
On this page
- The 15 checks we’d use to choose a Shopify developer
- Proof of work: portfolio, scope match and reviews
- Platform depth: Shopify Plus agency experience and current knowledge
- Ownership and access: code, store and handover
- Delivery quality: QA, performance and the team
- Commercials: pricing model, support and exit terms
- Shopify agency red flags we’d walk away from
- Scoring the shortlist: a worked example
- Frequently asked questions
- The short version
To choose a Shopify developer or agency well, score every shortlisted team on the same 15 checks, ask for evidence on each one instead of a promise, and rule out anyone who scores zero on code ownership, QA or work at your level of complexity. The checklist below is the one we would hand a founder who has three proposals on the desk and a launch date in the calendar.
Bad partner choices rarely look bad on signing day. The problems show up in month three: nobody can say which branch is live, the senior developer from the sales call has moved to another account, and “support” means a shared inbox. Every one of those can be checked before you sign.
Still deciding between a freelancer and an agency, or comparing hourly rates? Start with our guide on how to hire a Shopify developer. This post assumes you have a shortlist.
Key takeaways
- Judge evidence, not answers. For each check, ask for something you can open: a live URL, a pull request, a QA report, a support log.
- Shopify’s Partner tiers change on January 1, 2027. Select, Plus and Premier retire, and Silver, Gold, Platinum and Diamond are based on attributed GMV and shop counts. A tier tells you about volume, not code.
- Plus experience only matters if you will use Plus features. Ask for a checkout UI extension or Shopify Function the team shipped, not a logo wall.
- Match the pricing model to how certain your scope is. Fixed price on a vague brief usually ends in change requests.
- Weight the checks, score 0 to 2, and treat a zero on a heavy item as a knockout. In our hypothetical example, the agency with the best portfolio came second.
The 15 checks we’d use to choose a Shopify developer
The checklist covers five areas with three checks each: proof of work, platform depth, ownership, delivery quality and commercials. Each check has the evidence to ask for and a weight from 1 to 3 for how expensive it is to get wrong. The weights are our judgment from projects we have inherited, so adjust them for your store.
| # | Check | Evidence to ask for | Weight |
|---|---|---|---|
| 1 | Live stores they actually built | Three live URLs and what they did on each | 3 |
| 2 | Work at your complexity | A project with similar catalog size, integrations and markets | 3 |
| 3 | Verified reviews and references | Two client contacts, verified review profiles | 2 |
| 4 | Shopify Plus experience, if you need it | A checkout extension or Function they wrote | 2 |
| 5 | Current platform knowledge | What changed in the last 12 months and what they did about it | 2 |
| 6 | Build-versus-app judgment | A recommendation to not build something | 2 |
| 7 | Code in your repository | Repo in your GitHub organization, real commit history | 3 |
| 8 | Scoped store access | A collaborator request with a named role | 1 |
| 9 | IP assignment and handover pack | Contract clause, sample handover document | 2 |
| 10 | A written QA process | A real test plan or QA report from a past launch | 3 |
| 11 | Speed and accessibility targets | Acceptance criteria written into the SOW | 2 |
| 12 | A named team | Names, seniority and hours per week for the build team | 2 |
| 13 | Pricing model that fits your scope | Hours by phase, change-request rate | 2 |
| 14 | Support with defined response times | Severity levels, a warranty window, a sample monthly report | 2 |
| 15 | Clean exit terms | Termination clause and handover timeline | 1 |
Score each check 0 (missing), 1 (claimed but thin) or 2 (shown with evidence), times the weight. The maximum is 64.

Free download: we turned the 15 checks into a one-page printable scorecard with tick boxes and a column for each agency on your shortlist. Bring it to your discovery calls. Download the Shopify partner checklist (PDF)
Proof of work: portfolio, scope match and reviews
A portfolio proves something only when you can open the store, see what was built and call the client. Screenshots and logo walls prove the agency has a designer.
1. Live stores they actually built
Ask for three live URLs and a sentence on what the team did on each: full build, redesign, or a few sections. Then check. Open the store, open your browser console and type Shopify.theme. The returned object includes schema_name, which usually names the theme the store started from. A “fully custom build” that reports Dawn with a new color scheme is not a lie, exactly. It is a theme customization, and should be priced like one.
Then open each store on a mid-range phone, which is where most of its customers buy.
2. Work at your complexity
Ask for one project that looks like yours in the ways that create work: catalog size, number of integrations (ERP, 3PL, PIM), number of markets, B2B or subscriptions. Twenty beautiful 40-product stores do not prove a team can sync 12,000 SKUs with NetSuite. Ask what went wrong on that project. “Nothing” means it was small or they are not telling you.
3. Verified reviews and references
Call two references, and choose them yourself from the portfolio list if you can. On review sites, read how the reviews were collected. Clutch says it checks each reviewer’s identity and work history, and publishes reviews it could not confirm with a “Not Verified” label. Discount those.
Partner tiers are a different signal, and they are changing. Per Shopify’s Partner Program page, the Select, Plus and Premier tiers retire on January 1, 2027. From that date service partners are ranked Silver ($25K attributed GMV and 1 shop over one year), Gold ($5M and 5 shops), Platinum ($50M and 20 shops) or Diamond ($500M and 25 shops), the last three measured over four years. That measures how much merchant revenue a partner is attached to. It says nothing about whether their Liquid is maintainable.
Watch out: from January 2027, an agency still calling itself a “Shopify Plus Partner” in a proposal is describing a tier that no longer exists. Ask which 2027 tier they hold. Partners could preview it in their Dev Dashboard from November 4, 2026.
Platform depth: Shopify Plus agency experience and current knowledge
Platform depth means knowing what Shopify can do this quarter, what only Plus can do, and when an app beats custom code. Test it with questions about recent changes, not years of experience.
4. Shopify Plus experience, if you need it
Plus experience matters when your project uses Plus-only features. The clearest example is checkout. Shopify’s checkout UI extension docs state that extensions for the information, shipping and payment steps are available only to stores on Plus. If you need custom checkout fields, delivery logic in Shopify Functions, B2B catalogs or expansion stores, ask to see one extension or Function they wrote.
If you are on Grow or Advanced with no plans to move, weight this check at 1, or drop it. Our Shopify Plus vs Advanced comparison covers when the upgrade is worth it.
5. Current platform knowledge
Ask: “What changed on Shopify in the last 12 months, and what did you do about it?” A good team answers without thinking. For 2026 it should include the Shopify Scripts shutdown and the end of the old Thank you and Order status pages for non-Plus stores, both on the deprecation calendar in our Shopify store maintenance checklist. For app work, also ask how they handle API versions. Shopify releases a new API version every quarter and supports each stable version for at least 12 months. When one expires, requests fall forward to the oldest supported version, which is how integrations break quietly.
6. Build-versus-app judgment
The partner you want will talk you out of some work. Ask which part of your brief an existing app would cover, and which app they would remove. If every answer is “we’d build that,” the proposal is sized for their capacity, not your needs.
Ownership and access: code, store and handover
You own the repository, the store, the apps and the code. The partner gets scoped, removable access to each. Our hiring guide covers the contract clauses, so here we stick to selection.
7. Code in your repository
Ask to see the commit history on a recent project. With Shopify’s GitHub integration, each connected branch maps to a theme, and edits made in the admin are committed back to the branch by a shopify bot account. Only members of your GitHub organization with write access can connect branches, not outside collaborators, so the partner joins your org. And the code editor gives no conflict warnings, so ask how they keep your team’s theme editor edits from clashing with theirs.
8. Scoped store access
A good partner sends a collaborator request and asks you for the 4-digit code under Settings > Users > Security. Per Shopify’s collaborator documentation, you can edit the requested role before approving, collaborators do not count toward your staff limit, and they expire after 90 days without a login. A partner who asks for your owner password scores zero.
9. IP assignment and handover pack
Look for a contract clause that assigns custom code and designs to you on payment, then ask for a sample handover document. Ours includes the repo location, a list of custom sections and snippets, every app with its billing owner, metafield definitions, integration credentials and where they are stored, and known issues. If a partner has never written one, you will be the one reconstructing it later.
Delivery quality: QA, performance and the team
Delivery quality is what proposals describe most vaguely and clients test least. Ask for documents from a real launch, because a QA paragraph costs nothing to write.
10. A written QA process
Ask for a QA report from a recent launch. Look for a device and browser matrix, checkout test cases for each payment method and discount type, and a defect log with severities. On the code side, ask whether they run Shopify’s Theme Check linter in CI. It catches Liquid errors, unused snippets, deprecated tags and some performance issues before code merges. Then check the hours. A 300-hour build with 10 hours of QA has not budgeted for testing.
11. Speed and accessibility targets in the contract
“We build fast stores” is not a target. Ask what number they will commit to, and how it is measured. Shopify’s performance testing guide says to run Lighthouse at least three times per URL and take the median, since one run can swing 5 to 10 points, and to use the admin’s Web Performance dashboard for field data from real Chrome users. The same guide notes that Theme Store themes need an average Lighthouse performance score of at least 60 across home, product and collection pages. Treat that as the floor, not the goal. We write lab targets per template into the SOW and track field Core Web Vitals after launch.
For accessibility, ask which WCAG level they test against and whether anyone uses a screen reader. If you sell into the EU, it is a legal requirement.
12. A named team
Ask for the names, roles and weekly hours of the people who will build your store, and put them in the SOW with a clause that says you are told before anyone is swapped. Two to three shared working hours a day is enough for most projects. Zero is not. Large agencies often lose points here: the portfolio is excellent, and the people who built it are on someone else’s account.
Commercials: pricing model, support and exit terms
The right pricing model depends on how certain your scope is. Fixed price protects you only when both sides already agree on what “done” means.
13. Pricing model that fits your scope
| Model | Who carries overrun risk | Fits when | Watch for |
|---|---|---|---|
| Fixed price | Agency, on paper | Small, fully specified work: a theme setup, a set of sections | Padding, and a strict reading of scope that turns every question into a change request |
| Paid discovery, then fixed price per phase | Shared | Custom builds and migrations where the brief is still forming | Discovery that produces slides instead of a priced backlog |
| Time and materials with a cap | Shared, capped | Integrations, apps, anything with unknowns | No weekly burn report, so you find out at the cap |
| Monthly retainer | You | Post-launch CRO, maintenance, a steady feature backlog | Unused hours that vanish, and no monthly report of what was done |
Whatever the model, ask for hours by phase and the change-request rate. Hours are what you can compare. When our Shopify development team quotes a custom build, discovery is a separate, paid line, and the build estimate that follows is broken down by phase and template. For what those hours typically cost by region, see our Shopify development cost breakdown.

14. Support with defined response times
Get support terms in writing, with severity levels. Our minimum (a view, not an industry standard): a store that cannot take orders gets a response within an hour, any day of the week; a broken feature with a workaround within one business day; cosmetic issues in the next scheduled release. Add a warranty window for bugs in delivered work, commonly 30 days, and ask for a sample monthly report from an existing retainer client. It should list what was updated, tested and fixed, not just hours used.
15. Clean exit terms
Check the notice period and how fast the partner must hand over code, credentials and documentation. Thirty days’ notice with a handover inside ten working days is reasonable. A partner who resists a clear exit clause is telling you how they keep clients.
Comparing proposals right now?
Send us the proposals on your desk and we will score them against this checklist and tell you what each one leaves out, free and without a pitch attached.
Get your shortlist scored
Shopify agency red flags we’d walk away from
Walk away from any partner that won’t show evidence for a weight-3 check, fixes a price on a vague brief, or keeps an asset your store depends on in its own name. Other patterns we see at proposal stage:
- A fixed price with no discovery. If they priced a custom build from a 20-minute call, they priced an imaginary store.
- Case studies with percentages and no baseline. “Conversion up 140%” means nothing without the starting rate, the period and the traffic.
- Guaranteed rankings or conversion rates. Nobody controls Google or your customers.
- A page builder or private app holding your content. Ask what happens to your pages if the app is uninstalled. If the answer is “they disappear,” that is lock-in.
- Undisclosed app referral fees. Referral income is normal. Hiding it is not.
- No QA line in the estimate. Testing is either missing or hidden in “development.”
Scoring the shortlist: a worked example
A clearly hypothetical shortlist: say a home goods brand on Shopify Advanced, around 1,800 SKUs, a NetSuite integration and two markets, is choosing between three proposals:
- Agency A, well known, strong portfolio, fixed price. Could not name the build team; QA was one paragraph.
- Agency B, a low-cost freelancer collective. Cheapest by far, but code stays in its repository until final payment, and nothing this size in its portfolio.
- Agency C, a mid-size Shopify specialist. Paid discovery, then fixed price per phase. Showed a QA report and a NetSuite project, with fewer verified reviews.
Scored on the 15 checks, Agency A reaches 50 of 64 (78%), losing most of its points on the named team, QA and speed targets. Agency B reaches 28 (44%), and is out regardless, because it scored zero on check 7, a weight-3 item. Agency C reaches 58 (91%), dropping points only on reviews, Plus depth and support detail.

The gap between A and C sits almost entirely in delivery quality, the part a portfolio hides. And B’s price advantage disappears once you count the cost of rebuilding a theme you never got the source for.
From our team: have two people score separately, then compare. A gap of more than one point on a check means your question was not precise enough. Ask it again.
Frequently asked questions
What questions should I ask a Shopify developer?
Ask questions that require evidence: which three live stores did you build and what did you do on each, who exactly will write our code, where will the repository live, what changed on Shopify in the last year and how did you handle it, what would you not build, and can we see a QA report from a recent launch. Specific, checkable answers are what you are listening for.
What are red flags when hiring a Shopify agency?
The biggest red flags are a fixed price quoted without any discovery, no named build team, code kept in the agency’s own repository, case studies with percentages but no baseline, and guarantees on rankings or conversion rates. Also be wary of content locked inside a private app or page builder, and of any proposal that leads with a Partner tier Shopify has retired.
Is a Shopify Partner tier a good way to choose an agency?
A Partner tier is a useful filter but a poor decider. From January 1, 2027, Shopify ranks service partners as Silver, Gold, Platinum or Diamond based on attributed GMV and shop counts. That shows commercial track record and client volume. It does not measure code quality, QA or who will work on your store, so score those separately.
Do I need a Shopify Plus agency?
You need a Shopify Plus agency only if your project uses Plus-only features, such as checkout UI extensions on the information, shipping or payment steps, B2B catalogs or expansion stores. For a store on Grow or Advanced with no upgrade planned, a strong standard Shopify partner is usually better value. Ask any Plus agency to show an extension or Function it wrote.
Should I choose a fixed-price or hourly Shopify developer?
Choose fixed price when the scope is small and fully specified, and time and materials with a cap when the work has unknowns such as integrations or custom apps. For most custom builds, paid discovery followed by fixed price per phase works best, because it prices a defined backlog instead of a guess.
How many Shopify agencies should I compare?
Compare three agencies in depth. Fewer gives you no benchmark for price or approach, and more turns evaluation into a part-time job. Start with a longlist of six to eight, cut it using portfolio and scope match, then run the full 15-point scorecard and discovery calls on the final three.
The short version
Before your next call with a prospective partner, print the scorecard, decide your weights, and write down the evidence you will ask for on checks 1, 2, 7 and 10. Those four separate partners who can build your store from partners who can pitch it. Score separately with a colleague, and treat a zero on any weight-3 check as the end of the conversation, however good the price looks.
If you want a partner who will hand you the evidence for all fifteen, we are happy to be scored. We have built and supported stores for brands like Delsey, EcoFlow, Kerbholz and Maker & Moss, with strategy in Alpharetta and engineering in Mohali. Look through our shopify development services, or ask us for a sample handover pack and QA report before you decide.




