Chosen
What Should a Website Maintenance Plan Actually Include?
Compare maintenance plans by what gets checked, fixed and restored—not just the monthly price.

Milan Gohil
·
10 min read
OVERVIEW
Here is a short version of what I write.
A useful website maintenance plan defines who monitors the site, who fixes problems and how recovery works. The right scope depends on the platform, integrations and business impact of downtime.
Specify updates, security checks and functional testing.
Define backup frequency and test restoration.
Agree response arrangements, support limits and exclusions.
Request reports that explain issues found and work completed.
A website maintenance plan should answer one practical question:
If something changes, breaks, slows down, gets attacked or quietly stops working—who notices, who fixes it, and what happens next?
That is more useful than a proposal that simply says:
Monthly website maintenance — $300/month.
Because “maintenance” can mean completely different things. One provider may only install software updates. Another may include:
backups
security monitoring
uptime monitoring
form testing
broken-link checks
performance monitoring
search/indexing checks
small fixes
recovery support
Before comparing maintenance plans, find out what you are actually buying.

Not every website needs the same level of maintenance. A five-page brochure site and a store processing hundreds of orders should not have identical plans.
First: hosting is not automatically maintenance
Your hosting provider may keep the server running. That does not necessarily mean someone is checking:
whether your contact form still sends
whether a CMS update broke a page
whether an integration stopped syncing
whether Google stopped indexing an important page
whether a backup can actually be restored
whether a checkout flow still works
So when a proposal says:
Hosting + maintenance included
ask for the two responsibilities separately.
Want AI to review a maintenance proposal?
Paste the proposal or retainer into an AI tool with this prompt:
I am considering this website maintenance plan.
WEBSITE:
[PASTE WEBSITE URL]WEBSITE TYPE:
[BUSINESS WEBSITE / ECOMMERCE / MEMBERSHIP / SAAS / OTHER]IMPORTANT FUNCTIONS:
[FORMS / BOOKINGS / PAYMENTS / CRM / SEARCH / CUSTOMER LOGIN / OTHER]MAINTENANCE PROPOSAL:
[PASTE PROPOSAL]Review this from the perspective of the business buying the service.
Do not assume something is included unless the proposal explicitly says so.
For each of these areas classify the plan as:
INCLUDED
EXCLUDED
UNCLEAR
EXTRA COST
NOT RELEVANTReview:
1. Hosting
2. Domain/DNS
3. SSL
4. CMS/platform updates
5. Plugin/app/dependency updates
6. Update testing
7. Backup frequency
8. Backup retention
9. Backup restoration/testing
10. Security monitoring
11. Vulnerability response
12. Uptime monitoring
13. Broken-link/error monitoring
14. Forms
15. Checkout/booking testing
16. Third-party integrations
17. Performance monitoring
18. Search Console/indexing monitoring
19. Analytics/tracking checks
20. Small bug fixes
21. Content updates
22. Design changes
23. SEO work
24. Emergency support
25. Response times
26. Reporting
27. Ownership/access
28. Cancellation/handoffThen identify:
- what the plan actually protects
- important gaps
- vague wording
- work likely to become an extra charge
- what happens during an emergency
- questions I should ask before signingDo not assume I need every service.
Separate:
ESSENTIAL FOR THIS WEBSITE
USEFUL
OPTIONAL
The goal is not to buy the biggest maintenance plan. It is to make the responsibilities visible.
1. Updates should include more than clicking “Update”
Websites built on software that you control may depend on:
CMS core
themes
plugins
extensions
packages
libraries
server software
Those components change over time. Security fixes are one reason updates matter. CISA recommends regularly patching software and prioritising known exploited vulnerabilities, particularly on internet-facing systems. For WordPress specifically, WordPress recommends keeping core, plugins and themes current and also recommends having backups available in case an update causes problems. But a useful maintenance service should answer: WordPress: security hardening.
UPDATE AVAILABLE → Is it relevant/urgent? → Backup / recovery available? → Update → Check important website functions → Fix or rollback if something broke
“Updates included” is therefore less useful than knowing:
What happens if an update breaks the site?
2. Ask what gets tested after an update
Imagine a healthcare provider's website still loads after an update. Technically:
Website is online.
But the appointment form no longer submits. Maintenance failed if nobody checks the function that matters. What should be tested depends on the website.
Lead-generation website
Check things such as:
main contact form
quote form
phone/email actions
CRM delivery
Ecommerce
Check:
product
cart
discount
checkout
payment
order confirmation
Membership website
Check:
login
password reset
protected content
subscription/account actions
Do not ask for:
“full QA every week”
without a reason. Identify the critical journeys that must continue working.

Website | Critical journey |
|---|---|
Law firm | Service page → enquiry |
Clinic | Treatment page → appointment |
Manufacturer | Product → RFQ |
Ecommerce | Product → payment → confirmation |
Training company | Course → registration/payment |
SaaS | Signup → email → account access |
Your maintenance provider should know which ones matter.
3. Backups should answer “Can we recover?”, not only “Do we have backups?”
A dashboard saying:
Backup completed successfully.
is reassuring. But the reason you have a backup is restoration. A useful plan should make clear:
what is backed up
how frequently
how long backups are retained
where they are stored
who can restore them
what happens if the production website is compromised
whether restoration is ever tested
CISA recommends maintaining backups of critical data and testing their availability and integrity as part of recovery planning. NIST guidance similarly emphasizes conducting, maintaining and testing backups so businesses know they are usable when recovery is required. So ask: NIST: protecting data for small businesses.
When was the last successful restore test?
That is much more useful than:
“Do you take backups?”
How often should a website be backed up?
There is no useful universal answer. Ask:
How much data are we willing to lose?
A simple brochure site whose content changes twice a month has different requirements from a store receiving orders throughout the day. Think about:
SITE CHANGES RARELY → Lower backup frequency may be acceptable ORDERS / BOOKINGS / USER DATA CHANGE CONSTANTLY → Loss between backups matters much more
The backup schedule should reflect the business risk.
4. Security maintenance is not “install a security plugin”
Security is broader than one tool. Depending on the platform, maintenance may include:
software patching
vulnerability monitoring
admin-account review
MFA where supported
malware/security monitoring
suspicious-change monitoring
removing unsupported software
incident response
CISA's small-business guidance specifically highlights software updates, MFA, logging and backups among core cybersecurity practices. You do not need your website agency to become your entire cybersecurity department. But the contract should make clear:
What website security is being monitored, and what happens when something is found?
5. Monitoring only matters if someone receives the alert
An uptime monitor can tell you:
Website unavailable.
Great. Now ask: Who receives that notification? At 2 PM? At 2 AM? On weekends? How quickly are they expected to investigate? That is the difference between monitoring and response.

A maintenance plan may include monitoring without unlimited emergency repair. That is okay—as long as it is clear.
6. Your forms and integrations also need attention
Modern websites often depend on systems outside the website:
Website form → CRM → Automation → Email / sales team
The website can look perfect while this chain breaks. Common dependencies include:
CRM
email
booking system
payment gateway
marketing automation
maps
analytics
live chat
product feeds
APIs
For each important integration ask:
How would we know if this stopped working?
Sometimes the answer will be automated monitoring. Sometimes it will be a scheduled test. Sometimes manual business feedback is enough. The useful thing is making that decision intentionally.
7. Performance should be watched for regression
A website can become slower over time as people add:
scripts
tracking tags
fonts
videos
widgets
plugins
larger images
new functionality
Performance therefore is not always a one-time launch task. Google's web.dev guidance treats performance as a continuing cycle: evaluate, improve, then monitor for regressions. Current Core Web Vitals focus on:
LCP — loading
INP — responsiveness
CLS — visual stability.
You do not necessarily need monthly performance-development work. But for a commercially important site, it can be useful to know when performance becomes materially worse.
8. Search health is different from ongoing SEO
This distinction is important when comparing retainers.
Website/search-health maintenance
May include:
indexing errors
important pages unexpectedly disappearing
sitemap problems
structured-data errors
severe organic-traffic anomalies
crawl/indexing problems
Ongoing SEO
May include:
keyword research
new content
link acquisition
competitor research
content optimization
new landing pages
Those are different services. Google describes Search Console as a tool for monitoring and troubleshooting how Google crawls, indexes and displays a site, including indexing problems, traffic data and alerts. Google also recommends website managers monitor significant errors or spikes in Search Console's indexing and performance reports. So:
“Search Console monitoring”
does not automatically mean:
“SEO growth work.”
Make sure the proposal does not blur the two.
9. Small changes need a definition
Many maintenance retainers say:
Minor website changes included.
Ask what minor means. For example:
Could be maintenance
replace an image
change phone number
fix broken layout
update existing text
repair a form
add someone to an existing team template
Usually new project work
design a new landing page
build a calculator
add membership
connect a new CRM
create an ecommerce section
redesign navigation
There is no universal boundary. Put the boundary in the agreement so nobody has to argue later about whether something took “only ten minutes.”
10. Ask whether unused software is being removed
Maintenance should not only keep adding things. Over time websites accumulate:
abandoned plugins
old integrations
test accounts
old themes
scripts nobody remembers
staging environments
unused tags
Unsupported or unnecessary components create more things to maintain. CISA recommends replacing unsupported systems and keeping public-facing software current as part of reducing exposure to known vulnerabilities. Occasional cleanup can therefore be as important as installing the next update.
What should a maintenance report actually tell you?
You probably do not need a 24-page PDF showing:
37 plugins updated.
A useful report answers:

That tells the owner what matters.
How much maintenance does your website actually need?
Do not begin with:
Basic / Professional / Premium package.
Start with website risk.
LOW COMPLEXITY
Example:
8-page consultancy website
No login
One contact form
Hosted SaaS platform
Few changes
Maintenance may be relatively light.
MEDIUM COMPLEXITY
Example:
50-page business website
CMS
Regular articles
Several lead forms
CRM integration
Organic search traffic
Now monitoring, backups, integrations, search health and regular QA matter more.
HIGHER COMPLEXITY
Example:
Ecommerce
Customer accounts
Payments
Orders
Product feeds
Multiple integrations
Frequent changes
Downtime or broken functionality directly affects transactions. The maintenance plan should reflect that risk.

This is more useful than choosing a plan because it includes “20 maintenance hours.”
Before buying a maintenance plan, ask these 12 questions
1. What exactly are you monitoring?
Not just:
“We monitor the site.”
2. What gets updated?
CMS? Plugins? Server? Integrations?
3. What happens after an update?
Are important functions tested?
4. What is backed up?
Files? Database? Configuration?
5. How often are backups made?
Does that match how often our website changes?
6. Can those backups actually be restored?
Who does it?
7. What happens if the website goes down?
Who gets alerted and what response is included?
8. What happens if the website is hacked?
Detection, cleanup and restoration may be separate services.
9. Are forms/payments/bookings tested?
Which ones and how often?
10. Are performance and Search Console problems monitored?
Monitoring is different from ongoing optimisation.
11. How much small-change work is included?
Define it.
12. What happens when we cancel?
Make sure you retain the website, accounts and access you are supposed to control.
The maintenance plan should match the website you actually have
You do not need:
50-point monthly maintenance
because the package looks comprehensive. And you should not assume:
“The website was launched six months ago, so nothing should need attention.”
The useful plan is the one that protects the things the business depends on. For your website, identify:
WHAT CANNOT STOP WORKING?
WHAT DATA CANNOT BE LOST?
WHAT NEEDS SECURITY UPDATES?
WHAT PROBLEMS SHOULD WE KNOW ABOUT QUICKLY?
WHAT DO WE NEED TO RECOVER FROM?
WHAT CHANGES REGULARLY?
Then build the maintenance plan around those answers. Because the real purpose of website maintenance is not keeping a monthly checklist busy. It is making sure a website that matters to the business keeps doing its job—and can recover when it doesn't.
Sources and further reading
All Blogs
All Blogs & Articles
© 2026 Webshunya Studio®. All rights reserved.







