A monthly website maintenance report should tell you four things: what was updated, whether your site was backed up and whether anyone checked that the backup actually restores, how the site performed on uptime, speed and security, and what work was done on your requests along with the time it took. If a report doesn't cover those four, you can't tell whether your retainer is buying anything.
A useful report is specific. It names the plugin and the version it moved to, gives the date of the last restore test, lists downtime incidents with timestamps, and puts hours against each task. "All systems running smoothly, site updated and secured" is not a report. It's a sentence anyone could send without logging in.
Updates applied to the CMS, plugins and server
This section should be a list, not a summary. For each item you should see:
- The component: WordPress core, a named plugin like WooCommerce or Yoast SEO, the theme, or a server package like PHP.
- The version change: from 6.4.3 to 6.5.2, for example.
- The date it was applied.
- Whether it was tested first: on a staging copy, or straight on the live site.
- Anything that broke and how it was fixed.
The report should also list updates that were held back, and why. Sometimes holding back is the right call. A major WooCommerce release might conflict with your payment gateway plugin, so waiting a week for a patch is sensible. Just say so in writing.
Server-side updates often get skipped because the client never sees them. Ask about the PHP version your site runs on. If it's a version that no longer gets security fixes, that should be flagged in the report with a plan to move off it, even if the move needs your approval or a separate quote.
If you're on managed or shared hosting, the host may handle the server layer. The report should still say who is responsible for it, so nothing falls in the gap between your agency and your hosting provider.
Backups taken and whether a restore was tested
Most reports say "daily backups taken". That tells you very little. You want to know:
- Frequency and type: full site (files and database) or database only. An online store with daily orders needs more frequent database backups than a five-page brochure site.
- Where they're stored: on the same server as the site, or somewhere separate like Amazon S3, Google Drive or Backblaze. A backup on the same server dies with the server.
- How long they're kept: if malware sat quietly on your site for three weeks and you only keep seven days of backups, every copy you have is infected.
- The date of the last restore test, and the result.
That last point is the one most agencies leave out, and it matters most. A backup nobody has restored is a hope, not a backup. We've seen backup plugins report success for months while writing incomplete archives because of a server timeout. Nobody noticed until they needed one.
A restore test doesn't have to be monthly for every site. Quarterly is reasonable for a simple site. But the report should state when the last one happened, what was restored (to staging, ideally), and whether the restored site worked: pages loaded, forms submitted, logins worked.
Uptime, speed and security scan results
Uptime
You should see the uptime percentage for the month and, more usefully, a list of each downtime incident: when it started, how long it lasted, the cause, and what was done. Tools like UptimeRobot or Better Stack produce this automatically. A percentage on its own hides a lot. One four-hour outage during your busiest Saturday is worse than several two-minute blips at 3am, even if the totals look similar.
Speed
Look for Core Web Vitals numbers from real users where available, through PageSpeed Insights or Google Search Console, not just a lab score out of 100. The thresholds Google treats as "good" are:
| Metric | What it measures | Good |
|---|---|---|
| LCP (Largest Contentful Paint) | How fast the main content appears | 2.5 seconds or less |
| INP (Interaction to Next Paint) | How quickly the page responds to taps and clicks | 200 milliseconds or less |
| CLS (Cumulative Layout Shift) | How much the page jumps around while loading | 0.1 or less |
INP replaced FID as a Core Web Vital in March 2024. If your report still talks about FID, it's using an old template, and that says something about how much attention it's getting.
The useful part is the trend. Did speed drop this month? If so, why: a new plugin, a heavy banner image someone uploaded, a slow third-party chat widget? A report that shows numbers without explaining changes is only half done.
Security
The report should name the scanner used, such as Wordfence, Sucuri or the host's own malware scanning, and list what it found. Include blocked login attempts if relevant, any files flagged as modified, and plugins with known vulnerabilities. Also check that your SSL certificate's expiry date is noted. Expired certificates still take sites down more often than they should.
"No issues found" is a fine result. Just make sure it comes with the name of the scan and the date it ran.
The work done on your requests and the hours it took
Most retainers include a block of hours for changes you ask for: updating prices, adding a team member's photo, fixing a broken form, building a new landing page. This section is where you check whether you're getting those hours.
Each request should appear with:
- What you asked for and when
- What was done
- Time spent
- Status: done, in progress, or waiting on you
Then a running total: hours included in the plan, hours used, hours left, and whether unused hours roll over. If the agency also counts routine maintenance work (updates, backups, scans) against your hours, the report should show that split clearly. Otherwise you'll wonder why a month with two small text changes used up the whole block.
Watch for vague entries like "general site improvements – 3 hours". Ask what that was. A good agency will tell you without getting defensive.
This is usually the point where owners start asking what all this should cost and whether the hours are priced fairly. We've covered that separately in our guide to website maintenance cost in India, including what changes the price.
What to actually do with this
Ask for one sample report before you sign a retainer. If the agency can't show you one, or it's a single page of green ticks, expect the same after you've paid.
Read the report each month, even briefly. Check three things: did anything break, when was the last restore test, and how many hours are left. That takes five minutes.
Be honest about what you need. A five-page site that rarely changes, built on a simple setup, doesn't need a heavy monthly retainer with a block of development hours. Automatic minor updates, offsite backups, an uptime monitor and a quarterly check-up may be enough, and some good hosts bundle most of that. A static site or a hosted platform like Shopify or Wix removes most plugin and server work entirely. The fuller retainer makes sense when your site takes payments, captures leads you depend on, runs many plugins, or changes every week.
If you want a rough figure for your own site, the pricing calculator will give you a starting estimate.
















