Sourcing Shopify developers takes an afternoon. The Partner Directory, Upwork, two founder WhatsApp groups and one LinkedIn post will get you twenty candidates by dinner. Screening them is the part that decides whether you end up with a store that survives a theme update, and it is where almost every founder we speak to has already been burned once. So the useful answer to how to hire a Shopify developer is not a list of job boards. It is a list of questions whose wrong answers tell you, in about ninety seconds, that this person will hardcode your homepage and break your Liquid on their first sprint.
Below are the twelve we use, plus the answers that should end the call.
Where to find Shopify developers, and why the list is shorter than it looks
Four places produce most real candidates. The Shopify Partner Directory lists agencies and individual Partners with verified reviews attached to actual jobs, which is the only public signal that money changed hands. Upwork and Toptal give you individuals fast, with escrow, and with a rate premium for the platform's cut. The Shopify Community forums and the Shopify Developers Discord surface people who solve Liquid problems for strangers in public, which is a better proxy for competence than a portfolio PDF. And referrals from other founders in your category, which are the highest hit rate and the slowest.
Fiverr works for a one-off gig with a crisp spec: swap a font, fix a broken mega menu, add a size chart. It does not work for anything that touches your cart or your data model, because there is no continuity when the work needs revisiting in March.
Hiring an expert from the Partner Directory is worth doing properly. Filter by service and by language, then ignore the star rating for a minute and read the review text. You want reviews that mention specifics: a migration, a Recharge integration, a speed job, a B2B portal. Reviews that only say "great communication" tell you the person answers Slack, which is table stakes. Also check what the Partner actually sells. A directory listing that offers everything from logo design to Klaviyo flows to headless builds is a reseller, and your ticket may be subcontracted to someone you never interview.
The five questions that expose a theme-breaker
These are technical, but you do not need to be technical to score them. You are listening for whether the answer names a Shopify primitive or describes a hack.
1. "Where does your custom CSS live?" A bad answer edits base.css or drops a <style> block into theme.liquid. A good answer creates a separate asset file, or keeps changes inside section-level styles, and can explain what happens to those edits when the theme updates. This one question sorts about half the field.
2. "I need a 'Care instructions' field on 400 products. How?" If they say "put it in the description" or "use tags", stop. The answer is metafields with a defined namespace and type, surfaced in Liquid, editable by your ops person in the admin without touching code. Tags as a data store is the single most common cause of a catalogue nobody can filter three years later.
3. "My marketing manager needs to change this banner copy herself. Build it so she can." Watch for the developer who is happy to be your permanent copy-paste button. Every custom section should ship with schema settings and blocks so non-technical staff can edit text, images, links and order. Developers who skip schema are protecting their retainer.
4. "We're on Dawn. In eight months we'll want the newest version. What does that upgrade look like after your work?" The honest answer involves version control, a record of which core files were touched and why, and a warning that a major theme upgrade after heavy customisation is a rebuild of the diff, not a button. Anyone who says upgrades are painless has not done one.
5. "How would you edit our checkout?" The trap answer is checkout.liquid. Non-Plus stores never had it, and Shopify has retired it for Plus in favour of checkout extensibility, so a candidate who reaches for it is quoting a workflow that has been dead for years. Good answers name checkout UI extensions, Shopify Functions for discounts and delivery customisation, or Plus-only branding controls.
Four questions about process, because that is what actually fails
6. "Where do you work while you build?" Duplicated theme or a development theme via Shopify CLI, previewed on a share link, published only after sign-off. If the answer is "I'll just make the changes and you check", you are the QA environment.
7. "Who owns the code, and where does it live when we part ways?" Your GitHub or GitLab, your repo, your Shopify collaborator access revoked cleanly. Developers who keep the repo on their own account are holding a hostage they may never intend to use.
8. "Our store has 34 apps. What do you do about that?" Any developer who says app count is not their concern is not thinking about your Largest Contentful Paint. A good one asks which apps are load-bearing, checks what each injects into the storefront, and expects to remove three or four before writing a line of new code.
9. "Give me one thing you'd refuse to build." The candidates worth hiring have an answer ready, usually something like a custom cart drawer that duplicates cart logic, or a bespoke subscription engine when a good app exists. A developer with no refusals will build whatever the brief says and hand you the bill for the consequences.
Three questions for anyone selling to an Indian D2C brand
10. "How do you handle COD on a partial-prepaid campaign?" Real experience shows up fast here. They should know that COD gating by pincode, cart value or product tag lives at the payment-customisation layer, and that combining it with a partial prepaid flow usually means a Razorpay or Shiprocket app rather than custom code, plus a Function for the delivery rule.
11. "Where do GST and HSN codes go?" Correct answer: metafields plus your invoicing app, with a note that Shopify's tax engine handles Indian GST slabs unevenly on discounted and bundled line items and that invoices are best generated outside the storefront. A blank stare here means your CA will be doing manual work every month.
12. "It's the second week of October. Traffic is 6x. What broke?" Practitioners answer with specifics: an app script blocking render, collection pages with unpaginated Liquid loops, image assets served at 2400px, a third-party review widget timing out. People who have never worked a festive peak talk about "scaling the servers", which Shopify does for you.
How to vet a Shopify developer's portfolio in ten minutes
Screenshots prove nothing. Open the live URLs and do four things.
Run PageSpeed Insights on a product page, not the homepage, because product templates carry the app scripts. Then view source and search for the theme name: if six of their eight portfolio stores are lightly restyled Dawn or Impulse, they are a theme configurer, which is a legitimate and cheaper service, just not development. Third, resize the browser to 375px and use the filters and cart. Fourth, ask which specific parts of that store they built, and then ask for the name of the person who signed off, because a developer who worked on the project can tell you who they reported to.
If speed is the reason you are hiring, get a baseline before the interviews so you can judge answers against your own numbers. A store sitting at a 2.4s LCP on mobile does not need a rebuild, it needs the render-blocking scripts and the hero image fixed, and that is a two-week job. Our Shopify speed optimization work almost always starts with removing things, and if you want the score tracked continuously afterwards, SwiftStore handles the monitoring so regressions surface before your CAC does.
Freelancer, agency, part-time or in-house: the arithmetic
Here is the version nobody selling you agency retainers wants to write down.
A competent India-based Shopify freelancer bills roughly ₹800 to ₹2,500 per hour. Agency rates run ₹1,500 to ₹3,500 per hour, or fixed-price for defined scope. A full-time in-house Shopify developer at ₹12 LPA costs about ₹1,00,000 a month, and at roughly 160 working hours that is ₹625 an hour, the cheapest per-hour number on the page.
The catch is utilisation. If your store needs 30 hours of development a month, that in-house developer costs you ₹3,333 per productive hour, worse than the most expensive agency, and by month four they are bored and job-hunting. The break-even is somewhere around 100 hours of genuine monthly work, which most brands do not have until they are running multiple storefronts, a subscription programme or a serious B2B channel.
So: freelancer for defined, contained jobs. Agency when the work spans design, Liquid, apps, analytics and email and you do not want to be the integration layer. Part-time retainer, say 20 to 40 hours a month, for the long middle where most D2C brands live. Full-time hire only when you can name the next six months of roadmap. If the choice is close, the deciding factor is bus-count: one freelancer on holiday during Diwali week is a business risk, and it is why some brands keep a freelancer for features and an agency on standby. We break rates and scope down further on Shopify development cost in India, and if you would rather skip the sourcing entirely you can hire a Shopify developer through us on a monthly block.
A worked example, because "it depends" is useless. Say you want a build-your-own-hamper section: three tiers, live price calculation, custom line item properties, and an editable banner. That is roughly 18 hours of build, 4 hours of QA across devices, 2 hours of handover. At ₹1,800 an hour, 24 hours comes to ₹43,200. A quote of ₹12,000 means someone is hardcoding it into a template with no schema, and you will pay again in six months. A quote of ₹1,80,000 means it has been scoped as a bespoke app when it did not need to be.
Hiring for theme customization specifically
Theme work is the most commonly mis-hired category, because it sounds simple and is not. Two rules. First, write the brief as behaviour, not appearance: "the collection page shows filters for fabric, sleeve length and occasion, with counts, and the filtered URL is shareable" beats "make the filters nicer". Second, insist that anything reusable ships as a section with schema, so your team can place it on new pages without a ticket.
Be honest about whether you need a developer at all. If the requirement is filtering and search on a 3,000-SKU catalogue, an app like FilterPro costs a fraction of custom faceting and does not become your maintenance problem. We tell clients to buy before build maybe six times out of ten, and the ones who listen have simpler stores two years later.
A Shopify developer job description you can post today
Keep it short and disqualifying. Long JDs attract applicants, not candidates.
- Role: Shopify developer, part-time contract, 25 hours per month, remote, IST overlap of at least four hours.
- Store context: Shopify Basic, Dawn-based custom theme, 1,400 SKUs, ₹X crore annual GMV, Razorpay plus COD, Shiprocket, Klaviyo or Omnisend, one warehouse. Say the real numbers. Vague context gets vague quotes.
- First 60 days: reduce mobile LCP on product templates, rebuild the collection filtering, migrate three hardcoded content blocks to editable sections, remove unused apps.
- Must have: Liquid, metafields and metaobjects, Shopify CLI, Git, Ajax cart work, and at least one checkout extension shipped to production.
- Nice to have: Shopify Functions, GraphQL Admin API, headless with Hydrogen, Indian tax and COD flows.
- How to apply: send two live store URLs, say exactly what you built on each, and answer this: how do you make a custom section editable by a non-technical marketer? Applications without that answer will not be read.
That last line does more filtering than any interview round.
If you are a small business hiring your first developer
Do not hire for the store you want in 2027. Hire for the next four problems, in writing, with a budget cap. Start with a paid trial task of 6 to 10 hours: one real ticket from your backlog, priced fixed, delivered on a preview theme. What you learn from watching someone ship a small thing properly is worth more than three calls.
Two more things founders under ₹50 lakh in annual revenue get wrong. They buy a premium theme and then pay to undo half of it, when Dawn plus 20 hours of section work would have been faster and cheaper. And they hire a developer to fix a conversion problem that is actually a product photography and pricing problem, which no amount of Liquid will solve.
If you are unsure whether your next spend belongs in development, apps, or somewhere else entirely, send us the store URL through the free audit and we will tell you which of the three it is. If the answer is "you do not need a developer yet", that is what you will hear.