Uptime Monitoring for Agencies: Client Status Pages and Monthly Reports
Short answer: when a client's website goes down, the agency hears about it first β usually from the client. Monitoring every client site turns that around: you know before they do, you can show them a status page instead of answering calls, and at the end of the month you can send an availability report that proves what the maintenance retainer is for. This guide covers what to monitor for each client, how to set up client status pages, and how to measure availability in a way that holds up in a contract.
Product details verified against AlertSleep on 14 October 2026.
Why agencies need monitoring more than anyone
- You are the first call. Whoever built or hosts the site gets blamed for every outage, including the ones caused by the client's payment provider or DNS.
- A care plan needs evidence. "We kept your site running" is a promise. A monthly report with an availability figure and a list of incidents is proof β and proof is what gets a retainer renewed.
- Problems hide between visits. An expired certificate, a lapsed domain or a backup job that stopped weeks ago does not show up until someone needs it.
What to monitor for each client
| Check | What it catches | Alert |
|---|---|---|
| Website or API | The site is down, returns an error or is too slow | Email; a phone call for the few sites where downtime costs the client money |
| Keyword on a key page | The page loads but shows an error instead of the checkout or the login form | Email or phone call |
| SSL certificate | A certificate about to expire, a broken chain, the wrong hostname | |
| Domain expiry | A registration about to lapse β often on an account the client controls | Email, weeks ahead |
| Cron job heartbeat | Backups, imports or syncs that stopped running |
Each of these has its own guide: SSL certificate validity, domain expiration monitoring, cron job monitoring and phone call alerts.
Name every monitor after the client and the thing it checks β "Kowalski Shop Β· checkout" β so an alert, a report line or a 3 a.m. call tells you at once whose site it is.
A status page for every client
A public status page answers "is it down?" before the client picks up the phone, and makes your agency look like it has the situation under control. In AlertSleep each status page shows the monitors you choose, so every client gets a page with their own services and nobody else's:
- the client's logo and a page language of its own β one of six, independent of yours;
- current status and uptime history of the selected monitors;
- incident updates you write by hand β "We are aware of the problem, the hosting provider is working on it";
- planned maintenance shown in advance;
- visitors can subscribe by email to updates, so the client's own customers are informed without you writing a single message to them.
How many status pages and monitors an account can have depends on the plan; see pricing.
Measuring availability for a contract
If the contract promises 99.9% availability, the number has to be measured the same way every month and has to survive a disagreement. Two things matter.
Measure time, not checks. The share of successful checks is a useful health indicator, but it is not a statement of how long the site was down. A contract is about time: the minutes during which the client's customers could not use the site. AlertSleep's availability report is therefore calculated from incidents β when each outage started and ended β clipped to the reported period. That is the figure to put in front of a client. The uptime percentage in the dashboard counts successful checks and can differ slightly; it is the right figure for day-to-day health, not for a contract.
Agree on what counts. Most service agreements exclude announced maintenance. In AlertSleep a maintenance window freezes alerting and opens no incident, so planned work done inside a window is not counted as downtime. Announce it on the client's status page, and it shows up there too.
What the common targets allow in a 30-day month:
| Availability | Downtime allowed in 30 days |
|---|---|
| 99.5% | 3 hours 36 minutes |
| 99.9% | 43 minutes 12 seconds |
| 99.95% | 21 minutes 36 seconds |
| 99.99% | 4 minutes 19 seconds |
The free Uptime Calculator works this out for any target and period.
Monthly availability reports in AlertSleep
Under Availability reports in the dashboard you pick a period β last month, this month or a custom range of up to 366 days β and open the report:
- the headline availability and total downtime for the period;
- a table of every monitor with its availability, downtime, number of checks and response times;
- the list of incidents with start, end and duration;
- a daily bar across the whole period.
The report is a page made for printing: print it to PDF from the browser and attach it to the invoice, or download the data as CSV β one row per monitor per day, or the list of incidents. The report opens through a signed link that works without logging in and stays valid for 90 days, so it can be forwarded as it is. It is written in your account's language.
To get it without asking, turn on E-mail me a monthly availability report in your profile. On the first of each month you receive the headline figures for the previous month β availability, number of incidents, total downtime and the most affected service β with the link to the full report.
A report for one client: before opening the report, choose that client's monitors under Monitors and type their name under Client name. Every figure β availability, downtime, incidents, the daily bar and both CSV files β then covers only those monitors, and the client's name appears at the top with your account beneath it. The selection is part of the signed link, so whoever receives it cannot widen it. Leave the selection empty for a report on the whole account.
Turn monitoring into part of the service
- Put it in the offer. "24/7 monitoring, a public status page and a monthly availability report" is easier to sell β and easier to renew β than "maintenance".
- Agree the target and the exclusions in writing: the availability figure, how it is measured (time, from incidents), and that announced maintenance does not count.
- Set up the client's monitors on day one, including SSL, domain and backup jobs β and ask who owns the domain registration.
- Give the client their status page address and tell them to send it to their own customers when something breaks.
- Send the report every month, with a two-line summary: what happened, what you did.
FAQ
Can a client see only their own sites?
Yes. A status page shows only the monitors you add to it, and a report can be narrowed to one client's monitors and carry the client's name. The selection is part of the signed link, so the client cannot widen it to see anyone else.
Does the client need an AlertSleep account to open a report?
No. The report link works without logging in for 90 days.
How far back can a report go?
A single report covers up to 366 days. Daily availability is kept after detailed check data is removed, so a report on an earlier month stays complete.
Why does the report differ from the uptime percentage in the dashboard?
The report measures time from incidents; the dashboard counts successful checks. For a contract, use the report.
Can the status page use the client's own domain?
Not at the moment. The page is served from AlertSleep's status address, with the client's logo and in the client's language.
Share this article
About the Author
AlertSleep Team
Monitoring & DevOps
The AlertSleep team helps developers and ops teams keep their services online and communicate outages professionally.