Uptime monitoring for agencies, and what it leaves out

An agency looking after thirty client sites has the same problem thirty times over, and almost none of it is uptime.

The site is up. The certificate expires on a Sunday in March. The domain was registered by a client's previous developer and renews to an inbox nobody reads. A plugin update expired the CRM key behind the contact form, and the form has been answering 200 and swallowing every lead for eighteen days. A staging robots.txt shipped to production and search traffic is draining a few percent a day, which nobody will notice until the client asks why the phone has gone quiet.

Every one of those answers a perfectly healthy 200 to an uptime check.

What the category actually sells

Tools sold as uptime monitoring are good at the thing they are named after. They ping a URL on a schedule, they tell you when it stops answering, and the better ones watch the certificate expiry while they are there.

Two things follow from that shape, and both cost an agency money.

They charge by the monitor. A client with a website, an API, a mail server and a certificate is one thing to look after and four billable rows. A ten-client agency arrives at a bill for fifty monitors while thinking of itself as having ten things to watch.

The rest of the job is somebody else's product. Mail authentication is a separate subscription from a separate vendor. Security headers are a browser extension somebody remembers to run. Form testing is a calendar reminder that gets skipped. Search visibility is Search Console, checked when there is time. Four subscriptions, four dashboards, four sets of alerts in four formats, and nothing that can tell you the certificate and the mail record broke for the same reason.

The failures with the longest fuse

The failures that cost an agency a client are rarely outages. An outage is loud, everybody sees it, and it is fixed within the hour. What actually loses accounts is the slow one nobody was watching:

• A lapsed domain. Recovery is slow, expensive and occasionally impossible. The registrar lock coming off is the step before a domain is moved, and it is visible days in advance to anybody reading the registry.
• Mail that stopped being trusted. An SPF record that passes ten DNS lookups is a permanent error, and the record looks completely reasonable while it happens. Invoices simply stop arriving.
• A form that stopped delivering. Nothing breaks. The page renders, the submission posts, the response is a 200, and the leads go nowhere.
• A noindex that shipped. Traffic decays over a fortnight rather than falling off a cliff, which is exactly long enough for nobody to connect it to a deploy.

None of those are uptime problems, and all of them are the agency's problem the moment the client notices.

What we do differently

Priced per site rather than per monitor. A site carries its uptime checks, its certificate, its domain registration, its DNS, its mail records, its response headers and its search visibility, and it counts once.

Everything runs on every plan, including the free one. What a bigger plan buys is more sites, faster checks, longer history and more colleagues — never a category of problem you are not told about because you did not pay enough.

And the form is actually submitted. Not requested, not parsed — submitted, from a marked address, a few times a day, so a contact form that has quietly stopped delivering is caught in hours instead of at the end of the month.

See what is checked · See the pricing