Our published band is SAR 5,000 to 15,000 over 10 to 16 weeks. This article separates what the band covers, the seven decisions that push the number higher, and how to compare three proposals on the same lines.
Ask five teams what an ecommerce build costs in Saudi Arabia and you will get five numbers that do not mean the same thing. One sells a configured template, another builds a custom storefront on top of an existing platform, a third includes payment and shipping integrations, and a fourth leaves them outside the scope to discover later. A fair comparison starts with one definition: what is inside the scope, and what will be priced afterwards.
Our published band for this service is SAR 5,000 to 15,000, with a delivery window of 10 to 16 weeks. That figure does not buy every store you can imagine. It buys a complete store for a defined catalogue: a working purchase path, payment and shipping configured, Arabic and English interfaces, analytics, and a handover your team can operate. This article separates what the band covers, what pushes the number above it, and how to read proposals against the same lines.
What the published band covers
A complete store is not a homepage and a product grid. The scope we work to includes clear information architecture, product pages with images, dimensions, materials and return terms, a cart and checkout with few steps, a local payment gateway, shipping rules calculated by city and weight, policy pages, baseline analytics, and real test orders before launch.
It also includes Arabic and English together, not a later translation. That means an appropriate Arabic typeface, correct direction for fields and menus, and Arabic copy reviewed by a person rather than passed through machine translation. It includes handover too: access to the admin, a short training session, and an operating note that explains how to add a product, change a shipping rate, and export a report.
What the band does not cover is equally clear: a multi-vendor marketplace, a native mobile app, deep ERP or PIM integration, a custom loyalty programme, and paid media management. Those are separate projects, and folding them into one proposal is why the numbers you see disagree.
The seven decisions that move the number
Cost follows scope, and scope is set by decisions you can name before the meeting. Each one can be deferred or narrowed, but its effect on the schedule or the launch risk cannot be hidden.
| Decision | What it changes | How to control it |
|---|---|---|
| Catalogue size | Every product family needs fields, images, variants, and content review | Start with the products that carry most revenue; add the rest after launch |
| Selling model | Retail is not B2B with tiered prices or a minimum order | Pick one model for the first release; do not mix both |
| Payment gateways | Each gateway adds authorisation, refunds, reporting, and settlement | Start with one gateway that supports Mada and cards, then add instalments |
| Shipping and delivery | One warehouse differs from several warehouses and city rules | Write the shipping rules on paper before building the interface |
| Integrations | ERP, PIM, or accounting changes the entire data model | Defer non-essential integration; start with a simple daily export |
| Subscriptions | Recurring payment needs lifecycle, failure, and refund handling | Test it as a separate product, not a checkbox on checkout |
| Content and identity | Product photography and Arabic copy take time and an owner | Assign a content owner before kickoff; do not start with placeholders |
Platform first, then the build
The question is not “which platform is best” but “where will the selling model stop fitting the platform”. Local platforms such as Salla and Zid give you a faster launch and ready integration with Mada and e-invoicing, but they impose a data model and limits on customisation. Shopify gives you a wider ecosystem and more themes, but local integration needs extra work. A custom build gives you full freedom and full responsibility.
The platform decision has to be made before interface design, because every platform imposes a way of thinking about variants, categories, and discounts. If your catalogue is small and your path is standard, a local platform is usually cheaper and faster. If you have pricing logic, integrations, or a purchase experience the templates cannot hold, customisation starts to earn its place. Salla and Zid versus custom ecommerce explains the turning point, and Shopify versus a custom build compares the two on cost and control.
Payments and checkout are the riskiest part
From reading to shipping
Want this built, not just explained?
Bring us the problem behind the article. We will tell you what it takes, what it costs, and whether we are the right team.
The biggest reason an order is lost is not the design. It is payment. Baymard's 50-study average puts cart abandonment at 70.19%, and 22% of users have abandoned an order because checkout was too long or complex. The average checkout has 11.3 form fields. Every extra field and every unclear step costs you orders before the payment gateway is ever reached.
In Saudi Arabia add a local layer: Mada, Apple Pay, STC Pay, and cash on delivery or instalments depending on your audience. Do not test these in production. Test a successful card, a declined card, a full refund, a partial refund, an order with no address, and an order above the supported weight. If those cases have not passed, the store has not launched. Saudi checkout conversion fixes gives you the ten issues we find most often.
Measuring the effect is not optional. Deloitte's study for Google found that a 0.1 second improvement in mobile speed was associated with 8.4% higher retail conversions and 9.2% higher average order value. That does not mean every store will gain the same percentage, but it does mean speed is a commercial decision, not a technical improvement to defer.
What a good delivery includes
Before you pay, ask for this list. Every item protects your launch from a problem you will otherwise pay twice to fix.
A separate staging environment from the live store, with test data rather than customer data.
End-to-end payment tests for every method you will support, including failure and refund.
Shipping rules written down before they are built, and matched against what happens when the address changes.
A correct tax invoice in testing, not after the first customer complaint.
Reviewed Arabic copy by a person, not machine translation, with RTL tested on mobile.
Analytics and events for order, payment, and error, with a named owner who reads them.
An operating note that adds a product, changes a shipping rate, and exports a report without a developer.
A first-30-days plan for what is monitored, what is fixed, and who answers when payment breaks.
How to compare three proposals
Compare seven identical lines, not the final number. First: are the platform and the customisation clearly stated? Second: how many products and variants are inside the scope? Third: which payment and shipping gateways are included, and which are outside? Fourth: is invoicing or ERP integration in scope or a separate line? Fifth: are Arabic and English included with human review? Sixth: what acceptance tests will you sign off? Seventh: what happens after launch, and who owns the code and access?
Ask for a sample of a previous store, not a testimonial. Open it on mobile, add a product to the cart, go through to checkout, and look at the Arabic. Ask who will work on your project by name, and whether they stay from kickoff to handover. The team matters more than the template.
Frequently asked questions
Is SAR 5,000 to 15,000 enough for a complete store? It covers one store with a bounded catalogue, one purchase path, and standard integrations. A multi-vendor marketplace or deep ERP integration is outside the band and priced separately.
Does the price include design? It includes the full store interface, not a new brand identity. If you have no identity or product photography, that is a separate stage.
Do you work on Salla, Zid, or Shopify? Yes, and we compare them before building. The platform follows the selling model, not a personal preference.
How long does launch take? Ten to sixteen weeks depending on scope. Content and photography are the first cause of delay, not code.
What if we need changes after launch? We agree a care period and a change budget before kickoff, not after the first outage.
Do you run the ads? No. We build the store and the measurement, and work with your team or agency on campaigns.
Sources
- Baymard: cart abandonment rate — the 50-study 70.19% average.
- Baymard: checkout usability — 11.3 fields on average, and 22% who abandoned because checkout was too long.
- web.dev: Milliseconds make millions — 0.1 seconds and 8.4% higher retail conversions.
- Rasad's published price bands — the figures this article works from.
- Ecommerce development — scope, process, and delivery.
Cover image: Shixart1985, CC BY 2.0, via Wikimedia Commons.

