I thought a preferred-contractor panel would make our renovation-finance proposition more useful. Households often needed to find someone to carry out the work, and relevant contractor relationships could also introduce prospective borrowers to the company. The proposal had a plausible commercial account, including ₹18 crore of projected annual sourced loan volume. I supported it. The question I had not examined sufficiently was what a customer would think our preferred designation said about the contractor's work.
In research with thirty-eight prospective users, fourteen understood the panel as assurance from the lender about work quality. They saw more than a convenient list. The ranking and preferred label suggested that the company stood behind a service it did not itself deliver. Our technical review existed for the lender's defined purposes. It did not make us the household's contractor or establish the continuing capability implied by the presentation. The problem lay in the proposition I had helped advance.
Ask what the endorsement appears to mean
The commercial sponsor proposed a qualification under the list. Some colleagues argued that customers should know the contractor was a separate business. Those responses focused on the statement we intended to make. Research had shown the statement some users actually heard. Both needed to be examined, but intention could not simply cancel the evidence about interpretation. If the strongest part of the design carried an endorsement, a smaller qualification might not give the service the narrower meaning we wanted it to have.
I asked the team to describe what would happen if the work disappointed the household. Who would assess the complaint, what authority would the lender have and which remedy could it actually provide? Those questions were not an attempt to write a construction contract from the product function. They exposed the operational responsibility attached to the perceived promise. A feature that seems adjacent to lending can create expectations requiring another kind of capability, even when the software change itself is relatively small.
The panel might still have been possible under a different, properly assessed service model. That would have required its own commercial, legal, technical and operating work. It was not what the current proposition supplied. I did not want a useful future possibility to become an excuse for releasing the existing design as though the capability already existed. The decision before us concerned the product we could responsibly offer on the evidence and responsibilities we had, not the service we might one day choose to build.
That distinction also prevented an easy claim that the research had proved contractor relationships were always inappropriate. It had not. Fourteen interpretations within a defined group gave us a reason to reconsider this presentation and its unsupported meaning. Other arrangements could have different terms, evidence and service obligations. I wanted the response to remain precise enough to explain, rather than turn a local design finding into a broad rule that could not survive examination elsewhere.
Compare the promise with the actual capability
We withdrew the ranked referral proposition. Households would select their own contractors, while the lender's separate technical and credit requirements would be explained on their actual basis. The change cost ₹24 lakh in product and communication work. It also removed the ₹18 crore projected annual sourced volume from this proposal's business case. That amount was an opportunity forecast, not income already earned or a realised loss. Keeping its status clear did not make the commercial consequence irrelevant.
The team had invested effort in the panel, and withdrawing it was more difficult than adding a qualification. People reasonably wanted to know whether better wording could preserve the useful convenience. We considered that question against the way the ranking and preferred designation worked together. The answer was that the proposition as developed invited too much confidence in the lender's role. I supported a narrower offer instead of asking the customer to reconcile an endorsement with a statement that the company did not supply the responsibility the endorsement suggested.
I also needed to resist describing the withdrawal as a measured customer harm prevented. We had evidence of interpretation in research, not an observed series of failed building projects caused by the released panel. The decision concerned a risk in the proposed promise and a capability the company did not possess. That is a sufficient reason for responsible product action without inventing a financial benefit from events that did not occur. A more modest account can make the judgement easier to assess.
The professional boundaries remained important after withdrawal. A lender may need technical evidence about work it finances, but that evidence should not be repackaged as universal assurance about the contractor. Equally, saying the household selects a contractor does not erase responsibilities the lender actually holds under its own agreements and applicable requirements. The product should explain the service accurately with the relevant owners' input. It should not use a disclaimer as a way to make all responsibility disappear.
Retain the cost of withdrawing your own idea
I keep my initial support in the story because it explains a change in judgement. I had concentrated on the usefulness of the connection and the acquisition opportunity before testing the meaning of the designation. Research forced the service question into the decision. If the final account made me appear to have opposed the panel from the start, it would remove the most useful lesson: a senior product leader can assemble a persuasive case while leaving an important implication insufficiently examined.
For a board, I would frame the issue as the distance between the promise a product communicates and the organisation's ability to fulfil it. Which feature or phrase supplies the customer's confidence? What continuing work does that confidence imply? Does the company have the capability, authority and resources for it, and do the commercial assumptions include those demands? Directors can examine those questions without choosing every label. They can also ask whether management has allowed research to change scope when the answer is uncomfortable.
The narrower proposition lost an attractive source of expected volume. It gained an account of the service that did not depend on a central impression being undone by a small qualification. I would test that impression earlier in future, before a useful idea becomes a growth commitment. The product's job is not to imply every adjacent service a customer might value. It is to make the contribution the company can actually support clear enough that the customer knows what confidence to place in it.