Getting the licence is easy here. Making the unit economics work is not.
E-commerce is the easiest sector on this list to start and one of the hardest to make profitable. A licence is straightforward, the payment rails are mature, and customers are used to buying online. None of that is your problem. Your problem is that delivery is expensive, returns are high, and two large marketplaces have trained the market to expect next-day arrival at no visible cost.
That is survivable, and plenty of brands here do survive it. But it is survived with margin structure and repeat purchase, not with traffic. Any plan that begins with a customer acquisition number and stops there has skipped the part that decides the outcome.
It is survived with margin structure and repeat purchase, not with traffic.
An e-commerce licence is available on the mainland and in most free zones; for physical goods sold to UAE customers, mainland is often simpler since it avoids the import-or-distribute question. Where founders lose time is the activity list. Write it wider than your first product, since adding a category later is a paid amendment.
Write your licence's activity list wider than your first product. Adding a category later is an amendment with a fee.
The two large marketplaces bring you a buyer who is already searching and take a real commission in return; your own store keeps the margin and the data, and costs you every visitor you have to earn yourself. Treating it as a single choice is the mistake. Most brands that work here run both.
Marketplace, your own store, or both?
Brings a buyer who is already searching. Takes a real commission and owns the customer relationship in return.
Keeps the margin and the data, and costs you every visitor you have to earn yourself.
Marketplace for reach and price-competitive products, owned channel for the relationships worth building. Decide per product line, not for the whole business.
Import duty is generally 5% of landed value, with exemptions and higher rates on specific categories, but that is not what hurts. What hurts is clearance delay on a first shipment when the paperwork is not exactly right, because a container sitting in port is capital you cannot use. Get a customs code early and use a freight forwarder for the first two shipments.
The number that hurts is not the 5% duty, it is clearance delay on a first shipment. Use a freight forwarder for the first two and treat the fee as tuition.
Last-mile delivery costs real money per order and customers expect it free; returns on apparel and sized goods run high enough that a brand can be profitable on gross margin and still lose money per order. This is why average order value matters more here than conversion rate. Bundles and subscription formats exist because they make the delivery cost survivable.
A brand can be profitable on gross margin and still lose money per order once free delivery and high returns are counted properly.
A first sale typically costs more to buy than founders expect: marketplaces train shoppers to compare price across five competitors on one screen, and paid social is not cheap relative to average order values. Treat the first sale as buying a customer, not a profit event. Measure success on whether they buy again inside the margin left over.
Treat the first sale as buying a customer, not a profit event. Success is whether that customer buys again inside the margin left after the first.
Card payment is well served and settlement is predictable; the complication is cash on delivery, still a real share of orders, which carries a handling fee, a delay before you see the money, and a door refusal rate that lands the full delivery cost on you with no sale attached. Decide deliberately whether to offer it. Either choice is defensible, drifting into it is not.
Cash on delivery carries a fee, a delay, and a door refusal that lands the full delivery cost on you with no sale attached. Decide it deliberately.
VAT registration is mandatory once taxable supplies and imports pass AED 375,000 in a rolling 12 months.
The same number as the corporate tax band, measured on revenue rather than profit. Watch it monthly, not annually.
Read it on Federal Tax AuthoritySmall Business Relief treats a business with revenue up to AED 3,000,000 as having no taxable income, by election.
It is an election you make per tax period, not something applied for you, and it is revenue not profit that is tested.
Available only for tax periods ending on or before 31 December 2026.
Read it on Federal Tax AuthorityE-invoicing becomes mandatory in phases. A business under AED 50,000,000 of revenue appoints an accredited service provider by 31 March 2027 and issues electronic invoices from 1 July 2027.
It changes how you issue every invoice, so the decision belongs in the accounting stack you pick now rather than a migration later.
Appoint a provider by 31 March 2027; issue electronically from 1 July 2027.
Read it on Ministry of FinanceCorporate tax is 9% on taxable profit above AED 375,000, and 0% up to it.
A company earning under AED 375,000 of profit pays no corporate tax, but still has to register and file.
Read it on Federal Tax AuthorityChecked against the official sources on 26 September 2026. Rules change and this is general information, not tax or legal advice. We do not file anything for anybody. Every claim above links to the authority that publishes it, so you can read the rule yourself before acting on it.
E-invoicing arrives in two phases
The large-business phase starts first and is closer than it looks in a list of dates. Both rows read left to right: appoint a provider, then go live.
AED 50,000,000 and above
30 October 2026
Appoint a service provider
1 January 2027
Live
Under AED 50,000,000
31 March 2027
Appoint a service provider
1 July 2027
Issue electronically
If you are selling physical goods to customers in the UAE, mainland is usually simpler. Free zone works well when you are selling services, digital goods, or exporting.
Not until your taxable supplies pass the threshold in a rolling twelve months. Track it monthly, because the test is not your financial year.
It depends on your margin per order. It costs a handling fee, delays your cash, and carries a refusal rate that leaves you paying delivery on a sale that never happened. Model it before you decide.
Fintech and payments
The licence decides your product. Pick it before you build anything.
SaaS and developer tools
The lightest sector to set up here, and the one with the fewest local ceilings.
Logistics and trade
The sector where the country itself is the product advantage.
Property and construction
A large, digitising market with buyers who move slowly and pay properly.
Food, health and consumer
Approvals are the slow part. Distribution is the part that decides it.
Tell me what you are deciding. I will come back with which licence applies to you, what it actually costs in the first year, and the thing founders in your position most often get wrong. Thirty minutes, free, and no pitch.
If the licence is done and growth is the problem, the membership reads your numbers each month, gives you three Sprints a week and the courses to act on them, one step at a time.
See the membershipUpgradIQ, Inc. operates in the UAE as UpgradIQ FZC LLC. Back to the UAE guide
We use cookies
Some are needed to run the site and keep you signed in. Others help us understand what’s useful through analytics. We don’t run analytics until you agree. Cookie details
