Choose the freight you can serve before choosing the customers you want.

In 2021, the commercial team faced eighty-four tender invitations during a defined planning round. Sales wanted to preserve access to every account. Operations questioned whether several proposed shipment patterns matched the service we could credibly arrange. My concern was that marketing effort would follow the size of the invitation rather than the quality of the fit, leaving proposal teams stretched across opportunities with very different operating demands.

I proposed a joint review before full submissions were commissioned. Sales retained the bid decision, operations assessed the service conditions and finance examined the commercial assumptions. I was accountable for the marketing recommendation and the use of proposal resources. Thirty-two invitations were declined and fifty-two proceeded. The rejected group described US$96 million of prospective annual shipper freight spend. That was not our revenue and did not establish the amount we would have won by bidding.

The trade-off was real: saying no could reduce access to buyers and remove a chance to learn through the tender process. I therefore asked for a reason on each refusal and a condition that could justify reconsideration. Nine of the fifty-two submissions produced awards. We reported both the starting population and the selected population, rather than displaying only a more attractive conversion rate. We did not calculate a fictional profit from work never awarded.

The enterprise gained an explicit account of where pursuit effort would go and why. It did not gain proof that every refusal was correct. Looking back, I would have scheduled a review of declined opportunities at the outset instead of treating that as a later learning exercise. A selection policy needs a way to discover its own mistakes, especially when the missing outcome makes the decision easy to defend indefinitely.

A 2022 proposal refresh included a service statistic from 640 selected industrial shipments. Of those deliveries, 608 met the originally agreed window and thirty-two did not. Twenty of the latter had a documented later window amendment. A commercial colleague wanted the updated windows to determine the headline because they reflected the final customer conversation. COO Beatrice Calverden asked whether a new buyer would understand that convention without further explanation.

I recommended using 608 out of 640, or 95%, against the original window, with the twenty amendments described separately. This was a communication choice about a defined historical set, not a judgement on every operational exception. The alternative could make the result appear stronger while answering a different question. We retained the underlying records and asked operating colleagues to confirm the shipment selection and dates before the note entered proposal materials.

The immediate cost was a less flattering headline and additional explanation during sales discussions. I accepted it because the percentage would otherwise travel further than its definition. We also stated that the selected shipments did not represent all company activity or establish future performance for a different mix of lanes and requirements. No new service commitment was created by publishing the example.

The revised note gave buyers a basis for asking whether the record resembled their proposed work. It did not produce a measured increase in trust or contract wins, and I do not claim either. I would now involve a buyer-facing operating colleague earlier in the drafting stage, before a percentage acquires the status of approved campaign copy. The useful lesson was not to prefer the lower number automatically; it was to preserve the question the number actually answered.

The 2023 plan proposed US$420,000 for exhibition activity. Sales valued the visibility and access that a substantial presence could provide. My review of larger account discussions suggested another problem: after the meeting, the people inside the shipper's organisation were not always working from the same explanation of our proposition. Procurement, receiving operations and finance requested different details, and a single introductory deck did not make those differences easy to resolve.

I moved US$120,000 into a set of decision materials while retaining US$300,000 for one event. The materials described the same arrangement from the standpoint of different responsibilities. They included service boundaries, the relevant material needed for a useful quotation and questions that should return to a joint discussion. Operating colleagues reviewed the factual account, and sales selected eighteen existing prospects for the initial use. This was a bounded programme, not a claim that documents could replace relationships.

The disagreement concerned lost exhibition presence and the risk that a less visible activity would be difficult to evaluate. I did not promise a pipeline return to settle it. Eighteen invited accounts used the materials, and eleven returned documented questions. The questions helped us revise unclear explanations, but we had no controlled basis for assigning later contract outcomes to the programme. The remaining seven accounts did not provide the same feedback and were not treated as silent endorsements.

The change gave sales a more consistent resource for the conversation after the first introduction. It also exposed disagreements between buying participants that a polished deck had left implicit. I would have specified the feedback request more clearly before distribution, including what a non-response could and could not mean. The decision was worthwhile as a better-supported buying discussion; its commercial return remained a separate question.

In 2024 we reviewed a campaign aimed at manufacturers with irregular replenishment needs. The message had emphasised broad flexibility, while the operating team was encountering particular requests that required more preparation than the wording suggested. The issue was not that industrial replenishment had ceased to be attractive. It was that the campaign allowed different readers to imagine substantially different arrangements under the same phrase.

I suspended three planned campaign releases and commissioned a narrower proposition review within US$180,000 of existing marketing authority. Sales argued that a pause would leave useful demand unanswered. Operations argued that continuing without clarification would invite unsuitable enquiries and difficult expectations. We agreed to keep direct account conversations open while withdrawing the broad public message. Marketing did not change the company's contractual obligations or decide which shipments operations should accept.

The review examined forty selected enquiries and grouped them by the service condition that needed clarification. Twenty-six fitted the revised description, nine required a separate operating discussion and five were outside the scope being promoted. Those counts described the reviewed enquiries, not the entire market. Two revised releases subsequently went live; the third was cancelled. We did not infer a reduction in complaints or a gain in margin from this limited exercise.

The cost included delayed promotion and the effort of explaining a withdrawal to colleagues who had already prepared activity around it. The benefit was a more specific public account of the work we were inviting. I would have asked for examples of unsuitable interpretations before approving the original wording. A proposition can sound restrained and still be too broad if the reader must supply the operating detail from their own expectations.

In March 2026 I authorised a US$360,000 investigation into an industrial-parts segment. The commercial attraction was a set of recurring shipper needs, but the team had not yet established how responsibility for buying and receiving was divided within the prospective accounts. A launch proposal could easily have treated those organisations as a uniform audience and moved directly into campaign preparation.

I chose a twenty-four-interview programme before recommending public entry. Sales helped identify relevant participants, and operations helped frame questions about shipment conditions. By October, fifteen interviews had been completed and nine remained. We recorded who described the buying process and which parts of the account still required another participant's perspective. We did not treat an interested response as a purchase commitment or use the first interviews to announce market demand.

The trade-off is time. A slower recommendation can miss a commercial opening, and colleagues reasonably want to know when research will be enough. I have retained the planned completion point and asked the team to distinguish questions that could change the decision from information that would merely make the presentation fuller. The operating assessment and financial recommendation must follow their own responsible owners; marketing research cannot settle them by implication.

There is no launch result to report. The current outcome is fifteen completed interviews, a clearer account of unresolved buying questions and a remaining programme inside the original authority. If I began again, I would include the receiving function in the initial recruitment brief more explicitly. The case remains open because a leadership account should show how an executive keeps a decision reversible while evidence is still arriving, not turn an unfinished study into a completed growth story.