Choosing a dispatch system for your e-commerce delivery operation

How to know you need a system at all

Most delivery operations start with simple tools that genuinely suffice at first: a WhatsApp group for couriers, an Excel file for orders, and one energetic supervisor's memory. Then the symptoms creep in: an order lost between two messages, a customer calling three times about one shipment, COD amounts nobody can say precisely who is holding, and entire evenings spent reconciling numbers that refuse to reconcile.

These are not individual mistakes — they are structural signals that the operation outgrew its tools. The smart response is not hiring another supervisor, who would simply inherit the same chaos, but moving the whole cycle into a system designed for it. The next question is: which system? This is where many buyers go wrong by comparing feature lists. The better method is comparing systems against your operating day: does this system cover your day end to end, or does it cover fragments and leave you to patch the rest?

There is also a silent cost to postponing the decision itself: every extra month on scattered tools entrenches operating habits that get harder to change later, and piles up fragmented data you will not be able to analyze the day you need a decision — which zones are most profitable? Which couriers are most efficient? Where do delays cluster, and why? No WhatsApp archive answers those questions, however long it grows — and neither does an Excel file rebuilt every evening.

Axes one and two: a complete order lifecycle and COD control

Start the evaluation with lifecycle completeness. A serious system covers, without a break: order intake, courier assignment, coverage zones with pricing, multi-stop route optimization, shipping labels, proof of delivery at the door, and an automatic delivery monitor that catches stalled orders before the customer complains. And in Saudi Arabia specifically, ask how it handles prayer times: Pixa's dispatch module, for instance, schedules stops around closure windows using a built-in astronomical calculation that derives the times from location and date — because a plan that ignores 5 daily closure windows is not a plan.

Then scrutinize COD, because that is where a gap becomes a direct cash loss: is every amount tied to an order and a courier at the moment of collection? Does each courier carry a running balance the supervisor can see? Are settlements dated, with differences attributed to specific orders? A system without a closed COD cycle will send you straight back to the evening Excel file, however polished its other screens look.

And do not forget the person who will use the system more than anyone: the courier. With Pixa, the driver app runs as a mobile web app with OTP login — the day's tasks, proof of delivery, COD, and the daily inspection — with nothing to install from app stores, which makes onboarding a new courier a matter of minutes (native iOS/Android store apps are on the roadmap — coming soon).

Test geographic flexibility as well: can coverage zones be drawn on the map with independent boundaries and pricing per zone? Can the tariff differ between one district and another, one city and another, without a manual decision on every order? Disciplined geographic pricing is what turns expansion into a new city into a costed operational decision rather than a gamble whose results surface too late.

Axis three: your end customer's experience

Your store's customer never sees the operations screens. They see exactly three things: when the order arrived, how they knew it was on the way, and how smooth the handover felt. So evaluate the system explicitly from the customer's side:

• A public tracking link that opens with no login and shows order status as it changes — its direct effect is fewer where-is-my-order calls.
• Proactive stage notifications: in Pixa, across 5 channels — WhatsApp, SMS, email, app push, and browser push — with 3 interchangeable WhatsApp Business providers.
• Documented handover: proof of delivery that settles any later dispute before it turns into an angry call and word against word.

And since delivery is part of a wider operation, ask about the fleet dimension itself: live vehicle tracking with real-time streaming gives your dispatcher a single picture of couriers and orders together — the foundation of any quick intervention when something breaks mid-day.

Add the measurement question to this axis: does the system give you a custom report builder that can be scheduled to arrive by email or WhatsApp on a regular rhythm? What is not measured is not managed, and customer experience in particular is measured through simple indicators — on-window delivery, failed attempts, status-enquiry volume — that need to sit in front of you on a fixed cadence, not be extracted manually only when something already went wrong.

Axis four: integrations and local compliance

A system that does not integrate becomes an island fed by manual data entry. Check four doors:

• Connecting your store: a Partner API and Webhooks, available today, let orders flow automatically from your platform while statuses and tracking links flow back — more on the integrations page.
• WASL: the Transport General Authority's WASL mandate has made tracking an operating condition, not a luxury. Pixa ships built-in integration readiness — company, vehicle, and driver registration, location pushes through a queue that loses no message during outages, and a compliance report showing acceptance rates — see the compliance page.
• Billing: invoicing through ERPNext with ZATCA-compliant invoices, plus a prepaid wallet and a tiered tariff engine.
• Devices: if you already run a fleet with tracking hardware, anchor your decision with a compatibility question: Pixa supports 18 tested protocol families today (Teltonika, Queclink, GT06, Meitrack, and others), with an expansion methodology aimed at hundreds of protocols.

Add security and hosting to the list — in B2B contracts these are purchase conditions, not technical footnotes: hosting inside Saudi Arabia, server-enforced MFA, strict tenant isolation with a database-level guard, a tamper-evident hash-chained audit log, and recurring Snyk scans.

Behind all these doors stands the architecture question — and a platform built to scale tells you its numbers, not its slogans. Pixa, for instance, runs on more than 15 microservices, offers 7 specialised portals for every party in the ecosystem — platform, dealer, customer, driver, and a public portal among them — and ships more than 90 working screens in light and dark themes, built Arabic-first rather than localised later. These details are not technical vanity; they are the difference between a system that accompanies your growth and one whose ceiling you hit within a year.

An evaluation table you can take into any demo

Bring this table to every vendor presentation, and insist on a working answer rather than a marketing slide:

Notice that the "strong answer" column asks for observations, not promises: every row can be demonstrated live in front of you in a single session. A vendor who answers with an immediate working demo on a real screen tells you more about product maturity than any specification document, however elegant.

Pilot small, then decide on numbers

However good the demo, do not sign a long commitment before a pilot on your real orders: one zone or one team for a few weeks. Before starting, record your baseline — orders per courier per day, on-window delivery rate, where-is-my-order enquiry volume, and the age of unsettled COD balances — then compare the same numbers at the end. And when you look at the pricing model, look for clarity and the ability to scale with your growth, not just the entry number.

During the pilot, watch what demos rarely show: how quickly your team learns the screens without lengthy training, how often you needed vendor support and how the response felt, and how the system behaves on a peak day rather than a quiet one. In the end you are not buying a feature list — you are buying an operational relationship you will live with for years, and its quality is tested on the hard days, not the easy ones.

The right decision, in the end, is not picking "the most powerful system on the market". It is picking the system that matches your operating cycle, respects the specifics of your market, and proves its effect in numbers on your own orders. Everything else is slide headlines.

Want to run that pilot on your actual operation? Talk to the Pixa team — we will start from your operating day, not from a canned demo.