Follow the promise into the work it depends on

In 2017 I approved a tray-handling design after a demonstration organised around people responsible for purchasing our equipment. They could assess the intended business proposition and ask commercial questions. I allowed their approval to carry another implication: that the proposed handling routine had been adequately examined by the people expected to perform it. As head of applications, I owned that product recommendation. The mistake was not a condition imposed by an unnamed sales colleague; I had accepted the evidence as sufficient.

Later review involved eighteen operators. Eleven needed another person for the proposed handling routine; seven did not report that difficulty. Ruth challenged why those users had entered the inquiry so late. The commercial team argued that a redesign would disrupt promised delivery and that existing customers had already seen the cabinet. That concern was real, particularly for a small equipment business. Yet a successful demonstration for the purchaser did not remove the difference in operator experience. The findings were neither an injury count nor a universal safety assessment, but they made our incomplete product reasoning visible.

I supported a smaller-tray prototype with qualified engineering and usability review before a defined release. The correction programme cost ₹68 lakh, including ₹19 lakh customer credits. Engineering owners assessed the proposed design and actual handling conditions; customer quality owners retained their processing responsibilities. We did not tell operators simply to compensate for a design assumption that product research had failed to examine. Nor did we claim that a smaller tray automatically made every use suitable, improved food quality or reduced energy consumption. Those would have been separate claims needing appropriate evidence.

The change gave up the convenience of treating the demonstrated cabinet as finished. It required further work and an explanation of my earlier approval. I would now identify the operator research needed before the purchasing demonstration, rather than expect later user feedback to refine an otherwise settled proposition. The case changed how I understood approval itself. A buyer's authority answers whether an organisation can commit to a purchase; it does not necessarily answer whether the product fits the work another person must perform after that commitment. Both questions belong in the development account.

In 2022 our proposed product language described a broad set of crop-and-application combinations for a solar-assisted drying cabinet. Twelve combinations were being considered. Five had completed the defined application review, while seven remained unresolved. The commercial proposition made the whole set sound available. I wanted the release account to distinguish the applications supported within their reviewed conditions from those that still required investigation. A flexible product architecture did not complete every application assessment merely because the sales narrative could place the uses under one label.

The sales sponsor argued that a narrower release would make the equipment less attractive to processors whose crop mix changed through the year. That was a legitimate customer concern. Supporting only a limited set might not meet a particular processor's needs. I did not want to hide that limitation by presenting future possibilities as current capability. The decision needed to preserve the opportunity for further research while stating what the company could substantiate at launch. The seven unresolved combinations belonged in a development account, not in an implied approval supplied by the phrase multi-crop.

We limited the launch claim to the five reviewed combinations under their stated conditions and committed ₹1.6 crore to additional application research. The programme did not assume that all seven remaining combinations would ultimately qualify. Engineering and customer quality owners retained their assessments; my team was responsible for the product scope and the language connecting it with evidence. We published no operating recipe, preservation instruction or general food-safety claim. An actual use still required the relevant site, process and quality judgement, even where a product description identified a potentially suitable application.

The immediate result was a narrower offer and a clearer research agenda, rather than a larger approved market. Some prospects would need to wait or consider another route. That consequence belonged in the decision because the commercial objection was not solved simply by writing a careful qualification. I learned to distinguish variety in the opportunity from variety already supported by the release. A product team can keep future applications visible without allowing their commercial appeal to decide what has been demonstrated. The five and seven described different stages of work, and the proposition became more useful when it said so.

In 2023 we conducted forty-five interviews around the next product experience: fifteen purchaser-sponsors, fifteen operators and fifteen produce buyers. These were three groups with different relationships to the equipment, not forty-five interchangeable customers. Sponsors wanted a remote view of progress. Operators needed a clearer batch-record handoff. Produce buyers wanted documentation relevant to accepting the output. Combining the requests into one feature ranking would have made the research look decisive while concealing what each group was actually responsible for.

The commercial sponsor preferred the remote view because it would be visible during demonstrations and could help explain the investment to purchasers. Ruth argued that the operator handoff deserved earlier development, since the product experience depended on people retaining a clear account as work passed between them. I considered both claims. The product did need a persuasive purchase proposition, but a sponsor-facing display would not necessarily resolve the underlying handoff or supply the quality evidence a produce buyer required. We needed a sequence that recognised these different relationships instead of calling one request the winner of the research.

I supported a ₹2.6 crore programme prioritising the batch-record handoff and deferred the remote sponsor view for six months. Product colleagues had to explain that sequence to commercial teams, while customer quality owners continued to determine what documentation was relevant to their output. The handoff feature was not a quality certificate or a promise that every buyer would accept the produce. It was a defined improvement to the account moving through the processing work. We retained the sponsor's request in the roadmap rather than pretending that the research had shown it to be unnecessary.

The trade-off was a less immediately visible demonstration feature and a later delivery of the sponsor view. We did not claim a measured income, energy or quality benefit from the interview count or programme expenditure. Those outcomes would need separate evidence. The change was in the way product priorities were explained: whose responsibility a feature served, what it made easier to understand and what remained outside its authority. I would now structure the research around those relationships from the beginning, so that different participants' legitimate needs cannot be flattened into a convenient total that appears to settle the decision.

In 2024 thirty prospective processors were considering equipment. Sixteen wanted a trial route before deciding, while fourteen were prepared to pursue a purchase discussion. The latter figure did not represent completed orders. Sales colleagues preferred a purchase-first proposition, arguing that an additional trial programme would consume equipment, support attention and commercial time without a reliable return. I agreed that a trial was a commitment by the supplier. I disagreed that its only useful conclusion would be a customer deciding to buy.

We established the trial route for the sixteen prospects, with a programme expense of ₹87 lakh. Eleven completed the process and five withdrew. Of the eleven completers, six subsequently purchased and five did not. Those groups needed to remain visible. Reporting six purchases without the five decisions against purchase and five withdrawals would turn a learning route into a selective sales story. Equally, calling the other outcomes failures would assume that the product ought to fit every participant. The trial existed partly to make an unsuitable commitment less likely, without claiming that it had prevented a quantified loss.

I required product research to retain the reasons participants were willing to share for continuing, withdrawing or declining. The actual technical and customer quality assessments remained with qualified owners; participation did not establish a universal operating result. Commercial terms and responsibilities needed to be clear before a processor entered the programme. We did not present the trial as financing, assured output acceptance or an income promise. Nor did we divide programme expense by subsequent purchases and describe that arithmetic as a causal return from product research.

The outcome gave us a more complete account of fit than an order count alone. It also retained the cost and complexity sales had warned about. I would now define the useful decisions a trial can support before deciding how its success will be reported. A participant who learns that the equipment does not suit the intended work may have reached a valid conclusion. Product leadership should be able to explain why that evidence matters, while leaving the commercial organisation free to examine whether the programme is an appropriate way to spend resources. Honest endings are part of the product route.

In 2025 a revised equipment proposition implied that processors would gain a dependable buyer for their dried output. The commercial idea was attractive: equipment would appear as part of a route to market rather than as another investment the processor had to justify alone. Yet discussions with nine prospective produce buyers had produced only three expressions of nonbinding conditional interest. Six had expressed no interest. None of those categories could responsibly be presented as a general offtake arrangement for customers purchasing our equipment.

The channel sponsor argued that introductions were a useful part of a more complete offer and that removing the promise would weaken its appeal. I agreed about the potential value of introductions. The issue was the commitment being implied. A buyer might want to examine a product, quality documentation, quantities or commercial conditions before considering an order. An equipment purchase did not supply those decisions. I asked for the actual status of each discussion and the language customers were receiving, rather than allowing the ambition to provide a broader service to become evidence that the service already existed.

We removed the unsupported market-access promise and spent ₹54 lakh on product and channel rework. The revised account distinguished possible introductions from purchase contracts and left produce buyers' quality and commercial decisions with them. Legal and commercial owners reviewed actual terms; the product team aligned the offer with what those terms could support. We did not invent an offtake volume, customer income or payback period. The three conditional expressions remained relevant relationship evidence, but they did not become three assured outlets merely because that would make the equipment story easier to sell.

The immediate consequence was a less expansive proposition and more work for a processor considering the wider business case. That was precisely why the missing commitment mattered. The product should not let the customer discover after purchase that market access remained largely theirs to establish. I would examine the surrounding promise earlier when equipment is packaged as a business opportunity. A supplier can help with relationships and still be honest about their stage. Product leadership must distinguish what the equipment can support from what an independent buyer has actually agreed to do.