He Was Delivering at a Loss and Called It Growth
Delivery is a real advantage over supermarkets and a real cost. Here is how to run it so the orders you gain are worth having.
Delivery orders at Kumar's Chennai shop grew from a handful a week to about twenty-five a day over eighteen months. He treated the growth as an unambiguous success until he sat down and costed it.
Roughly a third of those orders lost money once his time, his helper's wages and fuel were counted. The other two-thirds were excellent business. He had been treating both the same way.
Delivery Is Two Different Businesses
A ₹900 weekly grocery order delivered on a planned run is highly profitable. A ₹70 order for one item delivered immediately, two streets away, is not — and the second kind multiplies faster, because it is easier to place.
- Large scheduled orders: strong margin contribution, predictable, batchable
- Small urgent orders: usually loss-making once time is counted
- Repeat weekly orders from regulars: the best business you have
- One-off distant orders: rarely worth the trip unless batched with others
Twenty-five deliveries a day sounded like a business. Eight of them were a hobby I was funding.
The Costs That Do Not Appear on the Bill
- Picking and packing time, usually during trading hours
- Travel time, which is the largest cost and the least visible
- Fuel and vehicle wear
- The cost of someone not being at the counter while the queue builds
- Returns and refusals, where the customer is not home
Travel time is the one shops consistently underestimate. A fifteen-minute round trip on a ₹70 order at a 12% margin loses money before fuel is counted.
Three Rules That Fix Most of It
A minimum order value
The single most effective change. Set it where the margin on a typical basket covers a delivery trip. Most customers accept it immediately, and the orders you lose are precisely the ones costing you money.
Fixed delivery windows
Two or three runs a day rather than on demand. This lets you batch several orders into one trip, which transforms the economics — the same journey now serves four customers instead of one.
A delivery charge below a threshold
Free above a value, a small charge below it. This is familiar to customers from every delivery app they use, so it needs no explanation and generates almost no resistance.
Group the run geographically rather than first-come-first-served. Four orders in one lane is one trip; four orders in four directions is four trips at the same order value. Simply sorting the list by area before setting out typically cuts a delivery round substantially.
Getting the Order Right Before It Leaves
A wrong delivery costs the goods, the return trip and a portion of the relationship. Most errors originate before packing.
- Record the order as an itemised list rather than working from a chat thread
- Confirm it back to the customer with prices before packing
- Check the packed bag against the list, not against memory
- Note substitutions explicitly and confirm them before delivering
- Mark payment status on the order, so nothing is delivered with an unclear balance
The priced confirmation removes almost all disputes at the door, because the amount was agreed before anyone travelled.
Why It Is Still Worth Doing
Delivery is one of the few areas where a neighbourhood shop competes directly with a delivery app and can win, because you already have the stock, the relationship and the proximity. What you do not have is their scale, which is exactly why the economics have to be deliberate rather than accidental.
Kumar introduced a minimum order, two fixed runs and a small charge below the threshold. His delivery count fell by about a fifth. His delivery profit roughly doubled.
Arali records orders against customers with items, prices and payment status, so what left the shop and what was owed are never in question. The routing and the thresholds are yours to set — but they are worth setting before the volume arrives, not after.
Frequently asked questions
Should a small shop charge for home delivery?
Yes, below a threshold. Free above a set order value and a small charge beneath it is a familiar structure customers recognise from delivery apps, so it generates almost no resistance. Combined with a minimum order value, it removes the small urgent orders that lose money once travel time is counted.
Why do small delivery orders lose money?
Travel time, which shops consistently underestimate. A fifteen-minute round trip on a ₹70 order at a 12% margin loses money before fuel is counted, and that is before picking time and the cost of someone being away from the counter. Small urgent orders also multiply faster than large ones, because they are easier to place.
How do I make home delivery more efficient?
Run two or three fixed windows a day rather than delivering on demand, and batch each run geographically rather than by order time. Four orders in one lane is one trip; four orders in four directions is four trips at the same value. Sorting by area before setting out typically cuts a round substantially.
How do I avoid mistakes and disputes on delivered orders?
Record the order as an itemised list rather than working from a chat thread, confirm it back with prices before packing, check the packed bag against the list rather than memory, flag substitutions explicitly, and mark payment status on the order. A priced confirmation before anyone travels removes almost all disputes at the door.