Short version: pay a fixed price for work that has a finish line, and pay a retainer for work that never does. A gateway migration, a theme rebuild, an Arabic RTL storefront, a Shopify Markets setup for AED and SAR — all of those end. Conversion work, catalogue hygiene, app-stack cleanup, bug triage during White Friday: those don't end, and every attempt we've seen to buy them as a fixed-scope contract has ended in a change-order argument by week five. That's the whole answer to shopify retainer vs fixed price project. The rest of this is how to tell which bucket your actual list of work falls into, and what each costs in this market.
The reason this question is live right now in Dubai and Riyadh: Shopify Payments became available to UAE merchants on the Advanced plan in late August, and a lot of stores that have been sitting on a third-party gateway for years are suddenly scoping a switch. Check your own admin before you plan anything, because eligibility depends on your entity and rollouts are uneven. But if you are eligible, the question of "do I hire someone for a month or sign something ongoing" arrives with it.
The gateway switch is a fixed-price project with a retainer tail
A payment gateway migration is one of the cleanest fixed-scope jobs on Shopify. You know the start state, you know the end state, and you can write the acceptance test in one line: orders capture, refund, and reconcile correctly in live mode. Scope it, price it, sign it.
What makes it messy in the UAE is not the code. It's the paperwork and the parallel rails. Your trade licence — free zone or mainland — determines what the underwriting team asks for, and a free-zone licence with no local bank relationship can add two weeks that no developer can compress. Then there's everything that has to keep working alongside the new gateway: Tabby and Tamara as separate integrations with their own order flows, cash on delivery with its own reconciliation, and 5% VAT that has to land correctly on the invoice regardless of which of those three the customer picked. We have shipped this and still found a VAT rounding edge case on Tabby instalment orders after go-live. Nobody catches all of it in UAT.
So the honest shape of the contract is: fixed price for the migration, then eight to twelve weeks of light standing capacity to catch what the test plan missed. Not a full retainer. A capped block of hours with an expiry date. If your agency insists the whole thing can be fixed-price with no tail, ask them what happens when a subscription app stops charging renewals on the new gateway. Watch the answer.
When to use a fixed scope Shopify contract
Fixed price works when three things are true. You can describe the finished thing in writing. The thing doesn't depend on traffic data you don't have yet. And you're willing to leave it alone once it's done.
Work that passes that test, in our experience:
- Theme build or migration from WooCommerce, Magento or a custom stack. The migration itself is finite even when the data is ugly.
- Arabic RTL storefront. Mirroring the layout, sorting out the fonts, getting the bilingual product fields structured properly — finite, and worth doing once, properly. The ongoing part is translating new products, which is a content job, not a dev job.
- Shopify Markets for AED and SAR with separate pricing and a Riyadh-specific shipping profile.
- Headless proof-of-concept, custom app for one internal workflow, ERP connector with a defined field map.
- A speed pass with a target number attached. "Get mobile LCP on the top three templates under 2.0s" is a scope. "Make the site faster" is not.
What kills fixed-price contracts is vagueness on your side, not greed on the agency's. If the brief says "improve the PDP", every developer in the region will price the cheapest possible interpretation and you'll spend the project negotiating what "improve" meant.
What a Shopify maintenance retainer actually includes
A retainer buys attention, not deliverables. That distinction is the one most founders miss, and it's why some retainers feel like a waste of money by month four.
A serious maintenance retainer covers: theme and app updates with a staging step, broken-thing triage with a stated response time, checkout and payment monitoring, Core Web Vitals watch, small merchandising and layout changes, and a monthly call where someone tells you what they saw. On a store doing real volume in the GCC it should also cover COD-related order flow issues, courier integration breaks with Aramex or your same-day partner, and the periodic Tabby or Tamara app update that quietly changes the cart drawer.
What it should not include, and what you should push back on if it's listed: new feature development. The moment a retainer has a feature roadmap inside it, the maintenance work gets deprioritised and you find out during Eid that nobody has looked at your error logs since Ramadan started. Keep them separate. Retainer for keeping the lights on and small improvements; separate fixed-price statements of work for anything that needs a design round.
How many hours are in a retainer, and what it costs
Ranges we quote in this market, and what you get for them:
- 10 hours a month. Enough to keep a stable store alive. Updates, monitoring, three or four small changes. Not enough to make anything better.
- 20–30 hours. Maintenance plus one meaningful improvement a month. This is where most D2C brands doing AED 300k–1.5M a month land.
- 60+ hours. A part-time embedded developer. Usually a sign you should be looking at a dedicated developer arrangement instead, which is cheaper per hour than a retainer and better suited to continuous work.
Do the arithmetic before you sign. Take a 10-hour retainer at, say, AED 250 an hour — AED 2,500 a month. Now subtract the standing call at 45 minutes, the monthly report at another 45, and the context-switching that any small-block engagement carries. You're left with roughly 8 hours of actual hands-on-keyboard time. Eight hours is one medium task. If your mental list for the month has four items on it, a 10-hour retainer will disappoint you and you'll blame the agency.
Same sum at 25 hours: AED 6,250 a month, minus the same 1.5 hours of overhead, leaves about 23 hours of build. That's a real month of progress. The overhead is fixed, so the smaller the retainer, the worse the ratio. Which is why we'd rather sell you a well-specced fixed-price project once a quarter than a token monthly fee that buys you nothing.
For reference on the project side, our breakdown of what Shopify work costs holds broadly for GCC engagements too, with the usual adjustment for local onsite time and Arabic content work.
Fixed price vs time and materials on a Shopify project
Time and materials is the third option and almost nobody asks for it, because it sounds like a blank cheque. It isn't, if you cap it.
Use T&M when the work is genuinely discovery-shaped: a performance investigation where you don't yet know whether the problem is the theme, a third-party script or an oversized collection page; an integration with an ERP whose API documentation is a PDF from 2019; a bug that only shows up on COD orders over a certain value. Fixed-pricing that work means the agency prices the worst case and you pay for risk that may not materialise.
The structure that works: T&M with a not-to-exceed ceiling and a checkpoint at 50%. If we hit the checkpoint and still don't know the answer, you get a written option to stop. We've used that on speed audits where the honest finding was "your problem is four apps you're not using" and the fix took two hours instead of the twenty we'd reserved.
The UAE calendar decides more of this than your budget does
Four peaks matter: Ramadan into Eid, White Friday in late November, DSF in the new year, and whatever your category's own spike is. Around each one, the right model flips.
Six to eight weeks out from a peak, you want fixed-price project work finishing. Theme changes, new landing templates, a gateway switch, Markets expansion into SAR. Get it live, get it tested, and then freeze.
Two weeks out and through the peak itself, fixed-price work is the wrong instrument entirely. Nothing you scope in advance will match what actually breaks. What you need is someone on call who can look at a checkout error at 11pm on White Friday, and that is a retainer or a specific peak-support agreement — not a project. We've had a client's third-party gateway start declining a chunk of transactions during a promo weekend because of a velocity rule nobody knew about. That's a 40-minute fix if someone's watching and a lost weekend if nobody is.
The planning consequence: don't schedule a gateway migration for the first week of November. Do it in September, run it through a quiet month, and let the store's normal order volume find the bugs before White Friday does. Same logic for Ramadan — brands in food, gifting and fashion see their behaviour change enough that you want the store stable well before the first week.
Should you hire a Shopify agency monthly at all
Not always, and we'll say so on the call. A store doing under roughly AED 150k a month, on a standard theme, with a stable app stack and no custom code, usually doesn't need standing capacity. You need someone you can call. A block of prepaid hours that roll over for six months serves you better than a monthly fee, and it costs less.
You do need a monthly arrangement when any of these are true: you have custom code in the theme that breaks when Shopify ships changes; you're running two or more storefronts across Dubai and Riyadh; your revenue is concentrated enough in four annual peaks that an hour of downtime is genuinely expensive; or you have no in-house technical person and your marketing manager has started editing Liquid. That last one is more common than anyone admits.
Retainer agreement terms worth arguing about
Most retainer disputes come from four clauses. Sort them before signing, not after.
- Rollover. Unused hours should roll for at least one month, capped at something like 50% of the monthly block. Pure use-it-or-lose-it punishes you for a quiet month you didn't choose.
- Response time versus resolution time. These are different promises. Get both in writing, with a shorter response window for anything that touches checkout. And define what counts as urgent — if checkout is down at 2am during DSF, is that covered, and at what rate?
- The rate card for overflow. When a month runs over, what's the hourly rate, and who approves it? Approval threshold in writing, or you'll get a surprise invoice.
- Notice and handover. 30 days both ways is fair. Handover should specify documented access, repo transfer and a written state-of-the-store note. An agency that won't commit to a clean exit is telling you something.
Also settle IP and code ownership on day one. Any custom app or theme code built for you should be yours. If an agency's standing app remains theirs and you licence it, that's fine — but it needs to be stated, because finding out during a handover is the expensive way.
A practical split for a UAE brand right now
If you're eligible for Shopify Payments and considering the move, the shape we'd recommend: fixed-price the migration, including VAT handling across COD, card and BNPL orders, with a defined test matrix. Add a capped 10-week support block after go-live. Then decide on an ongoing retainer in January, once you've been through White Friday and know what actually needed attention.
Deciding the other way round — signing a twelve-month retainer before you know your own failure modes — is how brands end up paying for capacity they never use.
If you want a second opinion on which of your open items are projects and which are standing work, our free store audit covers it, and the Dubai team page has the specifics on how we work with GCC brands. Bring your list. We'll tell you which half you shouldn't be paying monthly for.