The difference between an agency that manages 200 sites and one that merely hosts 200 sites is whether anyone finds out about problems before the client does. Monitoring is how you earn the care-plan fee every month, and at fleet scale it has to be designed, not bolted on one ping checker at a time. Here is what to watch, how to route the alerts, and what the dashboard should show.
The five things to monitor
1. Uptime and availability
The baseline, but do it properly:
- Check every site at least every minute from more than one region. Single-region checks miss regional CDN and DNS failures.
- Check a real page, not just the homepage returning 200. A keyword check (“Contact us” appears in the body) catches white screens, database errors, and maintenance-mode pages that still return 200.
- For stores and lead-gen sites, add a transactional check: load the checkout or submit a test form on a schedule.
- Monitor SSL expiry and domain expiry. Both cause outages that look like hacks to clients and are entirely preventable.
2. Performance
Uptime says the site loaded. Performance says whether it loaded well enough to keep visitors. Track Core Web Vitals (LCP, INP, CLS) per site with both synthetic checks and field data, and watch trend, not snapshot. A site that went from 1.8s to 3.1s LCP over three weeks has a cause, usually an update or an upload, and the timeline tells you which.
3. Security
Three signals, checked daily or continuously:
- Vulnerable components: cross-reference each site’s plugin and theme versions against published vulnerability feeds.
- Integrity: core file checksums, unexpected new admin users, changed
wp-config.php, new files inuploadsthat are not media. - Blacklist and malware status: Google Safe Browsing and malware scans. A blacklisted client site is a reputational emergency.
4. SEO and search rank
This is the one most maintenance setups skip, and it is the one clients care most about. Track for each site:
- Keyword positions for a handful of target terms (five to ten per site is plenty for a care plan).
- Indexing status and sudden drops in indexed pages.
- Broken links and 404s, which creep in after redesigns and migrations.
- Accidental
noindexafter a staging push. This happens more than anyone admits.
Rank tracking turns a maintenance report into a growth report. WP Maintain’s MainRank handles this alongside monitoring so you are not stitching in a separate SEO tool.
5. Content and visual changes
Capture a screenshot and a DOM snapshot of key pages on a schedule and diff them. This catches defacements, broken layouts after updates, missing hero images, and the client’s intern who deleted the pricing table. It is also your evidence when a client asks “when did this change?”
Alert routing: the part that makes or breaks it
A fleet that generates 400 alerts a day is a fleet nobody watches. Design the routing before you turn the checks on.
| Severity | Examples | Route | Response |
|---|---|---|---|
| Critical | Site down 3+ minutes, malware detected, blacklisted, checkout failing | Phone/SMS to on-call, plus ticket | Immediate |
| High | SSL expiring in 7 days, critical CVE in fleet, LCP failing on a critical site | Slack channel plus ticket | Same day |
| Medium | Visual diff above threshold, rank drop over 5 positions, 404 spike | Ticket only | Next business day |
| Low | Minor CWV drift, plugin update available | Weekly digest | Scheduled |
Rules that keep this sane:
- Confirm before alerting. Require two consecutive failed checks from two regions before a critical page. False positives destroy trust in the system within a week.
- Alert on change, not state. One “site down” alert and one “site recovered” alert, not a page every minute.
- Group by client. If a host has an outage, you want one alert saying “12 sites on Host X down,” not 12 pages.
- Maintenance windows. Suppress alerts during your scheduled update window and when a tech marks a site as being worked on.
- Tier the on-call. Critical sites get a human at night. Brochure sites can wait until morning, and the SLA you sold should say so.
Dashboards: one for you, one for the client
The operations dashboard (internal) answers three questions at a glance: What is broken right now? What is trending the wrong way? What is due (expiring certs, scheduled updates, overdue tickets)? Sort by severity, filter by client and host, and make it the first tab your team opens.
The client dashboard and monthly report (external, white-labeled) shows uptime percentage, updates applied, security scans passed, performance scores, and keyword movement. Keep it to one page. The purpose is to make the invoice feel obvious. A client who sees “99.98% uptime, 17 updates, 0 vulnerabilities, 3 keywords up” each month does not question the retainer.
Avoid the tool sprawl
A common agency setup is an uptime SaaS, a separate security scanner, a rank tracker, a screenshot service, and a per-site plugin for updates, all with different logins and no shared view. It works until it does not, and it never produces a clean client report.
Consolidation is the practical fix. WP Maintain was built for agencies and hosts running WP Care at scale: MainMonitor covers uptime, performance, security and change detection; MainRank handles keyword tracking; the same platform runs updates, backups and patching; and the client report comes out of one place. 1,000+ agencies and web hosts partner with Seahawk on this because the alternative is six tabs and a spreadsheet.
Rollout checklist
- Inventory the fleet and tag each site with a tier (critical, standard, low-touch).
- Turn on uptime with keyword checks for everything; transactional checks for critical sites.
- Add performance, security and visual monitoring fleet-wide.
- Add rank tracking for care-plan clients with target keywords on file.
- Configure routing by severity and tier; test the on-call path at 2 p.m. before trusting it at 2 a.m.
- Set a maintenance window and verify alerts suppress during it.
- Send the first monthly report and ask two clients what they would want to see added.
Next step
Monitoring is the proof behind your care plan. WP Maintain gives you MainMonitor for uptime, performance, security and change detection, MainRank for SEO, and a white-label report that ties it all together, across every site you manage. Connect your fleet and see what has been quietly going wrong.