A website is a small production system. It runs on software that receives security updates, depends on certificates and DNS, stores customer data and connects to payment and email services. Maintenance is the routine that keeps those dependencies from turning into an outage. The plans on the market differ less in the list of tasks than in what happens when something goes wrong at 9pm.
This comparison covers five common models, the criteria that separate them and the questions that expose the real scope behind a monthly fee.
The five common models
1. Break-fix or hourly support
What it includes: work is quoted when something breaks. There is no retainer and no proactive schedule.
Best fit: a static brochure site with no transactions, a known internal owner and a willingness to manage risk.
Limits: nobody is watching uptime, backups or updates between incidents. The hourly rate for an emergency is usually higher than planned work, and the business carries the availability risk.
2. Shared care plan
What it includes: core updates, a backup, uptime monitoring and a support desk, usually at a low monthly fee for many small sites.
Best fit: a small content site with standard plugins and no custom code.
Limits: the plan is standardized, so it may exclude malware cleanup, development hours, staging or performance work. Response times can be pooled across many customers.
3. Managed application maintenance
What it includes: a named owner, a maintenance calendar, staging, backups with restore tests, security patching, monitoring, a support allowance and monthly reporting.
Best fit: an ecommerce store, a lead-generation site or a custom application where downtime has a revenue cost.
Limits: higher fee and a scope that must be read carefully. Development work is usually separate from maintenance hours.
4. Hosting plus maintenance bundle
What it includes: hosting, updates, a control panel and support from one provider.
Best fit: a business that wants one invoice and a standard stack.
Limits: the bundle can hide the trade-offs. Moving hosting later may be harder, and the maintenance scope may be thinner than it looks.
5. In-house plus tools
What it includes: an internal person or team uses monitoring, backup and update tools, with external help for specialist work.
Best fit: a business with a technical team and enough sites to justify the role.
Limits: holiday cover, turnover and after-hours response. The tools do not replace judgement.
Comparison criteria
- Scope: updates, backups, security, performance, content and development hours.
- Response and resolution: the service level for a site down, a payment failure and a content change.
- Backup and restore: frequency, retention, off-site storage and a tested restore.
- Security: patching, malware scanning, web application firewall, access review and incident handling.
- Performance: Core Web Vitals monitoring, image and cache work, and a process for regressions.
- Change management: staging, review, rollback and a maintenance window.
- Reporting: what is measured, who reads it and what happens when a metric is bad.
- Contract: term, notice, price review, exclusions and exit assistance.
- Cost model: fixed fee, hours, per-site pricing, emergency rate and project work.
What good maintenance actually includes
- Inventory and ownership: domain, hosting, DNS, CMS, plugins, integrations and who holds each account.
- Updates: core, theme, plugin and language runtime, tested on staging before production.
- Backups: automated, encrypted, stored separately and restored in a test at least quarterly.
- Monitoring: uptime, certificate expiry, error rates and form or checkout failures.
- Security: malware scanning, firewall rules, access review and a patching record.
- Performance: a baseline, monthly checks and a fix plan for regressions.
- Content and forms: broken links, contact forms, email deliverability and policy pages.
- Reporting: a short monthly note with work done, incidents, risks and recommendations.
- Incident process: who is contacted, how the site is contained and how data is restored.
- Documentation: credentials, architecture and handover notes that the business owns.
Pricing questions
- What is included in the monthly fee, and what is billed separately?
- How many support hours are included, and what happens when they run out?
- What is the emergency rate, and is it capped?
- Is malware cleanup included, or quoted per incident?
- Are plugin, theme and hosting licences included?
- How often are backups tested, and how long are they retained?
- Is staging included? Is development work priced at a fixed rate or hourly?
- What is the notice period, and what does handover include?
A low monthly fee with unlimited updates but no restore test is not a maintenance plan. It is a subscription to activity.
Red flags
- The provider owns the domain, hosting or analytics account in its own name.
- No written scope, no service level and no named contact.
- "Unlimited" anything without a fair-use definition.
- No staging environment for a site with custom code or payments.
- Malware cleanup priced as an open-ended extra.
- No reporting and no record of changes.
- A long lock-in with no exit or data handover clause.
Decision guidance
- Brochure site, low traffic: a shared care plan plus a quarterly restore test is usually enough.
- Ecommerce: managed maintenance with security, performance and a defined incident response. Do not rely on a generic care plan.
- Custom application: maintenance must cover the application, dependencies and infrastructure, not just the CMS.
- Multiple sites: compare a per-site plan with an in-house owner and tooling; the break-even depends on volume and skills.
- Regulated or customer-facing: ask for evidence of access review, patching and incident handling, not just a monthly report.
For a business in Bangladesh, check that the provider can support your payment gateway, Bengali content workflow and the hours when customers actually shop. A provider in another time zone can work, but the escalation path must be explicit.
Handover checklist
Whichever model you choose, the business should hold the domain, hosting and DNS accounts, a current list of administrators, an export of the site and database, the deployment documentation and the analytics ownership. Test the handover before you need it. A provider that cannot hand over cleanly is a provider that has become a single point of failure.
Metrics
Track uptime against the agreed target, the time to apply critical patches, backup restore success, mean time to respond and resolve, the number of security findings closed, Core Web Vitals status, support hours used and the number of unplanned changes. Review them quarterly with the provider and agree on one improvement for the next quarter.
Where this fits in our work
If you need a maintenance model with a defined scope and an exit path, start with Custom Software and review the engagement options on our pricing page. For the incident side, read the website malware cleanup cost guide.
Sources
- WordPress Developer Resources, "Hardening WordPress", accessed 17 November 2026.
- OWASP, "Top 10 Web Application Security Risks", accessed 17 November 2026.
- NIST, "Cybersecurity Framework 2.0", accessed 17 November 2026.
- web.dev, "Web Vitals", accessed 17 November 2026.
Responses (0)
No comments yet. Add the first one below.
Leave a response
Your email address will not be published. Required fields are marked *.