Test Web Hosting Speed Before You Pay: A 1-Hour Method
Quick answer: Test a real customer site on the host from your customers' region, focusing on server response time (TTFB). Then buy the shortest plan with a refund window, copy your site there, and run the same tests before moving your domain.↗ Share on X
You can test a web host's speed before you pay by doing three things: measure a site the host already runs, check how fast its server answers from where your customers live, and then buy the cheapest plan with a refund window and test your own copy of your site during that window. The number that matters most is server response time, often called TTFB (time to first byte). If the host's own pages and its customer sites answer slowly from your region, your site will too. Below is a step-by-step plan you can finish in about an hour, plus a simple scorecard to compare two or three hosts side by side.
Why can't I trust the speed claims on the hosting page?
Cheap Web Hosting Renewal Rates: How to Find the Real Cost →
Where Is Your Web Server? Check It and Fix Slow Load Times →
Where Is Your Web Host's Server? Find It and Test Its Speed →Hosting sales pages show their best case: an empty test site, a server close to the test tool, and no traffic. Your site will be different. It will have images, plugins, a contact form, and visitors from your city or country.
Speed also depends on things the ad doesn't mention:
- Where the server is. A server in Germany will feel slow to customers in Texas.
- How crowded the server is. On cheap shared plans, many sites share one machine. If a neighbor gets a traffic spike, you can feel it.
- The software stack. Things like the web server type, PHP version and built-in caching change how fast pages are built.
- Your own site. A heavy theme or huge images will be slow on any host.
The goal of testing is to separate the host's part from your site's part, so you know who to blame later.
Smart software picks in your inbox
What numbers should I look at?
You don't need to understand every metric. Focus on these four:
| Metric | What it means in plain words | Rough target for a small business site |
|---|---|---|
| TTFB (time to first byte) | How long the server takes to start sending the page | Under about 0.6 seconds from your main customer region; under 0.2 is excellent |
| LCP (Largest Contentful Paint) | When the biggest thing on screen (usually the main image or headline) appears | Google calls 2.5 seconds or less "good" |
| Fully loaded time | When everything has finished loading | Useful to compare hosts, but less important than the two above |
| Consistency | Whether results stay similar across several tests and times of day | Small swings are fine; big jumps are a warning sign |
TTFB is the number most controlled by the host. LCP is shared between the host and your site's design. That's why TTFB is the best "before you buy" test.
Step 1: Find a real site running on the host
Shared vs VPS Hosting: 8 Signs Your Site Has Outgrown Shared →
Move a Non-WordPress Site to a New Host Without Downtime →
Slow Website? 5 Fixes to Try Before You Switch Your Host →Testing the host's own homepage tells you a little, but it's often on a special fast setup. A better test is a normal customer site.
How to find one:
1. Check the host's showcase or testimonials page. Many list customer sites.
2. Search for the host's name in site footers. Some customers leave "Hosted by..." in the footer.
3. Ask in the host's pre-sales chat. Say: "Can you share a demo site on the same plan I'm considering?" A host that refuses or dodges this is telling you something.
4. Confirm the site is really on that host. Free "who is hosting this site" lookup tools show the hosting company based on the server's address. Use one to double-check.
Try to find at least two sites, ideally on the same plan level you plan to buy.
Step 2: Run speed tests from your customers' location
Use free tools that let you pick the test location:
- WebPageTest lets you choose a city and a connection speed. It shows TTFB clearly near the top of the results.
- Google PageSpeed Insights shows LCP and other Core Web Vitals. If the site gets enough visitors, it also shows real-user data from Chrome users, which is more useful than lab tests.
- GTmetrix gives a waterfall chart that shows how long each file takes.
Do it like this:
1. Pick the test location closest to most of your customers.
2. Test each site three times and write down the middle result, not the best one.
3. Repeat at a different time of day, for example morning and evening. Shared servers are often slower at busy hours.
4. Write the numbers in a simple table: host name, site tested, TTFB, LCP, time of day.
If one host's customer sites show TTFB well above one second in your region, cross it off before spending any money.
Step 3: Ask pre-sales the right questions
A five-minute chat can save you a bad year. Ask these and write down the answers:
1. "Which data center locations can I choose?" Pick the one nearest your customers.
2. "Is server-level caching included on this plan?" Caching saves ready-made copies of pages so the server doesn't rebuild them every time.
3. "Do you include a CDN?" A CDN (content delivery network) stores copies of your images and files in many cities, so they load from somewhere close to the visitor.
4. "How many CPU cores and how much memory does this plan get?" Some cheap plans are very limited. If the host won't say, be careful.
5. "How long is the refund window, and are there exceptions?" Some hosts exclude domain fees or setup fees from refunds.
6. "What happens to my price at renewal?" The first-term price is often much lower than the renewal price.
Step 4: Test your own site during the refund window
This is the most accurate test, because it uses your real site.
1. Buy monthly or the shortest term with a clear refund period. Don't register your domain through the host yet, since domain fees are often not refundable.
2. Copy your site to the new host. For WordPress, a migration plugin can do this. Many hosts also offer free migration.
3. Use the temporary URL or a preview link the host gives you, so you can test without moving your domain.
4. Run the same tests from Step 2 on your copied site: three tests, two times of day, same test location.
5. Compare with your current host, if you have one. Same site, same tools, same location. Now the only thing that changed is the host.
6. Do a small load check. Click around the admin area, save a few pages and open the site on your phone over mobile data. If the dashboard feels sticky, that's a sign the server is weak.
Give yourself a reminder in your calendar two days before the refund window ends, so you don't miss it.
How do I compare two or three hosts fairly?
Use a simple scorecard. Give each item a score from 1 to 5:
| Check | Host A | Host B | Host C |
|---|---|---|---|
| TTFB from your region (customer sites) | |||
| TTFB from your region (your copied site) | |||
| LCP on your copied site | |||
| Consistency (morning vs evening) | |||
| Data center near your customers | |||
| Caching and CDN included | |||
| Renewal price you can live with | |||
| Support answered clearly and quickly |
Add up the scores. Speed matters, but a host that is slightly faster and twice the price at renewal may not be the best deal for a small business.
What if my site is slow on every host?
If your copied site has a good TTFB but a slow LCP on all hosts, the problem is probably the site itself. Common fixes:
- Shrink your images. Photos straight from a phone can be several megabytes each. Resize them and save them in a modern format like WebP.
- Remove plugins you don't use. Each plugin can add work for the server and extra files for the browser.
- Use a lighter theme. Some all-in-one themes load a lot of code on every page.
- Turn on caching. If the host doesn't do it at the server level, use a caching plugin.
- Limit fancy extras like big sliders, auto-playing video and many web fonts.
In this case, switching hosts won't fix much. Fixing the site will.
Your next step
Pick two or three hosts you're considering and do this today:
1. Find one customer site for each host and confirm it's really on that host.
2. Run three WebPageTest tests from your customers' city and write down the middle TTFB.
3. Cross off any host that's clearly slow in your region.
4. Ask the remaining hosts the six pre-sales questions.
5. Buy the best one on the shortest term with a refund window, copy your site, and test it before you move your domain.
An hour of testing now is much cheaper than moving your site again six months from now.
FAQ
What is a good TTFB for a small business website?
Under about 0.6 seconds from your main customer region is reasonable, and under 0.2 seconds is excellent. Test several times and use the middle result.
Which free tools can test hosting speed?
WebPageTest lets you pick a test city and shows TTFB clearly. Google PageSpeed Insights shows LCP and real-user data. GTmetrix shows a waterfall of each file's load time.
Should I register my domain with the host while testing?
Not yet. Domain fees are often excluded from refunds. Use the host's temporary URL or preview link to test, and move the domain only after you decide to stay.