Managing Alert Fatigue
Short answer: Too many low-signal alerts is almost always a selector problem or a severity-threshold problem. Fix the root cause instead of ignoring the digest — you'll miss what matters if you tune out.
Why alert fatigue happens
KompWatch captures everything that changes within the element you're tracking. If you set the selector to body, you get navigation tweaks, footer copyright year bumps, cookie-banner text updates, and A/B test variants alongside the pricing change you actually care about.
The fix is almost always one of three things:
- Narrow the CSS selector — scope to the content region, not the full page
- Raise the severity floor — stop receiving LOW-severity changes
- Reduce snapshot frequency — check less often for slow-moving pages
Step 1 — Fix your CSS selector first
This is the highest-leverage change you can make. Open Competitors → [Name] → Edit and update the selector.
| What you're tracking | Good selector |
|---|---|
| Pricing tiers and prices | #pricing, .pricing-table, [data-section="pricing"] |
| Feature list | .features, #features, main .feature-grid |
| Homepage hero / messaging | .hero, header .hero-content |
| Careers / job listings | #jobs, .positions-list, main |
| Changelog entries | .changelog-entries, article |
Avoid: body, div, main on pages with heavy navigation, live chat widgets, or personalization. These generate 5–10× more changes than a scoped selector.
Test your selector before saving: open your browser's DevTools console on the competitor's page and run document.querySelector('YOUR_SELECTOR'). If it returns a small, stable element, you're good.
Step 2 — Raise the confidence threshold
KompWatch's AI assigns a confidence score (0–100%) to every detected change — a measure of how likely the change is genuine vs. noise (A/B test variant, CDN drift, session token). You can raise the minimum confidence required to fire an instant alert:
- Go to Settings → Notifications → Alert Confidence Threshold
- Set to 85% for low-noise environments, or 50% for maximum coverage
Changes below 40% confidence are discarded automatically and never stored. Changes between 40–69% are stored but excluded from instant alerts by default. See AI Confidence Scoring → for the full breakdown.
Step 3 — Set a severity floor in your digest settings
Go to Settings → Digest Preferences and set the minimum severity to include:
| Severity floor | What you receive |
|---|---|
| LOW (default) | Everything — layout tweaks, typo fixes, minor copy changes |
| MEDIUM | Meaningful changes — pricing deltas, feature additions, messaging shifts |
| HIGH | Only significant moves — new tiers, major feature launches, positioning pivots |
| CRITICAL only | Rare, major events — free tier added, acquisition, price restructure |
For most teams, MEDIUM is the right floor. You'll get roughly 80% less volume and lose very little signal.
Step 4 — Reduce snapshot frequency for slow-moving pages
Not every competitor needs hourly snapshots. Pricing pages typically change a few times per quarter. For competitors you watch but don't consider primary threats:
- Pro plan: Change from every 6 hours to daily (contact support to adjust per-competitor cadence)
- Free plan: Already on daily — this is actually appropriate for most monitoring needs
Leave hourly cadence for your top 2–3 direct competitors only.
What NOT to do
Don't unsubscribe from the digest. The digest is the output — the problem is the input. Unsubscribing means you stop seeing legitimate changes.
Don't delete and re-add a competitor. You'll lose change history. Edit the selector in place instead.
Don't set everything to CRITICAL. CRITICAL severity is reserved for events KompWatch detects as structurally significant (pricing model changes, feature category additions, major content removal). You can't manually tag changes as CRITICAL — the AI classifier assigns severity based on what changed.
Diagnosing a noisy competitor
If a specific competitor generates 10+ low-value alerts per week:
- Go to Competitors → [Name] → Change History
- Filter by Severity: LOW
- Look at the
difffield — what's actually changing? Nav text? Cookie consent text? Dynamic ad content? - Update the selector to exclude that region, or switch to a more specific sub-selector
If the page is a JavaScript SPA with a lot of client-side dynamic content, see Monitoring JavaScript / SPA Sites for additional strategies.
Team digest settings
If you're on the Team plan and multiple teammates are getting overwhelmed:
- Have each team member set their own severity floor (digest preferences are per-user)
- Assign ownership of specific competitors to specific people — Sales owns pricing pages, Product owns feature pages
- Use the digest's filter by competitor feature for weekly team reviews instead of real-time alerts
Related articles
- AI Confidence Scoring — How KompWatch Filters Change Noise
- CSS Selectors — How to Scope What KompWatch Tracks
- Understanding Your Digest
- Which Pages to Monitor Per Competitor
- Monitoring JavaScript / SPA Sites
- Change Severity Levels — What LOW / MEDIUM / HIGH Mean
Questions? Email support@kompwatch.com — a team member will respond within 24 hours.