Skip to content
Retainer | Ongoing

Patching, upgrades, monitoring and the occasional thing that breaks at a bad time. For a store that works and needs to keep working, without a developer on payroll.

How engagements work

In one paragraph

What does eCommerce website maintenance actually cover?

Security patching on the platform and its extensions, version upgrades, dependency and certificate renewals, uptime and error monitoring, database and cache health, backup verification that actually restores, performance tuning as the catalogue grows, and incident response during agreed hours. Plus the small changes that accumulate: a payment method, a shipping rule, a template fix. Typically a reserved block of hours each month rather than an hourly call-out, because the work that prevents incidents is the work that gets skipped when it is billed by the emergency.

What this is

A commerce platform is not a finished object. Dependencies get security advisories, payment providers change APIs, certificates expire, the catalogue grows past what the current indexing handles, and the backup nobody has tested turns out not to restore.

None of that is urgent until it is all urgent at once, usually on the highest-traffic day of the year. The point of a retainer is that the preventable half never becomes an incident, and the unpreventable half has someone who already knows the system.

What this page is not

This is ongoing ownership, not a one-off fix. If something is broken right now and you need it diagnosed, that is better as an hourly engagement than a retainer.

What usually brings this in

  • The store has not been patched in a year.

    Known vulnerabilities in the platform and its extensions, and a card-skimming incident is the usual way a business finds out.

  • The developer who built it has gone.

    Nobody knows how it deploys, nothing is documented, and every change is a risk nobody wants to take.

  • When something breaks, nobody knows until a customer says so.

    No monitoring, so the time between a failure and finding out is however long it takes someone to complain.

  • Small changes wait weeks for a quote.

    A payment method or a shipping rule becomes a project, so the business stops asking and works around its own store.

What gets built

Security patching and upgrades
Platform and extension patches applied on a schedule and tested before they reach production, with the version position reported rather than assumed.
Monitoring that reaches a person
Uptime, error rates and the key transactional flows, alerting to somebody rather than accumulating in a dashboard nobody opens.
Backup verification
Restores tested, not just backups taken. An untested backup is a belief rather than a safeguard, and this is where that gets found out safely.
Performance as the catalogue grows
Cache, index and query health reviewed as the data changes, because the configuration that was right at 5,000 products is not right at 200,000.
The small changes
Payment methods, shipping rules, template fixes and content structure, inside the monthly allocation rather than quoted individually.
Incident response
Priority access during agreed hours, from someone who already knows the system rather than someone reading it for the first time under pressure.

Technologies

Platforms
  • Magento 2
  • Adobe Commerce
  • Shopify
  • WooCommerce
  • WordPress
Infrastructure
  • AWS
  • Nginx
  • Varnish
  • Redis
  • Cloudflare
Monitoring
  • Uptime monitoring
  • Error tracking
  • Log review
  • Backup verification
Release
  • Git
  • GitLab CI
  • Jenkins
  • Staging environments

How delivery runs

  1. Audit what you have

    Version position, patch status, extension inventory, monitoring, backups and deployment process. The output is a list, and often the first useful document the platform has ever had.

  2. Fix what is urgent

    Security patches and anything actively broken, before any routine work starts. This is usually the first month.

  3. Put the routine in place

    Patch schedule, monitoring, backup verification and a documented release process, so the preventable half stops being incidents.

  4. Run it, and report monthly

    A written summary of what changed, what was patched, what broke and where the hours went. Rolling monthly, cancellable with notice.

Questions

What happens in the first month?
An audit and then the urgent fixes. On a platform that has been unmaintained for a while the first month is almost entirely catching up, and the routine work starts once the backlog of security patches is clear.
Do unused hours roll over?
That is agreed per arrangement and I am flexible about it. What I will not do is treat a retainer as a budget to be consumed: the point is reserved capacity, and a month where little goes wrong is the arrangement working rather than value lost.
What if a large project comes up mid-retainer?
It gets scoped separately. A retainer consumed by one build is not a retainer, and pretending otherwise means the maintenance stops happening precisely when the platform is changing most.
Can you support a platform you did not build?
Yes, and that is most of this work. It starts with reading it, which is the audit in step one, because an undocumented platform in the order path is not something to start changing on day one.

Describe what is in front of you and you get a written scope and a figure, usually within a few days of a short call. No charge for either.

All services