Running a Delivery Fleet: Drivers, Routes and the COD Float

Once you have riders on the road, a delivery operation becomes three problems that recur every single day: who is driving, where they are going, and where the money is. The first two get most of the attention. The third — the cash riders collect on cash-on-delivery parcels, the "float" — is the one that quietly grows with the business and is by far the most likely to produce a loss you cannot explain. This guide takes them in the order they will hurt you.

1. Engaging drivers: get the classification right first

Before rates, uniforms or apps, settle what the relationship legally is. Malaysian operators typically run one of three models:

Classification is not a preference. If a rider works your hours, on your routes, under your instruction, using your equipment, calling them a contractor does not make them one — and a misclassification carries real statutory exposure for unpaid contributions. Where the arrangement is genuinely employment, budget for it properly: employer EPF, SOCSO and EIS are a material share of the cost of a rider-day, and the exact rates come off official schedules rather than round percentages. Our EPF, SOCSO and EIS guide has the tables, and the payroll calculator computes a real rider cost from a real wage.

The employment-law framework — hours, rest days, overtime, termination — is in our Employment Act guide, and it applies to riders like anyone else.

Retention is cheaper than recruitment

Rider churn is the hidden cost line in every fleet. A rider who leaves takes route knowledge, customer familiarity and a month of training with them. The levers that actually work are unglamorous: pay on time every time, make the per-drop rate predictable, do not blame riders for address failures that came from the merchant, and give the good ones the dense routes.

2. Routing: you are optimising density, not distance

The instinct is to minimise kilometres. The right objective is to maximise drops per hour, and those are not the same thing — a route with fewer kilometres but three high-rise buildings requiring guardhouse clearance is slower than a longer route through landed housing.

Practical rules that survive contact with a real day:

• Cluster by building and street, not by postcode. A postcode can span areas a rider cannot practically cover in one sweep.
• Sequence by access constraints. Offices before 5pm, condominiums when residents are home, businesses outside their lunch closure.
• Cap the route, then measure. Give a rider a fixed number of drops and track completion. A route that consistently finishes early is under-loaded; one that consistently fails its last five is over-loaded and generating return legs.
• Keep riders on the same area. Route knowledge — which guardhouse is difficult, which shop closes early — is worth more than any routing algorithm and it only accumulates if the rider stays put.
• Plan for re-attempts. They are not exceptions; they are a predictable percentage of every day, and a route with no slack turns one failure into three.

Measure two things and nothing else at first: drops per rider-hour and first-attempt success rate. Everything you might improve shows up in one of them.

3. COD float: the exposure that grows with you

Every rider carrying cash-on-delivery parcels is holding your money in their pocket, all day, in a system you can only reconcile after the fact. At ten parcels a day this is a nuisance. At five hundred it is the largest uncontrolled risk in the business.

Six controls, all cheap:

1. Cap the float per rider per day. A ceiling on the total COD value one rider can carry. Above it, split the parcels across riders or across days.
2. Reconcile daily, not weekly. Cash in, against consignment notes marked delivered, same evening. A one-day gap is findable; a one-week gap is an investigation.
3. Match on consignment note, never on amount. Two RM80 collections will reconcile perfectly against each other and be wrong.
4. Make handover a two-party event — counted and signed by two people, or banked directly by the rider with a slip photographed and sent.
5. Never let a shortfall roll over. An unexplained gap on Tuesday that is still open on Friday will be there next month too. Escalate on day one, every time.
6. Separate merchant remittance from rider collection in your books. You are holding merchant money in trust. Spending float on operating costs is a cash-flow disaster that looks like profit — see understanding cash flow.

The seller's-side view of why COD is expensive is in the real economics of COD; as the operator, you are the one holding the risk in between.

4. Proof of delivery, or you will lose every dispute

Timestamped, geotagged photo proof at the point of handover is not a premium feature — it is how disputes end. Without it, "the parcel never arrived" is unanswerable and you pay. With it, most disputes close in one message. The same evidence also tells you which riders' failures are genuine access problems and which are not, which is otherwise unknowable.

5. Vehicles and safety

• Confirm commercial-use cover. A personal motor policy may exclude delivery work; discovering that after an accident is the end of a small operator.
• Schedule maintenance rather than reacting. A bike off the road at 9am costs a full rider-day and the drops that day would have carried.
• Fuel and maintenance per drop is a real cost line. Track it — it is what turns an apparently profitable per-drop rate into a loss.
• Enforce basic safety. Helmets, no overloading, no unrealistic delivery windows that push riders to take risks. A rider injured on an impossible schedule is both a human cost and a liability.

Frequently asked questions

How many drops per day should a rider do?
There is no universal figure — it is entirely determined by route density and access type. Rather than chase a benchmark, calculate the break-even for your own cost base with the break-even calculator, then measure what your riders actually achieve on your actual routes. The gap between those two numbers is your business problem, stated precisely.

Should I buy bikes or have riders use their own?
Riders using their own vehicles reduces your capital outlay and shifts maintenance to them, and it makes commercial insurance cover the rider's responsibility — which you must verify rather than assume. Owning the fleet costs capital but gives you control over condition, branding and availability. Most small operators start with rider-owned and move to owned as volume justifies it.

How do I stop COD shortfalls without treating riders as suspects?
By making the controls systematic rather than personal. A float cap, a daily reconciliation and a two-party handover apply to everyone equally and are visibly about process, not trust. Shortfalls then surface as facts rather than accusations — which is better for the honest riders too, because it clears them.

What is the right per-drop rate to pay?
Work backwards from your revenue per drop and your target margin, then check the number against what riders can earn elsewhere in that area. A rate that is theoretically viable but below the local market simply produces churn, and churn is more expensive than the difference.

Do I need software from day one?
No — and starting with software usually means buying something before you know what you need. Run manually until you can state your drop density, failure rate and reconciliation pain precisely. Then choose, using build vs white-label for courier software as the framework.

Sources: employer statutory contribution obligations (EPF, SOCSO, EIS) as covered in our EPF/SOCSO/EIS guide, pinned against the EPF Act 1991 Third Schedule and the PERKESO contribution schedules; employment-law framework per the Employment Act 1955 as covered in our HR guide. Insurance scope for commercial delivery use varies by policy and insurer — confirm with your own insurer rather than assuming a personal policy extends to delivery work. Per-drop rates, rider costs and drop-density figures are operational variables, not published benchmarks; every number in this guide is illustrative of a method.