whether a business appears in search results at all — our own scoring gives it 20% of the weight, which we have published as probably wrong.

Fast is worth being for the people already on your site. It is not what gets them there. If the actual problem is that nobody can find you, that has a different answer.

Businesses in the index can see both figures side by side on their own scorecard, under “Lab speed against real users”, along with which metric is failing.