On 13 August 2024, Shopify stopped honouring checkout.liquid customizations on the Information, Shipping and Payment steps of checkout. The deprecation notice was published well ahead of the date, and the consequence was stated plainly: stores that had not upgraded to checkout extensibility were converted to the default Shopify Checkout without their customizations.
Plenty of brands handled it months early. Others found out because a delivery-instructions field they had paid for in 2021 quietly stopped appearing, and the contractor who built it had moved on. The work had been good. Nobody had been responsible for noticing that the platform underneath it changed.
That distinction decides most of these arrangements. A Shopify agency for established brands earns its price at the point where the store’s main risk stops being the quality of individual changes and becomes the reliability of everything around them: when one change going wrong, or one person going quiet for three weeks, costs more than the change was worth. Order volume, theme complexity, the number of connected systems and how much the business intends to change over the next two years all move that threshold.
What Usually Fails First, and Why It Is Rarely the Code
Reviews of freelancer arrangements tend to start by questioning the standard of the work, which is usually the wrong place to look. Independent Shopify specialists are often very good at what they were hired to do. What degrades is the structure around it: who remembers why a decision was made, who checks it before customers see it, and who is obliged to act when conditions change.
Context Lives in One Person’s Memory
A solo contributor rarely has a commercial reason to document reasoning. The variant selector behaves oddly on one collection because of a supplier feed quirk from two years ago; the shipping logic carries an exception for one country. None of it is written down, because the person who knows it has always been reachable. The cost surfaces when a second developer has to infer intent from code, and rebuilds rather than repairs because rebuilding is the cheaper route to confidence.
Nothing Is Reviewed Before Customers See It
One contributor writes the change, decides it is finished and publishes it, so the first real test is a live shopper. Defects found that way are not evenly distributed: those that break loudly get reported quickly, while those that merely make a step harder can persist for months. Microsoft Clarity describes rage clicks as “repeated clicks/taps in a specific area of a webpage in a short amount of time that does not result in any change on the page”, and notes that analysing them “can help identify potential problems with your site such as bugs, broken elements, or deadlinks”.
Separate Decisions Accumulate Into a Patchwork
Where several contractors have contributed over years, the result is often less a store than a set of independently maintained decisions. Our RIPO International case study describes that starting position: “Each country had its own separate website, maintained independently, with inconsistent branding and no unified conversion strategy.” Consolidation is a different kind of project from any change a single supplier was asked to make, and a per-task arrangement rarely proposes it unprompted.
Seven Signs the Freelancer Arrangement Has Reached Its Limit
No single one of these settles the question. Several together usually mean the store has outgrown the shape of the arrangement rather than the people in it.
- Delivery dates are set by one person’s calendar. Campaigns and seasonal changes cannot be planned around the store when timing depends on someone else’s other clients.
- Nobody can reconstruct what changed last month. Without a release record, every incident starts from zero, and working out what happened can take longer than the fix.
- Platform deadlines arrive as surprises. Deprecations are published in advance, so learning about one from a broken page suggests nobody is paid to read the release notes.
- Changes are published straight to the live theme. Editing what customers are using removes the chance to check work privately, and makes reverting a reconstruction rather than a rollback.
- Requests are described rather than diagnosed. When the brief is always “make this button green” and never “why are mobile users abandoning at delivery”, the supplier is a pair of hands, which works until the business needs an opinion.
- The business does not hold its own accounts. If the Shopify owner role, the registrar login or the analytics property sits under a contractor’s personal email, continuity depends on goodwill.
- Improvement work never reaches the top of the queue. Bugs and campaigns are urgent and win, so conversion work can go unstarted for a year without anyone deciding against it.
The last is worth separating, because a bigger budget alone does not solve it. A queue full of the immediate stays that way unless someone is explicitly accountable for work that is never urgent. We have written separately about what happens when a developer handles tickets but nobody owns growth.
What a Shopify Agency for Established Brands Should Be Able to Show You
Most of what separates agencies is invisible in a pitch and visible in a working arrangement. Each point below has a demonstrable answer: a screen, a repository, a document or a number. A vague answer marks the part of the relationship that will need defining in the contract.
Access Held in the Business’s Name
Shopify’s intended model for agency work is the collaborator account. Partners request access using a four-digit collaborator request code the merchant generates, and Shopify confirms that “collaborators don’t count towards your store’s user limit”. Permissions are set per collaborator, so theme and catalogue access can be granted without payouts or customer data. The practical benefit is that such access can be revoked by the business in one action, without changing any password the team uses.
A Release Process You Are Allowed to Inspect
Shopify’s own guidance is that version control “helps you to manage changes to your code over time”, and that with the GitHub integration “any changes that a merchant makes to the theme are tracked as commits to the branch”. A store run that way has a history and a route back; one edited directly in the admin has neither, which only becomes apparent on the day something must be undone quickly. Ask which branch is connected to the live theme, where work is previewed, and who approves a release. Ask also how a release is verified afterwards, since agreeing in advance what would count as success is what stops any later movement being read as support.
Commitments You Can Measure Rather Than Assume
Responsiveness is easy to promise and hard to compare. Salesforce defines first response time as how long it takes “to send the first meaningful reply after a customer submits a request”, and treats SLA compliance as a measure of “how often the team meets defined service level agreements”. Applied to an agency, that means agreeing separately what counts as an incident, what response applies to one, and what happens when the target is missed. A single blended promise covers a checkout outage and a copy change in one sentence.
An Opinion About the Roadmap, Not Only the Queue
For many established brands the reason to move is strategic rather than technical. A supplier that only executes produces what is requested, which caps the outcome at the quality of the requests. An agency worth a retainer should be able to say which of three requested items it would do last and why.
Key takeaway: the useful test is not whether an agency can do the work. It is whether the store would still be safely maintainable if every individual currently touching it stopped answering tomorrow.
Freelancers, a Specialist Agency or a Mix: How the Options Genuinely Differ
The choice is not binary, and the strongest setups for mid-sized brands are frequently hybrids. The table compares the four arrangements most established Shopify merchants weigh up.
| Criterion | Individual freelancers | Small specialist agency | Full-service ecommerce agency | Internal lead plus contractors |
|---|---|---|---|---|
| Best suited for | Defined changes on a stable store | Ongoing work in one dominant discipline | Multi-discipline programmes and platform change | Enough internal volume to justify a salary |
| Continuity if one person leaves | Low, context often leaves too | Moderate, depends on documentation | Higher, roles usually covered twice | Moderate, the lead keeps the context |
| Release and QA process | Varies widely by individual | Usually defined, sometimes light | Normally formalised with review and staging | Set by the lead, applied unevenly |
| Tracking platform deprecations | Rarely contracted for | Often informal | Commonly part of the retainer | Depends on the lead’s spare capacity |
| Strategy and prioritisation | Client-led | Shared within one discipline | Agency can own it when scope allows | Internal, keeps commercial judgment in-house |
| Speed on small isolated changes | Often fastest when the person is free | Fast | Slower, review adds steps | Fast for anything the lead does directly |
| Cost structure | Lowest rate, variable total | Mid-range retainer or sprint pricing | Highest commitment, broadest coverage | Salary plus variable contractor spend |
| Main limitation | Single point of failure | Gaps outside the core discipline | Paying for breadth a smaller store may not use | Recruitment risk and a narrow bench |
Two rows tend to decide it. Speed favours freelancers more often than agencies admit, since a competent individual with spare capacity usually turns round a small isolated change faster than a team with a review step. Continuity favours agencies, and matters more as the cost of an hour of downtime rises. A brand doing modest volume on a lightly customised theme may be entirely rational to accept single-point-of-failure risk for a lower cost.
Hourly rates are the least reliable part of the comparison. A lower rate applied to work that has to be redone can cost more in a year than a higher rate applied to work that is right first time. Compare annual cost against delivered change, including the internal hours spent coordinating either arrangement.
When Staying with Freelancers Remains the Better Decision
Switching is not automatically an upgrade, and moving for the wrong reason can cost a year. Staying is often stronger when the store is commercially stable, change requests are small and infrequent, and someone inside the business already holds the technical context. In that situation, moving to a Shopify agency for established brands buys coordination the business does not currently need.
Staying may also be right when the gap is strategic rather than operational. If the store runs reliably but nobody can say which improvements matter most, a bounded technical audit can answer that for a fraction of a retainer, and the existing contractors can implement the findings. The case for moving strengthens when structural work is approaching: a move to Shopify Plus, a checkout rebuild, an ERP integration or expansion into new markets. Those need several disciplines at once and a named owner for the sequence. Our article on what can be improved on standard Shopify checkout compared with Plus is worth reading before assuming the supplier is the constraint.
Moving Suppliers Without Losing the Store: An Eight-Step Transition
Most of the damage in a supplier change happens in the first fortnight, and almost all of it is avoidable. This sequence assumes a competent incoming agency, and covers what the business should control regardless.
- Inventory every account before anything is announced. Shopify admin, registrar, DNS, analytics, payment gateways, email platform, app subscriptions, anything billed to a personal card. Assembling that list after a relationship cools is far harder.
- Confirm the business holds the store owner role. Shopify’s guidance on changing or transferring ownership notes the new owner must already be a user in the admin before the transfer completes. Do it while the outgoing supplier is still engaged.
- Take custody of the code. Ask for the theme in a repository the business controls, plus source for any private apps. If the store has never used version control, establishing it is the first technical task, not a later improvement.
- Record behaviour, not just files. Have the outgoing supplier walk through anything unusual: custom discount logic, feed integrations, hard-coded exceptions. An hour of recorded explanation can save days.
- Agree an overlap period rather than a cut-off. A short paid handover window keeps the outgoing contractor reachable, and the relationship civil enough that follow-ups get answered.
- Scope the first 90 days as stabilisation. Access, release process, monitoring and known defects. Starting with a redesign changes the store and the supplier at once, making any movement in performance impossible to attribute.
- Fix the measurement baseline before the first change ships. Conversion rate by device, checkout completion, page performance and revenue per session for a comparable prior period. Without it the arrangement can only be judged on how the reporting feels.
- Write the exit terms at the start. Notice period, what is handed over, in what format, within how many days. Agreeing it while both sides are optimistic is easy, and it is the clause most likely to matter later.
Internal expert input required: add a verified example from a WD Market Shopify handover showing what was missing from the original theme documentation and how long reconstruction took.
Mistakes That Make the Switch More Expensive Than It Needs to Be
- Selecting on hourly rate. The most visible number is among the least predictive. Two very different rates can produce similar annual costs once rework and coordination are counted.
- Ending the previous relationship abruptly. Undocumented knowledge sits with the outgoing supplier, and access to it is worth more than the fees a clean break saves.
- Opening with a redesign. A supplier proposing a rebuild in week one has not had time to learn why the store is the way it is, and a rebuild discards behavioural history that took years to accumulate.
- Treating the onboarding audit as the deliverable. Findings never sequenced into work are an expense, not an asset. Agree at the outset how they become a prioritised plan with owners and dates.
- Leaving the app stack unexamined. Long-running stores often carry apps that are paid for and unused, or that duplicate each other. Reviewing them early frees budget and removes scripts that may slow the store.
These share a root: treating a change of supplier as a procurement event that ends at signature. It is closer to a transfer of responsibility, and the handover decides whether the next two years beat the last two.
Deciding Which Arrangement Fits the Next Two Years
The question was never whether freelancers can build a good Shopify store. Many can. It is whether the arrangement around the work still matches what the business depends on: how quickly a broken checkout must be fixed, whether platform change is somebody’s contracted responsibility, and whether the store stays maintainable if the current people stop answering. Where those answers are comfortable, staying is sensible. Where several of the signs above are recognisable and structural work is approaching, the arrangement is likely the constraint rather than the people. We publish shorter notes on decisions like this on the WD Market LinkedIn page.
From a Freelancer Patchwork to a Managed Shopify Programme
If several of the seven signs are recognisable in your own setup, the next step worth taking is a review of the current store, its access model and its release process, so the decision rests on findings rather than impressions. WD Market’s Shopify development service covers the build and consolidation side, and Shopify support and maintenance covers what a Shopify agency for established brands takes on month to month: monitoring, releases, platform deprecations and a prioritised improvement backlog.
To discuss what a transition would involve for your store, including what should be recovered from the current suppliers before anything changes, contact the WD Market team with a short description of your setup and order volume. Ongoing analysis on these topics is published on LinkedIn.
Questions Established Brands Ask Before Changing Shopify Suppliers
How large does a store need to be before an agency makes sense?
Revenue alone is a poor threshold. A better measure is what an hour of broken checkout costs compared with the monthly difference between the arrangements. Stores taking a steady daily order volume, running a heavily customised theme or feeding an ERP cross that line earlier than turnover suggests. A high-revenue store on a near-stock theme with few integrations may reasonably stay with individual specialists far longer.
Can an agency work alongside our existing freelancer?
Frequently, and it is often the least disruptive route. The requirement is a shared release process, so both parties work from the same source rather than overwriting each other in the admin. Divide responsibility by area rather than by task queue, for example the agency taking checkout and integrations while the freelancer keeps campaign pages, and name one person who decides sequencing when the two collide.
What should we ask for before the current supplier stops working with us?
Account access with the business named as owner, the theme code in a repository you control, source for any custom apps, credentials for third-party services, and a walkthrough of anything unusual. Ask while the relationship is active and paid. Requests made after the final invoice are answered slowly if at all, and reconstructing undocumented logic is a large hidden cost of any supplier change.
Will moving to an agency slow down small changes?
For small isolated tasks, often yes, because review and staging add time that one person editing the live theme does not spend. The trade is fewer changes that have to be reversed, and faster recovery when one goes wrong. If rapid small changes genuinely matter, raise it during evaluation and ask how the agency handles a fast lane for low-risk work.
Is a Shopify Plus partner necessary for a brand on a standard plan?
Not usually. Plus partner status signals experience with larger accounts and with functionality limited to that plan, which matters if a move to Plus is being considered. For a brand staying on a standard plan, the more informative signals are experience with a comparable catalogue size and integration complexity, and whether the agency can show how it reviews and releases work.