# How often should a website be checked? URL: https://ergostrat.com/en/monitoring/check-frequency.html Language: [Български](https://ergostrat.com/monitoring/check-frequency.html) · [English](https://ergostrat.com/en/monitoring/check-frequency.html) · [Français](https://ergostrat.com/fr/monitoring/check-frequency.html) · [Deutsch](https://ergostrat.com/de/monitoring/check-frequency.html) Ask what is being checked before asking how often. A responding homepage, a contact form that delivers a message and an automation that completes its run are different results. The agreement should give each check a frequency, an alert recipient and a response deadline. “Monitoring” alone leaves those questions open. ## What do published provider intervals tell you? WP Buffs describes uptime checks hundreds of times a day. That says how often it checks availability; it does not by itself prove a contact form delivers messages. SiteCare’s Diamond plan describes up to 5 user flows tested, typically hourly or every few hours depending on the flow. That is a functional-check schedule, rather than one universal interval for every website task. These are examples from the providers’ own pages, not a survey of the market or a recommendation to buy a particular plan. Use them to ask a more precise question: which pages, actions and expected results does your quotation cover? ## What can an hourly check miss? If a fault begins just after a successful check, it can remain undetected until a later check. A short failure between checks may never appear in that test’s results. More frequent polling can shorten the gap, but it cannot guarantee that a customer will never encounter the fault first. The test also needs to exercise the thing you care about. A homepage response does not establish that a form reaches the intended inbox. Agree a controlled test submission, the expected destination and how success is confirmed. Do not treat a thank-you message alone as proof of delivery. Track certificate and domain expiry separately. Their dates can be recorded in advance; they do not need to fail before someone acts. Let’s Encrypt ended its expiration notification emails on 4 June 2025. Check who monitors renewal and which notices you actually receive instead of assuming the certificate issuer will send a reminder. ## Separate checks, alerts and human response Write down when the test runs, how a failure is reported, who receives that report and when a person starts work. A polling interval is not a repair deadline. A service can run automatic checks outside the hours in which a person responds. Ask for separate automatic-monitoring and human-response hours in writing. Do not read “24/7 monitoring” as a promise of 24/7 human support. At ErgoStrat, the website is checked every hour and automations are checked on every run. If something fails, we start work and write to you by the end of the next working day (Sofia time) at the latest. That response is not a commitment to finish the repair by then; the time needed depends on the cause. ## Give each kind of check its own schedule - Availability: specify the page or endpoint, the test interval and what response is acceptable. - Key journeys: specify the form, sign-in or other agreed steps and how the result is verified. - Automations: check each run, and separately define how an expected run that never starts is detected. - Expiry dates: name the certificate and domain owner, the reminder point and the renewal responsibility. - Updates and backups: agree these as work of their own. A monitoring interval does not establish an update schedule or a tested backup. ErgoStrat’s Monitoring and Care service does not include version updates and backups as routine work; they receive a separate proposal. A monthly report summarises checks, failures and fixes. The reporting interval is different from the frequency of the checks recorded in it. ## Questions ### Can I compare providers from the word “monitoring” alone? No. Ask for the covered checks, intervals, alert routing and response hours in writing. A service description that leaves an interval unstated gives you a question to resolve, rather than a number to assume. ### Is certificate monitoring the same as an uptime check? No. Availability concerns the result of a request now. Certificate and domain expiry concern known dates and successful renewal. Specify both in the inventory, with an owner for each follow-up. ### Is checking every automation run enough? It tells you about executions that actually happen if the checks cover them. A missed scheduled start needs separate detection. Agree the expected timing and alert rule; a daily task does not become continuously verified merely because its errors create alerts. ### Does a monthly report mean the site is checked monthly? No. Checks and reporting have separate schedules. Ask for the actual check times and results in the report, alongside failures, action taken and any work still open. ## Sources - [WP Buffs: uptime monitoring](https://wpbuffs.com/plans/) - [SiteCare: user journey monitoring](https://sitecare.com/plans/) - [Let’s Encrypt: ending expiration notification emails](https://letsencrypt.org/2025/01/22/ending-expiration-emails) ## Related reading - [How much does website maintenance cost per month?](https://ergostrat.com/en/monitoring/maintenance-cost.html) - [How do you choose an automation partner?](https://ergostrat.com/en/automations/choose-automation-partner.html) ## Ask for the interval and the response deadline in writing With us, the website check runs every hour, and automations are checked on every run, written into the inventory before we start. [Ask about monitoring](mailto:contact@ergostrat.com?subject=Question%20after%20the%20Check%20Frequency%20article)