For UK teams

Outsourced PHP development for inherited UK systems

Assess an inherited PHP application before extending it. Set a technical baseline, prioritised change plan, review evidence and explicit operating duties.

Let’s talk

Move an existing PHP system forward

A professional speaking on a laptop video call at a desk.
Illustrative photography · LinkedIn Sales Solutions / Unsplash

Your application may still do important work even when its original developer has moved on or nobody is sure how to release the next change. Tell us which task your team depends on and what you need to improve. We can discuss an assessment, a specific feature or a continuing development arrangement around that work.

We start with the condition of the application and the people using it. A small, carefully checked change may be a sensible first step. If the code or operating arrangements make that difficult, you should understand why before committing to ongoing maintenance or a replacement.

Begin with technical discovery

We review the runtime, framework, dependencies, data model, hosting and release process. Just as useful is a demonstration from someone who uses the application: what do they enter, what result do they expect, and where do they have to work around a problem?

We also establish whether there is a usable development environment and a recoverable backup. Where important behaviour has no automated checks, the first scope may need focused tests around the part being changed. Those findings help us distinguish urgent risks, preparation work and features that can be planned next.

For example, an upgrade and a new export may compete for the same budget but solve different problems. The assessment should explain whether the upgrade is needed for that export, has a separate purpose, or can be considered later. Your team can then decide the order with the consequences in view.

Make changes reviewable

Before implementation, we agree what the change should do, who will check it and how an approved version reaches production. A reviewer needs examples they recognise from their work, including an exception that matters to the business.

For an integration, that includes what happens when a record cannot be sent and how someone investigates without exposing customer information in logs. For a data migration, it includes how the transferred information is checked and what recovery would involve if the result is wrong.

Laravel development and API integrations may be relevant to the existing architecture. We would discuss how they fit the application before proposing a framework change or rewrite.

Set support expectations explicitly

Tell us who currently monitors the application and what happens when someone reports a failure. We agree development work separately from monitoring, incident response and emergency availability. A development retainer does not automatically provide round-the-clock cover.

For a mix of custom applications and websites, see PHP and WordPress maintenance. Tell us about your application, the work it supports and the sort of help you need. We can then identify which responsibilities need to be discussed before a proposal.

Commission an assessment with an actionable output

Give the assessment a question your business needs answered. Perhaps you need to know whether an inherited ordering process can support an additional step, or why a report regularly needs manual correction. Agree which application areas will be inspected and what information the reviewer can access.

Your team supplies source access through an approved process, known defects, environment details and someone who can explain the operating rules. Existing documentation is helpful, but say where it is incomplete. We can identify whether the review also needs a separate environment, prepared sample data or cooperation from the incumbent supplier. Sensitive production data is not automatically required.

The report should leave you able to choose a next step. If missing access prevents an answer, it should say what is missing, who needs to supply it and which finding remains provisional. You should be able to distinguish a problem demonstrated during the review from a concern that still needs investigation.

Turn findings into a prioritised workstream

We can separate immediate risks, work needed before implementation and the features you originally requested. That makes it easier to see what a first phase would achieve and which tasks can wait.

For example, suppose your team wants to change an export used by another system. A first step could be to reproduce the current output, agree a sample of the proposed data and ask the recipient to confirm how it will be used. Review would cover valid records, invalid cases and the business owner’s confirmation that the result is useful. This is an example of how a scope could be organised.

A defined change can then move through implementation, review and an approved release. Where tests are missing, the agreed work may include checks around the affected behaviour. Your team should receive an explanation of those checks and any important limits, so the next change does not depend on remembering an undocumented conversation.

Keep maintenance and operations distinct

Agree who monitors the system, investigates incidents and authorises production changes. Those may be different people. A developer preparing a correction needs to know who can provide access, who can explain the reported behaviour and who can approve the release.

Handover can include environment requirements, dependency decisions and the recovery approach for the delivered work. The person responsible for operating your system should review those instructions and identify anything they cannot carry out with the access or tools they have.

A project may suit a defined change supported by the assessment. Continuing capacity may suit a prioritised queue with reviewers available to check completed work. Urgent incident response needs a separate availability discussion. We would set out those commitments so you can compare proposals that cover the same responsibilities.

Ask discovery to answer a funding decision

An assessment should help you decide where to put the next part of the budget. Ask which issue blocks your immediate requirement, what evidence supports the finding and what depends on another supplier. An upgrade, a temporary containment measure and a replacement each need an explanation of what they would resolve and leave outstanding.

Before the first change, agree the affected workflow, test environment and recovery conditions with the application owner. Include a check that useful existing behaviour still works. If the proposed change alters the way staff complete a task, make time for those staff to review it before release. Routine maintenance, new features and incident response each need their own agreed commitments.

Questions about inherited UK PHP applications

A few useful answers before you decide.

Do we need to replace the application to obtain support?
A rewrite is not the starting assumption. We first assess the application, the work it supports and the available release controls. The next step might be a targeted correction, an upgrade or a larger change with its own business case. You should understand the reason for that recommendation before choosing a scope.
Is a dependency upgrade the same as routine feature capacity?
The scope depends on compatibility, testing and release risk in your application. We separate that work from requested features and explain how it affects the queue. For example, if an upgrade must come first, your reviewer should know which workflows need checking and what will happen to the feature’s planned timing before agreeing the work.
What should the initial technical report contain?
It should answer the agreed business question using findings your team can follow. Look for confirmed issues, evidence, priorities, missing information and dependencies. Ask it to explain the proposed next scope and the reason for its order, so your team can make a funding decision without treating every observation as equally urgent.
Can you maintain code written by another supplier?
Potentially, after assessing access, condition and responsibilities. Bring what you know about the application and any agreement with the previous supplier. The review establishes whether targeted changes are practical and what must be resolved before we agree continuing support. We cannot commit to an unseen codebase at a fixed effort.
Should we upgrade everything before building a feature?
We assess which changes the requested feature actually depends on, alongside safety and compatibility needs. Your application owner can then compare the sequencing options. If unrelated upgrades are proposed at the same time, ask how their testing and recovery will be handled and why they belong in that release.
How will we know a bug is reproducible?
Describe the input, environment and steps that show the problem, then state what should happen instead. An anonymised example can help the developer and business reviewer check the same case. A screenshot may show the symptom; repeatable steps help establish whether a correction addresses it. We record any context still needed to reproduce the issue.
Does a successful staging test authorise production deployment?
No. The named release owner approves production changes through the agreed process. Staging results help them decide, alongside backup, recovery and operational readiness. Developer access alone is not permission to change a live business system; the release and its approval need to be arranged explicitly.
What if the assessment recommends a larger replacement?
Ask for the constraints, alternatives and evidence behind the recommendation. A replacement needs its own scope and business decision, including the useful behaviour that must be retained and the risks of moving data or operations. You may need further investigation before choosing between replacement and continued work on the existing system.
Let’s move things forward

Start a conversation

Tell us what you are working towards. We’ll help you find a practical next step.

Let’s talk