Do you need to hire a Shopify developer? Do you need to hire a Shopify developer?

The verdict

The eight jobs that genuinely need a developer

Nothing here is opinion. Each row is Shopify telling you, on its own site, that the job takes code.

Four of those rows are Shopify's own words pointing you at a developer — "requires development knowledge," "require custom code," "Hire a Shopify Partner," "only if you know HTML and CSS." One is Plus-only outright: checkout UI extensions on the information, shipping and payment steps. Functions are Plus-only when they ship in a custom app you commission; a public App Store app carrying a function runs on any plan. None of the eight is a taste question. Every one is a code artifact with a toolchain attached — CLI, Wasm, Liquid, GraphQL, TOML, Node.

If the thing you want done isn't on that list, you probably don't have a developer problem. You have a tooling problem, and it's cheaper.

And know the free-help budget before you buy anything. Shopify publishes it: 60 minutes of design support on Basic and above, "tracked in 15 minute increments," against a written won't-touch list (quoted in the FAQ). Past that hour, everything is somebody's billable time.

What those jobs cost — and what the price is really a function of

Shopify's Partner Directory is a price list, and almost nobody quotes it. Two partner profiles, read the same day, pricing the same ladder of jobs:

(shopify.com/partners/directory/partner/ultimate-apps and /partner/it-geeks, both read 26 July 2026. Self-declared starting prices, not averages. † IT-Geeks' custom-apps figure is published on the directory's Custom apps and integrations listing page — shopify.com/partners/directory/services/development-and-troubleshooting/custom-apps-integrations — not on its partner profile, which reads "Contact for pricing" for that service.)

Read the left column down its five priced services — $45 → $100 → $500 → $550 → $1,000 — and you have the whole argument of this page, priced by a Shopify Partner on Shopify's own website. The cheap end is fixing and restyling what already exists. The expensive end is building something new that talks to another system.

Read across and the same directory service line — "Store build or redesign" — starts at $500 with one partner and $10,000 with the other, on the same day. The label is identical; the scope behind it plainly isn't, and neither is the tier. Which is the honest answer to "how much does a Shopify developer cost?": the number tracks who you hire at least as tightly as what you ask for.

The directory's listing pages say the same thing at scale, all read 26 July 2026. Theme customisation — 3,994 partners, published starts from $49. Custom apps and integrations — 2,807 partners, published starts spanning $85 to $2,500, with most page-one partners in the $1,000–$2,500 band and the rest "Contact for pricing." Headless commerce — 1,830 partners, and on page one 16 of 16 published no price at all.

If you're buying hours instead of a project: US and Canadian agencies "typically bill $120 to $200/hour" (FactoryJet, 7 Jun 2026); CartCoders puts US/Canada senior and Plus work at $120–$200 too, and Blackbelt puts a Plus Certified Partner at $175–$300 (16 Jun and 8 May 2026). Shopify's own blog quotes Upwork at "$15–$35/hr (median rates)" for front-end and "$55+/hr" for full-stack (shopify.com, 27 May 2026) — and Upwork's own methodology footnote calls its figures "typical hourly rate ranges of certain historical contracts worldwide" (read via Wayback snapshot, 24 May 2026; upwork.com blocks automated fetches). So that is a global median, not a US rate. Don't compare the two as if they were one range.

If you're buying a retainer: Netalico publishes its own at "$2,700–$10,000 per month … with the average client around $4,500 per month" (read 26 Jul 2026). Blackbelt Commerce publishes a $3,000/mo floor (8 May 2026).

If you're hiring in-house: BLS puts web developers and digital designers at a $95,380 median in its 2026 Handbook — May 2024 wage data, wages only, no benefits or payroll tax. Its March 2026 employer-cost release puts wages at 69.9% of total compensation, so by arithmetic (not a BLS figure) that's roughly $136,000/yr loaded, about $11,400/month, for one generalist who is not a Shopify specialist.

The software that sits on top of the store — testing app, replay tool, analytics, page builder — is a separate bill with its own arithmetic. It's priced line by line in what a Shopify CRO stack actually costs.

What Shopify already does without a developer

Skipping this would make the rest of the page a brochure. All of it is real, and all of it narrows the second column.

The theme editor is genuinely capable. Merchants can "add, remove, edit, and rearrange content," and theme settings cover colours, typography, page width, section spacing, animations, cart style, buttons, cards, drawers and badges — plus a Custom CSS field. The boundary is Shopify's own sentence, and it's a fair one: "If your theme doesn't include settings for the changes that you want to make, and you're comfortable with HTML, CSS, and Liquid, then you can edit your theme code."

Shopify A/B tests themes and checkouts now. Since 5 June 2026, Rollouts can "schedule, publish, and A/B test new themes and checkout and customer account configurations" — including a test between two entirely different checkout setups. Anyone still telling you Shopify can't test its checkout is working from a stale page. Where it stops: experiments need the Grow plan or higher, cover online-store checkouts only, swap one whole configuration for another rather than testing an element, and report one metric on a checkout test — checkout conversion rate. "You can't customize the available metrics for your rollout." No revenue per visitor, no audience targeting, no significance reporting.

Shopify Flow is free and no-code — "a free app available on the Basic, Grow, Advanced, and Plus plans" — and it automates a real amount of ops work that used to be quoted as a custom app.

Page-builder apps already remove some of this cheaply. PageFly is $0 for one page or section and $24/mo on Builder, billed monthly — its pricing page carries no annual toggle. Replo is $0 free and $99/mo on Starter billed annually: its pricing page defaults to annual and advertises "Save 17%," so the monthly-billed rate is higher. Both read 26 July 2026. If your only unmet need is a landing page, that is the honest answer and it costs less than an hour of agency time.

The jobs that stopped needing a developer

This is the second column, and it is split — because half of it is true on the Shopify store you have today, and half of it needs the store to move. Blurring those two is how vendor pages lose readers, so here they are separately.

Two things this table refuses to blur.

One. On a store that stays on Shopify, Mtrix does not edit your theme, does not vary your prices and does not replace your checkout. Those are reasons to move, not things you bolt on. What runs on the store you have is measurement and experimentation — replay, analytics, experiments, error tracking and Core Web Vitals — installed as an npm package or a CDN script. The platform-versus-platform argument is a separate decision, and it's made properly in Mtrix vs Shopify.

Two. Authoring the variant's markup on a Liquid theme is the one step we checked before claiming it. Applying an experiment adds a CSS class and two data attributes — and what that class does is authored in the Mtrix editor. A non-technical user can build the variant's visual change without writing CSS or JS.

If you do move, the import runs off your GitHub repo where there is one, off a scrape of the live site where there isn't, or off Figma for a brand starting fresh — and a person at Mtrix finishes the setup rather than handing over a half-built store. The most complex stores are fully live within a week.

What survives it: order and customer history both transfer — imported during setup, not left behind in the old admin. Third-party app functionality and subscription logic are supported natively — the jobs they do, not app-for-app — and each store's specific stack is mapped on the migration call before anything is committed.

What Mtrix does not take off your developer's plate

Nobody at Mtrix writes your Rust function. There is no managed service and no done-for-you tier — onboarding, email, a shared Slack channel and a named contact on the account, and then your team runs the programme. If you need someone else to run experiments for you, that's a retainer, and the arithmetic is here.

Shopify's app ecosystem has thousands of apps. Mtrix has nothing comparable. If five niche apps run your subscriptions, your reviews and your 3PL, staying on Shopify is a rational decision and you should make it without guilt.

Rendering a variant server-side is still an engineering task. Mtrix decides the allocation on the server for every test, always. Rendering the chosen variant on the server — the part that removes the flash of original content — needs an SSR integration: roughly 35 lines in a server load function on Hydrogen, Next.js or SvelteKit, and it isn't possible on a Liquid theme at all. That's a developer, once.

Mtrix is web only. No native iOS or Android SDK. If your revenue runs through an app, this is not your tool.

Session replay and error data are retained for one month (events and purchases run three years). If you need a year of replay for compliance, price that in.

And a good theme developer who already knows your store is worth keeping. The point of this page isn't to fire anyone. It's to stop routing the $45–$100 tier of work through a $120–$200/hr queue, so the person you're paying spends the week on the eight jobs only they can do.

Hire, retain, or neither

Hire a developer — project-priced — if your job is on the eight-job list above. Get a fixed quote and know what's in it: fixed-price quotes carry a 20–40% risk buffer that "the agency pockets if nothing goes wrong" (BoltRamp, 15 Jun 2026). That's the price of certainty, and worth paying when the scope genuinely is uncertain.

Put them on retainer if you consistently exceed 15–20 hours of Shopify work a month — BoltRamp's published threshold for when "monthly retainers become economically favorable" (15 Jun 2026). Below that you're pre-paying for capacity you won't use.

Hire nobody, and buy tooling instead, if what's actually blocking you is page changes, landing pages, tests and measurement. That's the majority of most queues. AgencyPitch's published $5M–$25M package is $35,000 to build plus $4,500/mo on a six-month minimum, and its retainer line item reads "Conversion optimization retainer (1-2 A/B tests/mo)" (23 Apr 2026). Mtrix puts no cap on experiments on any plan, and once it's on your store a test goes live about five minutes after you configure it. It bills monthly or annually, flat across three tiers, with nothing metered on sessions, orders or profiles — so the bill doesn't move when the traffic does.

And if your budget is genuinely zero, take the free route and come back later: Shopify Flow costs nothing, PageFly and Replo both have $0 tiers, and Microsoft Clarity gives you session recordings and heatmaps free — "Clarity is a free service forever. You never encounter traffic limits" (learn.microsoft.com, read 26 Jul 2026). Mtrix has no free tier — one month free, no credit card, then it's paid. Come back when the tests are worth more than the subscription.

FAQ

* Measured across 350,000+ sessions, tested side by side against Shopify's checkout in the same period, over two years of tests. Mtrix's own client results, not a guarantee of what a new store will see.