The order is only the beginning of the retail result.

The first campaign dashboard showed SGD8.4 million of customer orders. The digital team wanted to repeat the promotion quickly, while its audience response remained strong. The initial discussion used the order figure beside costs that had already been collected. I asked the team to hold the repeat decision until we could see the defined cohort through cancellations and its normal return window. The question was not whether the campaign had attracted demand, but what the company retained from serving it.

Customer-service and fulfilment colleagues helped reconcile cancelled and returned orders. Net sales settled at SGD7.2 million after SGD1.2 million of cancellations and refunds. Product cost relating to retained sales, after the recovery treatment for returned goods, was SGD4.1 million. Fulfilment, payment and return-handling costs totalled SGD1 million, with SGD1.3 million of campaign media expenditure. That left SGD800,000 of management contribution before central expenses. It was not the company's statutory profit for the period.

Some commercial colleagues felt the delay risked losing momentum. We therefore continued ordinary trading while postponing the proposed repeat promotion, rather than stopping all activity. The review found that a small set of bulky products had created disproportionate return-handling work. Product and service teams examined presentation and packaging issues; finance did not infer the cause of every return from its cost. The account kept those operating questions open instead of using the revised margin as a complete explanation.

The board accepted a narrower follow-up test rather than an immediate repeat at the original scale. That choice gave up some possible short-term order volume and required additional preparation. The first campaign remained commercially positive, and we did not relabel it a failure to make the review appear valuable. Its contribution was simply less than an order-based account implied.

I took away a discipline about timing. A commercial decision may need an early view, but that view should identify the costs and customer choices still unresolved. Waiting for every fact is not always practical; pretending that the opening dashboard is the final result is not a sound substitute.

I would require the campaign’s full economic boundary at the first approval meeting, before the headline sales expectation becomes the reference against which every omitted cost must compete.

A proposal to raise the free-delivery threshold promised a quick reduction in delivery subsidy. The operating team could identify the smaller baskets that cost the company most to fulfil. The commercial team worried that an abrupt change would make ordinary purchases feel penalised and reduce conversion. I asked for a bounded test that examined contribution and customer response together, rather than approve the proposal from the delivery-cost line alone.

We ran a twelve-week pilot in a defined customer segment, with a contemporaneous comparison group. The pilot recorded 40,000 completed orders and SGD2.4 million net sales. Its contribution after product, delivery, payment and acquisition costs was SGD286,000, compared with SGD258,000 in the comparison group on the same management basis. The SGD28,000 difference was an observed comparison, not a proven causal effect: customer mix and purchasing patterns were not perfectly identical. SGD70,000 of implementation and communication cost was recorded separately and had not been recovered by that difference.

Customer-service contacts about delivery charges were higher in the pilot, and conversion was lower. The larger retained baskets did not mean every visitor was better served. Service and digital colleagues reviewed how the condition was explained and where collection could offer a practical alternative. I resisted a recommendation to call the test a success solely because contribution per completed order had risen; that measure omitted people who had abandoned their basket.

We retained a revised threshold for a narrower set of circumstances and improved the explanation before extending the trial. The team accepted additional communication work and a slower rollout. The revised decision did not promise that the next cohort would reproduce the same contribution difference, and the implementation cost remained in the account.

My role was to keep the unit economics and the customer's experience in the same discussion. A delivery promise can be commercially unsustainable, but changing it also changes the offer. The useful result was a better-defined choice about that offer, supported by a test whose limitations the board could understand.

The next delivery proposal would begin with finance and service colleagues agreeing the unit of contribution together, rather than reconciling their different accounts after the promise is priced.

One store's lease was approaching expiry while direct location sales were weakening. The property team had negotiated a five-year renewal involving SGD4.8 million of rent. The store also handled collection, returns and product advice for customers whose later purchases often appeared online. Closing it from its direct-sales account alone could remove useful work; renewing for five years without understanding that role could preserve an expensive location on the strength of an untested story.

I worked with property, customer and digital leaders to compare renewal, closure and a shorter extension. Linked customer activity gave evidence of cross-channel use but could not prove that every nearby online purchase depended on the store. We kept that limitation in the paper. The landlord offered an eighteen-month extension at SGD1.65 million, a higher annualised rent than the longer proposal. A focused refresh would cost another SGD220,000. Legal colleagues reviewed the terms and the obligations that would remain at the next decision point.

The operating team preferred the certainty of the longer lease, while some investor representatives questioned paying more per year for time. We proposed the shorter term with a defined review of the location's service role, customer activity and alternatives. The board accepted that proposal. It purchased an opportunity to learn, rather than establish that the site deserved a permanent place in the network.

The refresh was completed within the SGD220,000 budget. The extension began with the agreed rent schedule and a review calendar before its expiry. Those completed steps are the observed outcome described here. A final long-term location decision had not yet been made, so the profile does not claim that the extension improved lifetime store profitability or prevented a closure cost.

The difference between SGD4.8 million and SGD1.65 million was not a saving: the commitments covered different periods. My contribution was to make the price and purpose of the shorter choice explicit, including what would have to be decided later. Time can be worth purchasing when the board knows what evidence it intends to obtain.

I would ask the property team to cost a shorter commitment alongside the original lease proposal, bringing the value of a future decision into the discussion earlier.

A seasonal homewares range came with a supplier proposal for 100,000 units at SGD9.20 each. The merchandise plan showed an attractive margin at the intended selling price. A smaller commitment of 60,000 units would cost SGD10 per unit. Buying colleagues understandably preferred the larger order's price and availability. I asked the review to compare total exposure under different demand patterns, including markdown and storage consequences if the full quantity did not sell as planned.

The opening commitments were SGD920,000 and SGD600,000 respectively. The SGD320,000 difference described cash not committed at that point; it was not profit or a realised saving. We also examined replenishment lead times and the supplier's capacity. The smaller order left a risk that later demand could not be served at the same cost or speed. I wanted the commercial team to own that risk alongside the reduced opening exposure rather than hear finance present flexibility as free.

The chief executive approved the smaller initial order. When early demand supported more stock, we bought a further 15,000 units at SGD10.40, adding SGD156,000 to the commitment. Total purchasing for the two orders was therefore SGD756,000 for 75,000 units. The weighted unit cost exceeded the original large-order price, and some short periods of limited availability occurred while replenishment arrived. Those were consequences of the selected approach, not details removed from the eventual success account.

At the review point, 69,000 units had sold and 6,000 remained. We had not completed the selling cycle, so the final markdown outcome and total range profit were still unknown. We did not compare the unsold balance with an imaginary outcome in which all 100,000 units would certainly have remained at full price. The board saw the actual commitment sequence and the limits of the result then available.

The case helped us improve how buying papers describe uncertainty. A volume discount deserves serious consideration, but it should compete with a credible quantity alternative. Finance made the exposure visible; merchandise colleagues continued to supply the product judgement and respond to customers as the season developed.

I would obtain the supplier’s quantity options before fixing the seasonal buying envelope, so a higher unit price can be weighed against unsold stock without reopening an approved plan.

A seasonal gift-card campaign brought SGD2.2 million of cash into the business. The treasury forecast welcomed the receipt during a period of substantial merchandise purchasing. An early commercial presentation placed the amount beside completed product sales, making the campaign appear to have created an immediate revenue increase of the same size. I asked for a separate account of cash received, customer redemption and the obligations still outstanding.

By the review date, customers had redeemed SGD1.4 million against purchases, leaving SGD800,000 of unredeemed card value. The figures describe face-value movement in the defined campaign, with no assumed expiry benefit or breakage income. Reporting colleagues assessed the relevant accounting treatment and the product transactions separately. The receipt alone did not establish a completed sale of goods or free surplus available without future service cost. The business would still need inventory and operating capacity when customers used the remaining cards.

Commercial leaders were concerned that a more qualified presentation would obscure the campaign's usefulness. I agreed that it had provided cash earlier and could introduce customers to the range. We kept those benefits in the discussion while making the remaining obligation visible. Treasury included expected redemption timing in the cash and stock plan, with scenarios for a faster concentration of use after the holiday period. We did not assume that the historical redemption pattern would repeat exactly.

The revised board account separated the campaign receipt from net product revenue and reported the remaining face value alongside the planned service response. Customer-service teams retained the published terms; finance did not seek to improve the result by restricting legitimate redemption. The review did not claim a measured increase in customer lifetime value, since no sufficiently mature comparison supported that conclusion.

My judgement was to preserve the commercial opportunity without allowing its funding benefit to erase the future work. Advance cash can help a retailer, but it comes with a relationship to honour. Directors should be able to see both the improved timing and the resources the company will need when customers return to complete their purchases.

For the next gift-card campaign, I would agree the service obligation and its cash implications before launch, preventing receipt timing from setting expectations about money available for other uses.