How fast are Czech small-business websites? We measured 231 of them (2026)
We measured the websites of 231 Czech small businesses that Google shows for searches like “car service Brno”. The median mobile speed score is 58 out of 100, and 1% of the sites meet Google's limit for loading the main content.

In short
- The median mobile Lighthouse performance score of a Czech small-business website is 58 out of 100. Only 4% reach the green threshold of 90.
- In the mobile lab test the main content appears after 8.7 s (median). Google's 2.5 s limit is met by 1.3% of the sites.
- Real visitors see it faster: 1.9 s on the half of the sites where Google collects field data. Even there, one in four is rated slow.
- One in five homepages has no H1 heading, one in six has no meta description and almost half have no structured data.
Czech small-business websites are slow on mobile. We measured 231 sites across 12 trades and 5 cities, and the median Lighthouse performance score is 58 out of 100. In the mobile test the main content takes 8.7 seconds to appear, while Google treats a page as fast below 2.5 seconds. Three of the 224 sites we could measure made it.
These are our own numbers. Nobody else has published a sample like this for Czech business websites, so below you will find exactly how we measured, and where the method is weak.
How did we pick and measure the sites?
We wanted the sites a customer actually sees when looking for a local service, so the sample comes straight from Google results:
- 12 trades: car service, dentist, lawyer, accountant, hair salon, restaurant, guesthouse, builder, plumber, gym, florist and real estate agency.
- 5 cities: Prague, Brno, Ostrava, Plzeň and Olomouc.
- A “trade city” query in Czech, for example “autoservis Brno”. From the results we took the first five websites of the businesses themselves. Directories, comparison sites, social networks and the head offices of national chains were skipped.
That gave 231 homepages. For restaurants and guesthouses the first page of results is taken over by directories and booking portals, so those trades have fewer business sites in the sample.
On 26 September 2026 every site was measured twice:
- PageSpeed Insights in mobile mode. This is Google’s Lighthouse, loading the page as a mid-range phone on a slower mobile connection. It returns a performance score from 0 to 100 and timings such as Largest Contentful Paint, the moment the main content shows up. Where Google has enough data from real visitors, it adds that too.
- Our quick check, the same one as in our free website audit. It fetches the homepage once and reads what the page says about itself: server response, headings, descriptions, structured data and images.
Lighthouse loaded 224 sites, the quick check 225. The rest did not respond or turned us away. We do not name individual businesses; we compare the whole.
How fast are Czech business websites on mobile?
| What we measured | Median | What good looks like |
|---|---|---|
Lighthouse performance score (mobile) |
58 / 100 |
90 or more |
Main content on screen (LCP, lab) |
8.7 s |
under 2.5 s |
First content on screen (FCP, lab) |
3.9 s |
under 1.8 s |
Main thread blocking (TBT) |
167 ms |
under 200 ms |
Page weight |
2.4 MB |
under 1.6 MB |
Server response (TTFB) |
0.46 s |
under 0.6 s |
The spread tells more than the median:
- 4% of sites score 90 or more. 28% are below 50, in the red zone.
- 1.3% show the main content within 2.5 s. 89% need more than 4 s, and 43% more than 10 s.
- A quarter of the sites weigh over 4.3 MB, and one in five over 5 MB.
- Server response is the healthiest number: 58% of sites answer within 0.6 s. The problem is almost always the page itself, and hosting comes a distant second.
Layout stability (CLS) does best of all: on 84% of sites nothing jumps around while loading.
Why does the lab say 8.7 s and real visitors 1.9 s?
For 112 sites, half of the sample, Google has field data from real Chrome users. There the median time to the main content is 1.9 seconds, and 73% of those sites fit under 2.5 s.
The gap to the lab test is big, for three reasons:
- The test deliberately simulates worse conditions: an older phone and a slow mobile connection. It shows how the site behaves for the people who have it hardest.
- Many real visitors are on wi-fi or fast 4G and 5G, and have part of the page cached from an earlier visit.
- Google only has field data for busier sites. Smaller sites, which tend to be slower too, are missing from it.
Even by real visitors’ numbers, Google rates 24% of those sites as slow and another 36% as average. Fewer than half are fast.
What this means for you is simple. The lab shows what someone in the countryside, on a train or in the metro goes through. When your site takes eight seconds for them, they close it and call the competitor whose number sits right there in Google Maps.
Which trades have the slowest websites?
| Trade | Sites | Score (median) | Main content (median) |
|---|---|---|---|
Accountant |
14 |
62 |
8.1 s |
Lawyer |
20 |
61.5 |
8.4 s |
Plumber |
20 |
61.5 |
6.8 s |
Dentist |
25 |
60 |
7.7 s |
Hair salon |
18 |
59 |
7.3 s |
Car service |
22 |
58 |
10.0 s |
Florist |
19 |
58 |
8.5 s |
Gym |
22 |
56.5 |
10.8 s |
Builder |
21 |
54 |
7.5 s |
Guesthouse |
11 |
52 |
14.7 s |
Real estate agency |
20 |
52 |
9.0 s |
Restaurants are left out: search results gave us only 6 of their own sites, too few to conclude anything.
The differences between trades are small. One pattern stands out, though: the trades that sell with photos do worst. Guesthouses, estate agents, gyms and car services put large galleries, sliders and background videos on the homepage. Every full-size photo straight from the camera costs an extra second on mobile. Accountants and lawyers have simpler pages and come out better, though they are not fast either.
Does the city matter?
It does not. The median score is 59 in Prague, 58 in Brno, 61 in Ostrava, 58 in Plzeň and 58 in Olomouc. Prague businesses do not have faster sites than the rest, even though they often pay more for them. Speed depends on how a site is built, and that has little to do with the city or the budget.
What do Czech business websites most often get wrong?
The quick check reads the basics that Google and AI assistants build a search result from:
| What we check | Missing or wrong |
|---|---|
H1 heading on the homepage |
missing on 20%, more than one H1 on another 22% |
Meta description |
missing on 16%, too short or too long on another 19% |
Share preview (Open Graph) |
missing entirely on 31% |
Structured data (schema.org) |
missing on 46% |
Canonical URL |
missing on 27% |
Sitemap (sitemap.xml) |
missing on 14% |
Image dimensions |
missing on at least some images on 43% of sites |
Scripts on the page |
more than 30 on 39% of sites; a sensible limit is 12 |
The good news: nearly every site runs on HTTPS (99.6%) and has a mobile layout (97%). That is a given now. The median Lighthouse SEO score is 100, but it only checks a technical minimum. Headings, descriptions and structured data weigh little in it, and they are exactly what decides how a site looks in search results.
What does this mean for your website?
If your site beats the median, you are in the better half of the market. That still does not make it fast. Google sets the bar far above what the average Czech business does today. That is the opportunity: you only need to be faster than the competitors that show up for your trade and city.
Here is an afternoon’s plan:
- Measure. Enter your address in our free website audit. You get the same numbers as in this study and can compare them with the tables above.
- Shrink the five biggest images on the homepage. For photo-heavy trades this is the fastest win. A photo for mobile does not need more than 1,600 px in width.
- Go through the scripts. Chats, pixels, maps and widgets nobody uses can go. The rest should load after the content.
- Add the H1, the description and structured data. It is an hour of work, and it shows in the search results.
Why a sub-second site pays off, with our own numbers and the research behind them, is in What a sub-second load time actually does to conversion. When it turns out your site is held together by plugins and a theme, only a rebuild helps. How we do it is on the website development page.
Frequently asked questions about Czech website speed
What is the average speed of a Czech business website?
According to our September 2026 measurement of 231 sites, the median mobile performance score is 58 out of 100. The main content appears after 8.7 seconds in the mobile lab test and after 1.9 seconds for real visitors.
How many websites meet Google’s recommendation?
The 2.5-second limit for the main content in the mobile test is met by 1.3% of the sites. By real visitors’ data it is met by 73% of the sites that have such data, which is only half of the sample.
Are websites in Prague better than elsewhere?
No. The median score differs by three points at most between Prague, Brno, Ostrava, Plzeň and Olomouc.
Can I measure my own website the same way?
Yes. Our free website audit uses the same two tools: PageSpeed Insights in mobile mode and our quick check of the homepage.
Bottom line
A typical Czech business website is slower on mobile than it should be, and people on a weaker signal pay for it. Run your address through our free website audit and compare it with the medians from this study.
Need a site that can pull this off?
Websites & e-shops

