The first weekend is an incomplete period
A successful promotion has an immediate atmosphere. The order count rises, colleagues share the result and the next campaign begins to feel like an obvious decision. Finance can appear unhelpful when it asks what remains unresolved. Yet the opening weekend is often a poor boundary for understanding the transaction. Some orders will be cancelled, some goods returned and some service work performed later. Those possibilities do not make the early demand unreal. They make it incomplete as an account of the economic result.
At Seravel Home, one defined campaign cohort began with SGD8.4 million of orders. That number was useful for understanding the response and the operating load ahead. It was not equivalent to revenue the company would ultimately retain. Cancellations and refunds later totalled SGD1.2 million, leaving SGD7.2 million of net sales. The difference mattered before we decided whether to repeat the same offer. A campaign can be commercially worthwhile at the lower figure, but management should know which figure its recommendation depends on and when the uncertainty will become clearer.
The initial review brought an order total together with costs collected on a different timetable. That can create a particularly persuasive impression because the arithmetic appears straightforward. The problem is the relationship between the numbers. A product cost based on retained sales cannot simply be set against gross orders without addressing the returns between them. Equally, charging the original cost of every returned item while also recognising the value recovered can distort the account in the other direction. The management view needs a consistent boundary before its apparent precision is useful.
I asked the team to follow the cohort through the relevant return window. We continued ordinary trading while postponing the proposed repeat promotion, so the request was not an instruction to stop the business until perfect knowledge arrived. Digital colleagues reasonably worried about losing audience momentum. Service and fulfilment teams knew that the customer response was still developing. Bringing those perspectives together helped us decide which uncertainty was material to the next commitment and which work could proceed while we examined it.
Follow the cohort without losing the customer
The completed management account showed SGD4.1 million product cost for retained sales, after the recovery treatment for returned goods. Fulfilment, payment and return-handling costs were SGD1 million, and campaign media expenditure was SGD1.3 million. Against SGD7.2 million net sales, that left SGD800,000 of contribution before central expenses. The amount did not represent statutory company profit or an independent valuation of the brand benefit. It answered a narrower question about the defined campaign and the costs included in its account.
That boundary is important because a retailer can misuse a contribution figure in more than one way. It may celebrate a positive amount without asking whether scarce capacity could have earned more elsewhere. It may reject a campaign after adding broad central allocations that would remain regardless. It may compare two periods with different return maturity or product mix and call the difference improvement. I want colleagues to know what the measure is designed to support before it becomes a verdict on the commercial team or the entire channel.
The returns themselves deserved attention beyond their financial effect. A small group of bulky items created disproportionate handling work. Product and service colleagues examined whether presentation, expectations or packaging contributed to the pattern. Finance could identify where the costs were concentrated, but it could not infer the customer's experience from a ledger line. A return may be legitimate even when it is expensive, and the response should not be to make the published promise difficult to use. The commercial offer and the ability to fulfil it belong in the same discussion.
Nor should an improvement be claimed before the evidence exists. A revised product description may sound sensible without having a demonstrated effect on returns. A narrower range may reduce handling but also change who buys. We therefore kept proposed actions separate from observed results. The original campaign's SGD800,000 contribution remained the historical account. The next test would have its own cohort and costs, including preparation. That separation prevents a good idea from appearing in the record as a completed benefit merely because management has approved it.
Make the next decision at the right level
The board chose a narrower follow-up test rather than repeat the promotion immediately at its original scale. The decision accepted a possible loss of short-term order volume in exchange for a better-defined offer and more useful evidence. It did not establish that the original campaign had failed. I think that qualification matters for trust between finance and commercial leaders. If every review must produce either a celebration or a reversal, colleagues become reluctant to discuss the uncertain middle where most useful operating judgement takes place.
An early dashboard still has a role. It helps teams plan labour, inventory and customer support, and it can reveal an unexpected response quickly enough to act. I would improve it by showing the parts of the result that are still provisional and the dates when they should mature. A director should not have to remember an unwritten caveat about return timing while looking at a confident sales total. The presentation can be clear and timely without implying that every order has already become a completed economic outcome.
There is also a question of incentives. If the commercial team is rewarded solely for opening orders and the service team carries the later returns cost, each department can appear to perform rationally while the company receives an unattractive result. The remedy is not necessarily one enormous measure assigned to everyone. It is a shared understanding of the journey and a review that connects the consequences. People should see how the offer, the customer expectation and the operating work affect the result they are being asked to improve.
For another consumer business, I would begin by asking when a sale is sufficiently settled for the decision at hand. The answer may differ across products, channels and customer terms. I would then examine the costs included, the recovery assumptions and the alternatives the board is considering. The lesson from our campaign is not to wait a fixed number of days in every circumstance. It is to let the next commitment be informed by a coherent account of what the customer kept, what the company did and what remained after serving that demand.