Generic alerts shift the work
A message that only says 'page changed' forces the recipient to open the site, find the edit and decide whether it was expected before any response can begin.
Notification routing with evidence attached
Send the change to the person who can act.
Turn a detected page, API, SEO or accessibility change into an evidence-rich notification, then route it through email, Slack, Discord or a webhook without treating every edit as an emergency.
- Destinations
- 4 channels
- Decision
- Rules + confidence
- History
- Delivery recorded
Alert decision
acme.com/pricing · checked 14:32 UTC
Pricing · High impact
Pro price changed from $12 to $15.
threshold matchedmeaning changed
Delivery routes
Emailpricing@acme.com Sent
Slack#pricing-watch Sent
Webhook/events/changes Sent
A notification without evidence creates a second investigation.
The useful alert is not simply that a URL changed. It names the affected surface, shows the exact before and after state, explains why it qualified and reaches the owner while the context is still fresh.
Every team values a different signal
A price edit belongs with commerce, a canonical change with SEO and an unexpected download link with incident response. One undifferentiated channel quickly becomes background noise.
Routine churn trains people to ignore alerts
Rotating timestamps, recommendation widgets and harmless rewording can exhaust attention long before a contractual or revenue-impacting update arrives.
Qualify the event before it leaves OnChange.
Each delivery can carry the diff, classification and destination context needed for a reviewer or downstream system to make the next decision.
Region and value targeting
Watch a stable CSS or XPath region, tracked number, keyword or stock label so unrelated page furniture does not control the notification.
Human-readable change context
Alerts can include the changed text, section, category, severity and AI summary while keeping a direct path to the complete evidence record.
Semantic-only mode
On Pro and above, meaning-equivalent rewording can be retained silently while changes that alter obligations, availability or value continue to notify.
Owner-specific destinations
Use email for individual review, Slack or Discord for team visibility and webhooks when a qualified change should enter another system.
Structured webhook payloads
Send monitor, change, classification and evidence identifiers to an automation endpoint instead of scraping an email body for machine-readable context.
Delivery status and retry history
Inspect whether a destination accepted the notification and retry failed deliveries from the product workflow when the receiving service recovers.
From change event to owned response.
The delivery path remains traceable from the source monitor through qualification and into each configured destination.
- 01
Detect and preserve
Capture the new state, calculate the relevant diff and create the timestamped change record before notification logic runs.
- 02
Apply alert policy
Evaluate thresholds, keywords, natural-language conditions, confidence and optional semantic meaning against the recorded event.
- 03
Compose for the channel
Build a concise email or team message, or a structured webhook payload, with a link back to the full evidence.
- 04
Track the delivery
Record destination outcomes so the dashboard can distinguish a quiet rule decision from a failed external delivery.
Notification speed begins with check frequency.
A website alert cannot arrive before the monitor checks the source, and a successful OnChange handoff cannot control the final behavior of an external email or chat provider.
- Choose an interval based on consequence and source stability. A policy page may need daily checks while a production price or availability signal may justify seconds or minutes.
- Use selectors and normalization before adding more rules. Removing irrelevant input is more dependable than asking every downstream filter to understand noisy content.
- Webhook recipients should validate requests, return clear status codes and remain idempotent so a retry cannot apply the same external action twice.
- Semantic and AI features support triage, but the exact diff remains the review source when contractual, accessibility or incident decisions carry material risk.
Questions about website change alerts
Before you add the first monitor
- Where can OnChange send website change alerts?
- Every published plan can deliver alerts through email, Slack, Discord and custom webhooks. A monitor can use more than one destination, so a reviewer can receive email while an operational change also enters a team channel or automation endpoint.
- How quickly will I receive a page-change notification?
- Delivery begins after the next scheduled check detects and records a qualifying change. The fastest interval is 30 minutes on Free, one minute on Pro, 30 seconds on Business and 10 seconds on Enterprise. Network and destination-provider latency can add a short delay after detection.
- How does OnChange reduce false or noisy alerts?
- You can target a stable page region, normalize known volatile values, set change thresholds and use keyword or value rules. Pro and higher plans can also enable semantic-only alerts, which preserve wording-only changes in history without notifying as if the meaning changed.
- Can a detected change be recorded without sending an alert?
- Yes. Detection and delivery are separate decisions. A change can remain in the timeline when it falls below a threshold, matches a noise rule, looks like cosmetic re-rendering or is judged meaning-equivalent in semantic mode. That keeps the evidence honest without making every event urgent.
Route one real page change today.
All four alert destinations are included on the Free plan. Start with the page and owner that already depend on manual checking.