Keyword monitoring that catches pages that load but are broken.
A page can load cleanly and still be broken — an error message sitting where your content should be. Uptimia reads the page on every check, requiring the words that must be there and flagging the error strings that must never appear, and alerts you the moment either rule breaks.
Monitoring Logs
every 30 s · rotating locations
Toronto14:32:18251 msDown
Frankfurt14:32:16409 msDown
London14:32:14368 msDown
New York14:32:11211 msDownFrankfurt14:31:41405 msUpLondon14:31:11371 msUpNew York14:30:41208 msUpResponse Time
avg 560 ms · last 7 daysThe page that loaded fine — with nothing on it
Every dashboard stayed green. The product page answered instantly, exactly as it always had — and served a catalog with nothing in it.
What a keyword rule can catch
Require the words that only appear when a page is healthy, or forbid the ones that only appear when it’s broken.
Confirmation, alerting and proof
Confirmed before it pages you
One region loading a half-built page never pages you. Uptimia re-reads the body from other checkpoints — only when they all report the same missing or forbidden text does the incident open. One clean read closes it.
- Confirmed from up to 3 regions — a slow half-render can’t page you
- Several keyword rules per monitor — require some words, forbid others
- Runs on the same check — no extra request, same 30-second schedule
agree
The right person hears about it
A confirmed content failure goes straight to Slack, SMS, PagerDuty or any of the 12 channels. Unanswered, the ladder climbs to the next person on your schedule; when the page reads clean again, a recovery notice closes it out.
- Escalation ladders with one-click acknowledgement — no login mid-firefight
- Maintenance windows — a deploy that rewrites copy never pages anyone
- Recovery notice with the failure duration when the page reads clean again
The page it actually served
Each incident keeps the response body from the failing check — so you fix the page you saw, not the one that works when you reload it.
- The response body and headers — the exact bytes the check read, kept
- The rules on the monitor — listed with the incident, beside the body they were read against
- Hand it to anyone — export as PDF or HTML, or share a public incident link
HTTP/1.1 200 OK
<div id="catalog">
<p>No products found.</p>
required "Add to cart" required "Proceed to pay" forbidden "database error"
Read from where your customers are
A CDN or edge cache can serve a stale error page in one region while origin looks perfect. Reading the body from 171+ locations catches the broken page only some visitors see.
- Checks as often as every 30 seconds — a bad deploy can’t hide for long
- The body read per checkpoint — every check is stamped with the region that read it
- At least 6 locations read every monitor, always
A page that loads can still be broken.Read what it sent.
Keyword rules on every uptime check, every region, every alert channel — free for 30 days, and none of it is a paid add-on.
How keyword monitoring works
Add it to any uptime monitor in under a minute — no extra check, no code.
Add the words that matter
On any HTTP monitor, list the text that must appear and the text that must not.
Choose who hears about it
Choose the people and the channels — ladders and quiet hours are optional.
Get paged when the page reads wrong
A confirmed keyword failure opens one incident and alerts every chosen channel.
Also included
Full REST API
Set keyword rules from CI or scripts — required and forbidden text, per monitor.
Transaction monitoring
For text that only renders after JavaScript runs — checked in a real browser.
Scheduled reports
Uptime and incident reports, branded with your logo and colours.
Public status pages
Tell customers what is happening — and what you have already fixed.
Maintenance windows
A content change during a deploy never pages anyone.
Every monitor type in one account
Keyword checks sit beside uptime, SSL, transaction, DNS and heartbeat monitors — same contacts, groups and roles.
Alerts where your team already works
One content failure, delivered everywhere — the same contacts and channels you already use.
Microsoft Teams12 channels, one contact list — a keyword failure reaches them all.
Browse the full integrations directory →What is keyword monitoring?
Keyword monitoring is a content check that confirms a page’s response body contains the text it should — or none of the text it shouldn’t — on every uptime check. So a page that returns 200 OK while showing a database error, an empty grid or a maintenance notice is caught as down, not counted as up.
How does a keyword check run?
Each check fetches the page and scans the response body for your required and forbidden text, alongside the status code and response time.
When the words are wrong
confirmed 3/3 → content incident opens
What can a keyword rule check?
Two rules cover most broken-but-200 pages: require text that only appears when the page is right, or forbid text that only appears when it’s wrong.
Guide: choosing keywords that don’t false-alarm →| Rule | You require | Catches |
|---|---|---|
| must exist | “Add to cart” | catalog didn’t render |
| must exist | your price string | pricing widget broke |
| must not exist | “database error” | DB connection lost |
| must not exist | “Exception” | stack trace leaked |
| must not exist | “Be right back” | stray maintenance page |
Keyword monitoring FAQ
01What is keyword monitoring?+
02How does a keyword check work?+
03What’s the difference between “must exist” and “must not exist”?+
04Does it work on pages rendered by JavaScript?+
05Is keyword monitoring a separate monitor?+
06Can I check for more than one keyword?+
07Can I use it on a JSON API?+
08What do I see when a keyword check fails?+
09How do you avoid false alarms from a slow page?+
10Is there a free plan?+
11Will a page that only breaks in one region page me?+
Check what your pages actually say.
Add a keyword rule to any monitor — be the first to know when a page reads wrong.