When do you need a browser check instead of a URL check?
A URL monitor tells you whether an address responded. A browser check uses a script to open the site, do something such as choosing an appointment and look for the result a visitor should see.
| Check | Example outcome |
|---|---|
| URL monitor | The public booking page loads and responds. |
| API check | The request that looks up appointment times returns the expected answer. |
| Browser check | A visitor can choose a sample time and see that time in a confirmation. |
How do you check a booking without creating a real one?
Use a separate test copy with its own sample records and test email or payment connections. Ask for one short script that opens the booking page, chooses one sample time and checks for the confirmation.
Checkly Browser Checks run scripts made with Playwright, a tool that controls a browser. Start with an isolated app copy and a dedicated test account if sign-in is required. Keep its password in the tool’s secret settings. Avoid putting customer data in scripts or responses that may show up in results.
Before you put the check on a schedule, run it in that disposable environment and review both a passing result and a failure you set up deliberately. A repeated check needs fresh sample data or a safe way to reuse it. If you cannot verify where an action saves data or sends a message, start with a check that reads a page.
Where should a failed run send a message?
Attach an alert channel to the check or its group and decide which failures need a message. The alert can point you to the run so you can see what failed.
Start with the channel someone will actually see. Check which channels, intervals and public locations your plan supports. If Checkly tries again before alerting, look at the first result and the retry. A passing retry shows that the later check worked. Check whether customers were affected before closing the issue.
What counts as a run on your bill?
Uptime monitors count as saved monitors. The scheduled API and browser checks have monthly run allowances. Check how often the script runs and how many locations run it at the same time.
Hobby includes 10 uptime monitors, 10,000 API runs and 1,000 browser runs per month. Checkly says it stops running Hobby synthetic checks when your run credits are exhausted. The public Starter price is listed as $24 per month billed annually. Confirm the amount and currency shown for your account. Starter and Team can keep running at overage rates after prepaid runs are used.
A browser check scheduled every hour from one location runs 720 times over 30 days before retries. If your plan runs that check from two locations in parallel, the example uses 1,440 runs. Retries count as extra runs.
| What you schedule | How Checkly counts it |
|---|---|
| Ordinary browser check | Each script execution counts as one run. |
| Playwright Check Suite | The whole suite uses a run for each started 30 seconds of execution. A 45-second suite uses two runs. |
| API or multistep check | A single API check counts once per execution. In a multistep check, each request counts. |
What should you keep outside the Checkly account?
Keep a copy of the checks and expected results in your project files. Checkly can generate code for supported dashboard-managed resources, and its API can retrieve check results separately.
Keep the original Playwright project for a Check Suite. Importing cannot recover its source from the uploaded bundle. Before changing how checks are managed, review the import plan and avoid deploying duplicates.
Give teammates the account role they need. Someone who only reads results should not have to edit checks. Keep test credentials out of exported files and confirm which alert connections you need to set up again.