Payment gateways in Egypt: Kashier vs Paymob vs Fawry vs Fawaterak vs Geidea

Every founder launching e-commerce in Egypt hits the same wall somewhere around week three: the product works, the store is built, and now money has to move. Someone asks which payment gateway to use. They get five different answers from five people who each integrated exactly one.

I have integrated most of these in production, and the honest summary is that the fee is rarely the deciding factor. The published rates cluster within a few tenths of a percent of each other. What actually differs, and what actually hurts later, is settlement timing, what happens when a payment fails, whether the gateway supports your business model at all, and how much of your own engineering time the integration eats.

The quick answer

If you want the decision without the reasoning, here it is, and the rest of the article is why:

• Standard online store, cards and wallets, want to be live this month: Kashier or Paymob. Both publish 2.75% + 3 EGP, both have real CMS plugins, both onboard quickly.
• Your customers do not all have cards: Fawry, or Fawry alongside a card gateway. Reference-number payment at a corner shop is not a legacy feature in Egypt; for a large slice of the market it is the payment method.
• Marketplace, split payments, sub-merchants: Paymob. It is the one with an explicit marketplace product rather than a suggestion that you build it yourself.
• You also have physical stores or need Apple Pay first-class: Geidea, which comes at online payments from the acquiring and POS side.
• Invoicing, small business, minimal engineering: Fawaterak. It is built around getting a link in front of a customer, not around a custom checkout.

Side-by-side comparison

Fees: what the numbers actually mean

Three of the five converge on 2.75%. That is not a coincidence: it is roughly what the local card economics support once interchange and scheme fees are paid. So when someone quotes you 2.75% + 3 EGP, they are quoting you the market, not gouging you.

The fixed component is where small-basket businesses get quietly hurt. Three pounds on a 1,500 EGP order is noise. Three pounds on a 60 EGP order is 5% on top of the 2.75%, so you are paying nearly 8% for the privilege of selling something cheap. If your average order value is low, the fixed fee is the line item to negotiate, not the percentage.

Do this calculation before you sign anything:

effective_rate = (aov * pct_fee + fixed_fee) / aov

AOV   350 EGP @ 2.75% + 3  →  3.61%
AOV   120 EGP @ 2.75% + 3  →  5.25%
AOV 2,000 EGP @ 2.75% + 3  →  2.90%

The other cost nobody quotes you is failure. A gateway with a weaker 3-D Secure flow or a flakier issuer connection loses you real orders, and a 3% drop in authorisation rate costs more than the entire fee difference between any two providers on this list. Ask for authorisation rates on your card mix. If a provider will not discuss them, that is information too.

Also confirm, in writing: refund fees, chargeback fees, whether the percentage is refunded when you refund an order (usually not), and the cross-border rate, since international cards almost always price higher than local ones, and if you sell to Gulf customers that becomes your real blended rate.

Settlement: the number that matters more

This is the section I wish someone had written for me the first time.

Kashier documents settlement three working days after the transaction. Paymob's standard plan settles weekly. Fawry and Geidea set terms in the merchant agreement. On paper that is a difference of a few days. In practice it decides whether your business works.

If you buy inventory, pay suppliers, or pay drivers out of the same revenue you are collecting, settlement lag is your working capital requirement. Weekly settlement on 500,000 EGP of monthly revenue means roughly 115,000 EGP of your cash is permanently sitting at the gateway. That money is not lost, but you have to finance it, and financing it costs more than 0.25% of a fee difference.

Rule of thumb: cash trapped in settlement = (monthly revenue ÷ 30) × settlement days. Compute it for both options before you pick the cheaper rate.

Also ask what triggers a hold. Every provider reserves the right to hold funds when they see risk: a spike in volume, a rise in chargebacks, a business category they get nervous about. Find out the specific triggers and the release process before you are on the phone about it during a launch week.

Payment methods and Apple Pay

Card acceptance is table stakes everywhere on this list. The differences are at the edges, and in Egypt the edges are large.

Mobile wallets

Vodafone Cash, Orange Money, Etisalat and the bank wallets are a serious share of online payments and skew heavily toward customers who do not use cards online. Paymob has the broadest wallet story; Kashier and Fawaterak support wallets too. If your target customer is outside the card-carrying urban segment, wallet coverage matters more than checkout aesthetics.

Cash and reference numbers

This is Fawry's structural advantage and it is worth being blunt about it: a customer gets a reference number, walks to a pharmacy or a kiosk, hands over cash, and your order is paid. No card, no bank account, no trust in online payment required. Reference-number payments also tend to carry lower processing fees than online card payments, because there is no card network in the path.

The cost is latency and abandonment. The order sits unpaid until the customer physically goes somewhere. You need expiry logic, inventory reservation policy, and a reconciliation flow. Plan for a meaningful share of reference numbers never being paid.

Apple Pay and Google Pay

Geidea documents Apple Pay support in Egypt including a Direct API integration, alongside Google Pay and Meeza. Paymob supports Apple Pay as part of its wallet acceptance. For the others, treat Apple Pay as something to confirm for your specific merchant account rather than assume, since availability depends on the acquirer and on how your account is configured, and it changes faster than anybody's documentation.

Apple Pay is not a vanity feature. On iOS traffic it measurably lifts conversion, because it removes card entry and 3-D Secure friction in one step. If a large share of your traffic is iPhone, and for premium Egyptian e-commerce it usually is, this can matter more than your fee.

Installments and BNPL

Installment plans through banks, and BNPL providers, materially raise average order value on higher-ticket goods. Paymob has the widest coverage here. Note that installment transactions usually price differently from standard card transactions: sometimes paid by you, sometimes by the customer. Confirm which.

Integration effort, gateway by gateway

Rough engineering effort for a competent backend developer, assuming you already have an application to plug into and you are doing it properly (server-side verification, webhook handling, idempotency, reconciliation) rather than dropping in a plugin and hoping.

Add to all of these: onboarding and KYC, which is usually the real critical path. Commercial registration, tax card, bank account details, and a compliance review. That is days to weeks of calendar time regardless of which provider you choose. Start it before you write the integration, not after.

Whatever you pick, build these four things or you will regret it:

1. Server-side verification. Never mark an order paid because the browser said so. Verify against the gateway or a signed webhook, always.
2. Idempotent webhook handling. Webhooks arrive twice. Key on the gateway's transaction reference and make double-delivery a no-op.
3. Your own payment ledger. One row per payment attempt with the gateway reference, status transitions, and timestamps. When you dispute a settlement six months from now, this is the only thing that saves you.
4. Daily reconciliation. Compare what your database thinks was paid against the gateway's settlement report. The gap is always non-zero and you want to find it in a job, not from an angry customer.

Where each one is the wrong choice

The section vendor comparison pages skip. This is where the value is.

Kashier is the wrong choice when…

You are building a marketplace that needs sub-merchant onboarding and split payouts, or you need the widest possible set of alternative payment rails. Kashier is clean and fast at the standard e-commerce job; it is not where you go for complex money movement.

Paymob is the wrong choice when…

You are small, in a hurry, and weekly settlement would strangle your cash flow, or you need a minimal integration and do not want to reason about a product surface built for far larger merchants. Breadth has a cost, and on the standard plan part of that cost is waiting a week for your money.

Fawry is the wrong choice when…

Your customers are card-native and expect an instant, modern checkout, and you would be taking on the reference-number reconciliation burden for a payment method almost nobody in your audience uses. Fawry's strength is reach into cash. If your product is a B2B SaaS subscription sold to funded startups, that strength is irrelevant to you.

Fawaterak is the wrong choice when…

You need a deeply customised checkout, complex payment flows, or marketplace mechanics. Its simplicity is the product. Outgrow the simplicity and you outgrow the tool.

Geidea is the wrong choice when…

You are a small online-only merchant with no POS story and no leverage to negotiate. Unpublished pricing generally means pricing that rewards volume, and the enterprise-shaped onboarding is overhead you do not need if all you want is to sell 200 orders a month online.

Pick by business model

Questions to ask before you sign

Print this. Ask every provider the same list, in writing.

1. What is my rate on local cards, international cards, wallets, and installments, each stated separately?
2. What is the fixed per-transaction fee, and does it apply to refunds?
3. Exactly how many days from transaction to money in my bank account?
4. What triggers a settlement hold, and what is the release process?
5. What is the refund fee and the chargeback fee?
6. Is Apple Pay available on my account, and what does enabling it require?
7. What authorisation rate do merchants like me see?
8. Is there a sandbox with test cards, and can I have access before signing?
9. What does the settlement report contain, and can I pull it via API?
10. Who do I call at 2am on launch night, and what is the response commitment?

The answers to questions 3, 4, and 10 will tell you more about how the next two years go than the answer to question 1.