Test asset-heavy fixed costs with GoLemon's 28-month record
GoLemon built its own grocery fulfilment, grew customers, and still could not fund the fixed costs. That record gives founders a clear kill criterion.
For founders and investors weighing asset-heavy startups, GoLemon's shutdown is a fixed-cost warning: it ended operations after 28 months, having registered 40,000 customers and moved more than ₦2 billion in groceries. GoLemon said it was ending operations because it could not secure the additional funding needed to keep the business running. The startup began in March 2024 with a founding team of four former senior Paystack executives.
Its founding team brought payments experience, but the model still had to solve a harder problem: owned grocery fulfilment. GoLemon built its own fulfilment infrastructure and offered an inventory of more than 7,000 items. That choice moved the cost base from variable marketplace costs into fixed infrastructure costs, so the business needed enough order flow to pay for the assets even when demand slowed.
Payments experience did not remove the fixed bill
Checkout expertise can make an order easier to complete, but it does not shrink the warehouse, the stock, or the delivery network. Grocery delivery is unforgiving when the assets are owned: the costs continue whether baskets arrive in clusters or trickle in one at a time. GoLemon's peak headcount was 37 employees. Every role added to that team increased the monthly cost that had to be covered before an extra basket could contribute to the company's survival.
GoLemon's chief executive said the typical customer basket was ₦50,000 and delivery fees could be as low as ₦300. That pricing made the service feel generous, but it also left less revenue per order to absorb the cost of moving goods through owned infrastructure. The real question was whether the combined contribution from all baskets could carry the fixed costs of the system that delivered them.
Order density decides whether owned fulfilment pays for itself
When a company owns fulfilment, the warehouse does not pause on a slow week. Fixed costs keep running even when orders are thin. For an order-density business, the key question is whether enough baskets pass through the same routes, shelves, and staff to spread those costs. A startup can have many registered customers and still fail if orders are too sparse across neighbourhoods, hours, and delivery paths.
The same infrastructure can be efficient when baskets cluster and expensive when they do not. The test is whether the route, the shelf, and the staff are paid for by the orders that actually arrive. GoLemon had customers, movement, and a team, but its record does not show that total contribution margin from all orders covered the fixed costs of the infrastructure it had built.
The kill test is contribution margin against infrastructure
Founders and investors should ask whether the sum of all baskets covers the fixed costs of the assets that make the baskets possible, not whether a single basket is profitable. Contribution margin is the amount left after the direct costs of an order, and it must be summed across realistic total order volume, not best-case volume.
If the total contribution margin is smaller than the fixed bill, the model is not covered, even if each order looks healthy. A large customer count can hide a thin contribution margin. A high average basket can hide a delivery network that is too expensive to keep running. The fixed bill is the number that decides whether the next round is useful or wasteful.
- Estimate the contribution margin per order after product cost, delivery cost, and payment cost.
- Multiply that margin by realistic total order volume, not best-case volume.
- Compare the result to the fixed cost of owned fulfilment, inventory, and core staff.
- If the total does not cover the fixed cost, treat that as the kill criterion.
GoLemon's wind-down shows why the test matters. When the company described its wind-down, about 20% of staff had already secured new roles. That is the moment the fixed-cost coverage test should have become the kill criterion: if total contribution margin from all orders cannot cover fixed infrastructure costs, the asset-heavy model is not viable, no matter how many customers it has registered.