Website handover to a new provider: what to commission for your online store
You have agreed on store support with a new team, but every fix still starts with a message to the previous developer. One person knows where product data exchange runs, another has server access, and the hosting bill goes to an external email address. In this situation, a website handover is a separate piece of work to commission: defining exactly what the new provider must receive, verify and subsequently handle independently.
First agree on the scope of support, the outcome of onboarding and how to verify it. The owner then understands what they are paying for, and the new provider does not bury unknown work in an overall estimate. Changing providers does not in itself require a new design or a different platform.
Commission team independence, not the delivery of an archive
An offer to “take over website support” can cover very different services. Sometimes it means only access to support requests and billable hours. Sometimes it includes reviewing code, deploying a test copy, clarifying integrations and transferring responsibility for agreed operations. Comparing such offers by their monthly fee without clarifying the work included is misleading: cheaper support may omit exactly what the team needs before it can start working.
Define the outcome in business terms: “the new team can change an agreed shipping rule, check an order and explain who responds to an error without help from the previous provider.” This is not a promise to fix every existing shortcoming. It is a boundary of independent work that can be demonstrated and accepted. Use it to determine the list of materials, access rights, checks and responsible people.
Record separately what remains outside the new contract. For example, the business management system is supported by its integrator, the payment account is administered by the finance department, and the owner continues to pay for the domain. The new web team needs the contacts and escalation process, but should not silently become responsible for other providers’ services. Appoint one business-side coordinator during the transition to agree on these boundaries and priorities.
Trace one order through everyone responsible
Instead of a long list of technologies, ask to see the sales journey. Where does the customer see stock availability, where does the price come from, who calculates shipping, and when does the order appear in the business management system? Each handoff needs a system owner, an authorised way to check it and a sign of a successful outcome. This makes it easier to spot a service left out of the contract that the store nevertheless needs to fulfil an order.
Alongside this journey, record critical daily operations: updating the catalogue, receiving payment notifications, transferring orders and generating shipping documents. Do not expect the owner to read server logs. Ask the technical lead to explain how the team distinguishes “the data exchange started” from “the required order reached the manager.” That difference determines whether support notices a problem before a customer complains.
Responsibility during the transition should be agreed before the old contract ends. Who handles urgent requests today, who handles them while both teams are involved, and who takes over after acceptance? If two teams change the website simultaneously, they need a shared record of planned work and an approval rule. Otherwise, a change absent from the test copy may reach the live store without the new provider knowing.
The working package: what the business should retain
It is useful to organise a handover package around actions rather than file formats. Changing functionality requires current code, dependencies and startup instructions. Recovery requires matching data and file backups and a tested procedure. Contacting a supplier requires a company account, administrative rights or an authorised contact. A list of logins alone confirms none of these capabilities.
In the register, record the service, its owner, the role required, the access-granting process and the verification result. Keep passwords and keys in an agreed secure vault, not in a spreadsheet forwarded to everyone involved. Developer access to code, permission to pay for hosting and the ability to recover an administrator account are different powers. The business needs to know who holds each of them.
For a repository, it is important to check not only the new owner but also collaborators and automated connections. For example, the GitHub documentation explicitly states that after a repository transfer, the previous owner is added as a collaborator, other collaborators remain, as do associated secrets and deploy keys. Thus, changing ownership does not revoke old access. Agree on a separate revocation process after checking the necessary replacements.
A Drupal store also needs this package before approaching GL.ua (Glyanets): the company specialises in Drupal website development and support. The repository alone does not replace a recovery package. Official Drupal documentation distinguishes the database, user-uploaded files, a known working code version and a tested rollback procedure. Record the backup contents, creation time and responsible person. Test recovery in an agreed environment, without experimenting on live orders.
Distinguish three states of support readiness
The phrase “everything has been handed over” hides the difference between having materials and being able to act. A short table is useful in negotiations: it helps agree on a payment milestone and avoids describing incomplete onboarding as full support.
| State | What is already available | What has not yet been demonstrated |
|---|---|---|
| Materials received | Code, service records and documentation are available | That they match the live store |
| Operation reproduced | The new team runs the agreed version and checks scenarios on a test copy | That support responsibilities and the response process have been accepted |
| Support scope accepted | Verified actions, responsible people, contacts and a list of open issues are in place | That the system has no defects: a handover does not establish this |
Ask for a demonstration that the test and live versions match. If the server contains changes absent from the repository, first establish their purpose. Do not approve replacing live code merely because the archived version looks “more correct.” This is separate work addressing a specific question, not a reason to rewrite the store immediately.
Documentation also needs to be tested through action. Have a specialist from the new team perform an agreed step using the instructions, while the previous provider only explains gaps. If the work stops without their verbal guidance, update the record. The goal is instructions that can be followed again, not a video call after which the knowledge once again remains with one person.
A hypothetical example: changing shipping without changing the store
Imagine a Drupal store transferring support to another team. The owner wants to change the free-shipping rule but does not know where it is defined: in website settings or in the business management system. This is a hypothetical example, not a GL.ua case study. The new provider’s task is first to establish the source of the rule and only then estimate the work required to change it.
An agreed check could look like this:
- Record the current rule and identify the system in which it is defined, together with the person responsible for the business condition.
- On an isolated copy, prepare baskets below, at and above the threshold, including a discount and an unavailable shipping method.
- Compare the expected amount in the basket, the placed order and the record received by the manager.
- Show the owner the result and agree on who authorises the change in the live store and what to do if results differ.
This scope tests a specific team capability while revealing dependencies. If an external system defines the rule, there is no need to duplicate it on the website just to achieve a quick result. The estimate must include the involvement of the person responsible for that system. If this requires reworking the entire data exchange, the owner gets a separate decision about extending the work rather than an unexpected bill after the start.
The first change should be useful, but it should not be forced into the live store merely to tick a box. Sometimes checking an already agreed process in a test environment is enough. If the new team recommends migrating the platform at the same time, ask them to separate the reason for migration from the difficulty of getting to grips with someone else’s code. That is a different decision with its own criteria, including those discussed in the article on acceptance and portability when migrating to Adobe Commerce.
How to pay for uncertainty without commissioning an endless audit
When the store’s condition is unknown, a fixed price “for the entire handover” may include a large contingency or leave out difficult parts. It is more practical to separate a bounded introduction to the system from resolving the problems found. For the first stage, agree on the questions, available materials, spending limit and deliverable: a list of verified items, specific blockers and an estimate for subsequent work. An audit is hard to finish without a decision it is meant to inform.
In the estimate, separate the previous provider’s explanation time, the new team’s work with code and environments, integration checks and preparation of working instructions. Test hosting, necessary licences or assistance from the business management system supplier may be separate expenses. Fixing old defects and subsequent monthly support should not silently merge into the handover fee.
The timeline also depends on external decisions: obtaining account rights, integrator involvement and the ability to reproduce the environment. Ask for the team’s working time to be separated from time spent waiting on these dependencies. If the previous developer is unavailable, some behaviour will have to be established through authorised materials and controlled tests. This route is not knowledge transfer and must be estimated with the unknowns in mind.
End the transition with a specific support decision
At the end, you need a concise record: the scope accepted by the new team, what it demonstrated, which services remain with other providers and who resolves open issues. For the previous team’s temporary access, specify a responsible person and a condition for ending it. Do not randomly delete a shared technical account: first check which processes use it and what will replace it.
The next step after handover is to assess support performance, not collect archives again. A separate technical support check through a simulated incident will help here. It answers a different question: how the team acts when a problem occurs, rather than what needs to be handed over at the start.
GL.ua (Glyanets) develops and supports Drupal websites, including online stores with integrations. If you plan to hand over a Drupal store to a new provider, start the conversation with the support scope and the first business task. To discuss Drupal website support and development with GL.ua, prepare a description of the platform, a list of external systems and an account of what currently depends on the previous developer. This is enough for a substantive start without sending passwords, code or customer data through the contact form.
Just one step to your perfect website



