The verdict in three sentences
From 3,000 jobs a day, a no-code back office hits its limits: automation quotas, slowness, double entry and manual invoicing. A custom back office costs 120,000 to 250,000 AED (about 30,000 to 62,000 EUR) in 2026 over 10 to 16 weeks, against about 4,800 AED a month in subscriptions and several days of manual work each week. The right strategy is to rebuild dispatch and invoicing first, then migrate the rest step by step without stopping operations.
Why no-code saturates, and when
No-code tools (online databases, automations, forms) are perfect to validate a model with 5 to 10 riders. At 40 riders and 3,000 jobs a day, each job triggers 5 to 8 automated operations: 15,000 to 24,000 operations daily, far beyond standard plans.
| Indicator (2026 order of magnitude) | Current no-code stack | Custom back office |
|---|---|---|
| Recurring cost | 4,800 AED a month, rising with volume | Hosting 600 to 1,600 AED a month |
| Daily operations supported | 5,000 to 10,000 before overage | 100,000 and more |
| Time to assign a job | 30 to 90 seconds, often manual | Under 2 seconds, automatic |
| B2B client invoicing | Exports and fixes, 2 to 3 days a month | Invoices generated the next day |
| Data entry errors | 1 to 3% of jobs | Under 0.3% |
| Data ownership | With the vendor, limited export | Your own PostgreSQL database |
The most reliable warning sign: when one person spends more than 10 hours a week repairing automations or re-entering data, rebuilding pays off.
Back office priorities and their budget
Rebuilding everything at once is the surest way to fail the migration. Modules rank by impact on margin.
| Module | Priority | Indicative 2026 budget | Timeline |
|---|---|---|---|
| Dispatch and job assignment | 1 | 32,000 to 60,000 AED | 3 to 4 weeks |
| Rider app (proof of delivery, photo, signature) | 1 | 25,000 to 50,000 AED | 3 to 4 weeks |
| B2B client invoicing and rate cards | 2 | 20,000 to 40,000 AED | 2 to 3 weeks |
| Client portal (tracking, bulk orders, API) | 2 | 20,000 to 40,000 AED | 2 to 3 weeks |
| Rider and contractor payouts | 3 | 12,000 to 30,000 AED | 1 to 2 weeks |
| Dashboards and reporting | 3 | 11,000 to 30,000 AED | 1 to 2 weeks |
| Total | 120,000 to 250,000 AED | 10 to 16 weeks |
Migration runs in parallel: the new dispatch first serves one pilot client for two weeks, then 100% of volume. No-code tools are cancelled as you go, which cuts the monthly bill from the third month.
Technical choices that matter to a CTO
Aim for a relational database (PostgreSQL), a documented API from day one for your e-commerce clients, a task queue for notifications, and a rider app as a progressive web app or React Native depending on background geolocation needs. Require the code repository in your name and automated tests on pricing and payout calculations, the two areas where a mistake immediately costs money. In the UAE, also plan for Arabic and English interfaces and cash-on-delivery reconciliation.
To benchmark against a nearshore or offshore team, Kolonell's 11/10/2026 grid puts a complex application between 7,000,000 and 15,000,000 FCFA and above (about 10,670 to 22,870 EUR and above), with Premium maintenance at 120,000 FCFA per month (about 180 EUR).
Mini case study
Need a professional website?
Kolonell builds websites that attract clients, optimized for the Sénégalese market. Free quote in 2 minutes.
Omar, CTO of a B2B last-mile delivery startup in Dubai, manages 40 riders and 3,000 jobs a day. Current costs: 4,800 AED a month of no-code tools, plus one person half-time on invoicing and fixes (about 8,000 AED a month). Total: 12,800 AED a month, or 153,600 AED a year. He invests 190,000 AED in a custom back office, with 1,200 AED a month of hosting and a 3,200 AED monthly maintenance plan. Net monthly savings: 12,800 - 4,400 = 8,400 AED, or 100,800 AED a year, for a payback of about 23 months on savings alone. The biggest gain sits outside the calculation: automatic dispatch that adds 1 to 2 jobs per rider per day.
FAQ
At what volume should we leave no-code?
Usually between 1,500 and 3,000 jobs a day, or as soon as subscriptions exceed 3,200 to 4,000 AED a month. Time spent fixing automations is often a better indicator than volume.
Can we migrate without interrupting deliveries?
Yes, with a module-by-module migration and a pilot client. Plan 2 to 4 weeks of parallel running, which adds about 10% to the budget but avoids any downtime.
Should the rider app be native?
Not necessarily. A progressive web app is enough for proof of delivery and costs 20 to 30% less. Native makes sense for continuous background GPS tracking.
How much does maintaining such a back office cost?
Plan 15 to 20% of the initial budget per year, i.e. 24,000 to 50,000 AED for a 160,000 to 250,000 AED project, plus hosting (600 to 1,600 AED a month).
Do freelance riders create specific constraints?
Yes: the platform must handle contractor status, invoices and transparency on how payouts are calculated. Build these rules into the payout module from the start.
Let's scope your project. Share your current stack, volumes and priorities: we will propose a module-by-module migration plan, priced and scheduled. Detailed quote within 48 h. WhatsApp +221 77 596 93 33.
Mohamed Bah
Fondateur, Kolonell
Passionate about digital and entrepreneurship in Africa, Mohamed has been helping Sénégalese businesses with their digital transformation since 2020. Founder of Kolonell, he believes every SME deserves a professional and accessible online présence.
