How to Choose a Shopify Agency Without Getting Burned

Last updated October 23, 2026 · By Chad Norris

Disclosure: the author runs Tide Flow, an Amazon advertising agency, and sells Shopify apps under Orbitloft. Neither is a Shopify store-build agency, but both are services businesses, so read the disclosure page before weighing anything here. This guide names no winners.

The short answer: the agency matters less than the scope. Write down what “done” means, get two or three written proposals against that same page, ask each firm who will do the work, and pay against milestones. Everything below is how to do that, and how to spot the proposals that will cost you more than they say.

Start with the scope, not the agency

Most “bad agency” stories are really “vague scope” stories. Before you contact anyone, write one page that answers:

• The outcome. “Rebuild our product page so the size guide, reviews and subscribe option sit above the fold” is a scope. “Make the store better” is not.
• The deliverables. Theme sections, custom templates, app installs and configuration, data migration, redirects, tracking, a staging store.
• The constraints. Your current theme, the apps you already run, your deadline, your budget ceiling.
• What you will provide. Brand assets, copy, product data, decisions. Late client inputs are a common source of slipped timelines.

If you cannot write this page, the first paid engagement should be a short discovery or audit, not the build.

Fixed price, hourly or retainer

The pattern that holds up best for larger work is a fixed price per milestone, each with a written acceptance test, so a misunderstanding costs you one phase instead of the whole project.

What to ask every firm you talk to

Put the same questions to each candidate and compare the answers side by side.

1. Who does the work? Names, roles and whether they are employees or subcontractors. The person on the sales call is often not the person writing code.
2. Show a comparable project and let me talk to that client. Comparable means similar size, similar apps, similar problem. A reference you can actually call beats a logo wall.
3. How do you handle access? Shopify stores can grant outside developers staff or collaborator access; you should know exactly which permissions are given and be able to revoke them. Ask for named accounts, never a shared login.
4. Where does the work happen first? Changes should be built on a duplicate theme or a development store and previewed before anything goes live. A firm that edits your live theme is taking a risk with your revenue.
5. What do I own at the end? Custom code, design files and documentation. Get it in writing.
6. What happens after launch? Bug-fix window, how long it lasts, and what it costs afterward.
7. What will you not do? Good firms say no to some requests. A firm that agrees with everything has not scoped carefully.
8. How do you estimate? Ask them to walk through how the number was built. Vague answers here predict surprises later.

Red flags in proposals

• A price with no line items, or a timeline with no dependencies.
• Promises of specific revenue or conversion results. No one controls your demand, and a firm that guarantees outcomes is selling confidence, not information.
• Pressure to sign before you can compare other proposals.
• A request for your store owner login or for payment in full up front on a custom build.
• No staging or review step before launch.
• A portfolio that shows only screenshots with no live links, so you cannot check speed, mobile layout or accessibility yourself.
• Reluctance to put ownership of the code in the contract.

None of these is proof of bad faith alone. Two or three together is a reason to keep looking.

Check the work yourself

Open two or three of the live sites a firm cites and test them on your phone. Look at load time, the cart, the checkout hand-off and whether the layout holds up at narrow widths. A review page cannot tell you that; five minutes of use can. If the firm’s showcase work is on a different platform, ask what Shopify-specific projects they have shipped recently.

How agency lists are biased (ours included)

Every place you can find an agency has a filter that favors something other than your outcome.

The Partner Directory page describes four tiers (Select, Plus, Premier and Platinum) based on “history of experience and proven success on Shopify.” That is useful as a signal of track record on the platform. It is not a statement that a higher tier will fit your project or budget; a small specialist at a lower tier can be the right match for a narrow job.

Our own incentives are listed on the disclosure page: we earn nothing from the agencies in the directory today, we do not sell placement, and we do not rank. If that changes, the page and the card will say so.

A simple selection process

1. Write the one-page scope.
2. Build a candidate list of five or six from at least two different sources (directory, referral, search).
3. Send each the same brief and the same questions, with a deadline for replies.
4. Shortlist two or three on answer quality, not price. Slow, vague replies before you have paid are what you will get after you pay.
5. Hold a call with each; ask to speak with a recent comparable client.
6. Compare proposals line by line against your scope. Differences usually reveal who understood the problem.
7. Start with a paid, bounded first milestone before committing to the whole project.

Mistakes that burn merchants

• Hiring on design portfolio alone. A beautiful theme with slow pages and broken app compatibility costs you sales.
• Letting the agency pick your apps. Apps carry monthly fees and affect speed. Ask for each recommended app to be justified in terms of your goal. Our guide to Shopify subscription apps shows how quickly recurring fees stack up.
• Skipping the headless question. Some firms steer merchants toward custom front ends because they are more billable. Read when headless is worth it first and ask the firm to justify why a theme will not do.
• No named internal owner. Someone on your side must make decisions within a day or two, or the schedule slips regardless of the firm.
• Paying for everything before launch. Hold back a final payment until acceptance tests pass on the live store.

If you are in the Amazon world instead, the parallel decision is covered in how to choose an Amazon PPC agency.

Reading a proposal line by line

A good proposal makes it easy to compare. Look for these elements, and ask for any that are missing:

• Assumptions. Which apps, which theme, how many templates, how many languages or currencies. Differences in assumptions explain most price gaps between bids.
• Exclusions. What the price does not include: copywriting, photography, data cleanup, redirects from an old platform, ongoing support.
• Dependencies. What the firm needs from you and by when.
• Testing. Which browsers and devices are checked, who tests checkout, and how bugs found after launch are handled.
• Handover. Documentation, training for your team and where the files live.

If two proposals differ by a large amount, do not just choose the cheaper one. Ask both firms to explain the gap against the same scope. One of them has usually assumed something different, and finding out which is a cheap way to learn how carefully each firm reads.

What to do this week

Write the one-page scope, pick three sources, and send the same brief to five firms with your eight questions. Compare the replies, then run a small paid first milestone with the most credible one. Confirm all directory details and program descriptions on the official Shopify Partners directory before you rely on them.