Each proposition left a different responsibility to resolve.

The 2023 range review at the fictional Qivaren Industrial examined seventy-two catalogue variants. Qiaorong Ke favoured a single successor story that sales colleagues could explain without navigating a long history of earlier products. The commercial attraction was clear: a smaller set of propositions could be easier to choose, quote and support. The initial analysis identified forty-eight variants as candidates for six common platform families. Twenty-four special-duty variants needed separate assessment. I wanted that distinction to remain part of the programme decision, rather than allowing the common candidates to stand for the entire range.

The special-duty group existed for reasons that had to be examined with engineering, service and customer owners. Some earlier choices might prove unnecessary; others might still serve a requirement the common proposition did not address. The product office could coordinate those questions and decide how the range was presented, but it could not declare interchangeability from a catalogue analysis. I asked for the family proposition and the continuing exceptions to be described together. That made the proposed simplification more complicated to communicate, which was exactly the cost Qiaorong wanted us to confront.

We committed CNY 42 million to six family development programmes and retained the twenty-four variants outside the common proposition pending their distinct assessments. The forty-eight remained candidates for that development work, not an announcement that every earlier product could already be replaced. Engineering validation and quality release retained their normal owners. We also kept the existing service obligations visible while the new work proceeded. A family name could organise the roadmap, but it did not extinguish a customer's need for an earlier configuration.

The consequence was a less sweeping launch narrative and continued complexity in the portfolio. I accepted both because the simpler story was not yet supported across the full set. The decision established a development scope, not realised savings or a completed universal conversion. The account should let a reader see what the organisation chose to fund and what it deliberately left unresolved. My lesson was to judge simplification at the point where the next colleague must use it. If sales can tell a shorter story only because service inherits an unexplained exception, the product decision has moved complexity rather than removed it.

In 2000, while I was a product manager at the fictional Lanveth Controls, a replacement selector covered 150 legacy references. The review had produced 116 functional mappings and left thirty-four references unresolved. I allowed the mapped group to be described as drop-in replacements. Minzhen Qiu challenged that description after identifying eighteen of the 116 that required connector changes. The selector contained useful work, but I had attached a promise about the customer's replacement task that the mapping alone did not establish.

My first instinct was to preserve the simple message and explain the exceptions in supporting notes. Minzhen argued that the phrase itself would shape expectations before anyone reached those notes. She was right. Functional correspondence and an unchanged connection were different questions, and neither alone was a full engineering suitability assessment. I needed to change the main claim rather than ask a qualification to undo it later. The error belonged to my product communication, not to colleagues who had correctly recorded the narrower mapping.

We withdrew the drop-in label and supported CNY 1.8 million review. The corrected account showed ninety-eight direct-connection candidates, eighteen references needing adapter assessment and thirty-four still unresolved. Those categories sum to the original 150; none disappears to make the result look cleaner. The ninety-eight were still candidates for relevant engineering review, not certified replacements for every installation. Customer conditions, engineering validation and the actual change process remained with their responsible owners. The public example deliberately contains no connection procedure or instruction to substitute equipment.

The correction required revisiting selector material and conversations that had already used the stronger description. It slowed a proposition I had supported and made the replacement route less convenient to explain. I accepted that consequence instead of treating the published wording as a commitment that the facts must now satisfy. The outcome was a more accurate account of the available mappings and the work still needed. No reduction in failures or completed conversion of the full set is invented. The case changed my approach to successor products: the customer should understand the assessment and changes required before the attraction of the newer item becomes a promise about how easily it can take the old one's place.

A 2015 portfolio proposal at Qivaren Industrial planned to retire a legacy range over twelve months. The customer account covered 480 installations. Three hundred and sixty had an available planned migration path; 120 could not complete the transition within that period. Colleagues supporting the original timetable wanted to stop dividing product and service attention between generations. Their argument had substance. Keeping an older range alive required people, parts planning and continued explanations to customers when the newer proposition was supposed to take its place.

I asked the team to start with the 120 installations that did not fit the timetable. A product sunset was an internal portfolio choice with consequences outside the catalogue. The fact that a successor existed did not establish that every customer could adopt it within the same window. Service and customer teams needed to explain the remaining constraints through their approved relationships, while engineering assessed the relevant transition questions. I did not propose that every legacy product should be supported indefinitely. I proposed that this retirement account should contain a realistic bridge for the identified group.

We retained a twenty-four-month service bridge for those 120 installations with CNY 16 million programme expense. The 360 with planned paths remained a separate group; a plan was not recorded as a completed migration. The additional year extended support work and postponed some of the organisational simplicity the original proposal sought. Commercial terms, service commitments and technical suitability retained their respective owners. The product programme coordinated the phased account without making a general lifetime-support promise or claiming that a single approach would fit every installation.

The decision therefore changed the retirement timetable for a defined part of the installed base. It did not establish that all 480 customers later moved or that the extended support produced a measured financial return. I would want a board or search reader to see both the benefit sought and the cost retained. A sunset can be strategically sensible while its first proposed schedule is impractical for part of the customer population. The useful leadership work is to explain how the institution will manage that difference, rather than turn the launch of a successor into an assumption that the earlier product's obligations have already ended.

In 2021, the product team examined a machine-monitoring proposition with ninety maintenance groups. Fifty-four required a local dashboard for the work under discussion, while thirty-six requested remote access. A unified remote-service offer appealed to sponsors because it would give the range a clearer recurring proposition and reduce the number of versions to explain. I wanted us to distinguish that commercial preference from the requirements collected. The fifty-four groups asking for local use had not simply failed to understand a feature the product team considered more advanced.

We explored the work those groups needed to perform and the conditions under which remote access would be considered. The product discussion had to include the relevant customer and specialist owners. I did not treat a request for remote capability as permission to activate it, or assume that a local option required no technical assessment. The choice concerned the shape of the offer. Security, connection arrangements and engineering approval remained separate questions. A sales narrative should not make those decisions by describing one configuration as the inevitable next generation.

We supported CNY 28 million development around a local core and a separately reviewed remote option. In a thirty-site pilot, eighteen selected local use and twelve selected remote capability under the applicable review. That split gave the team concrete configurations to examine; it was not a conversion rate for the entire market. The programme still needed to learn from operation and from customers outside the pilot. We made no claim that the choice established cybersecurity certification, universal suitability or recurring revenue merely because a remote option was available.

The selected design retained more product explanation and assessment work than a single compulsory proposition. Sponsors gave up some simplicity in the offer, while customers could discuss the capability that matched the work they had described. The outcome was the configured pilot and the deliberate option structure. It was not proof that the local choice would always be preferable or that the remote service would succeed commercially. I learned to treat a customer's decision to decline a feature as useful product evidence. The portfolio can remain ambitious without making every customer adopt the same ambition on the same terms.

The 2025 range proposal attached CNY 180 million potential orders to an assumed 1000 units across thirty-six options. It was an attractive reason to discuss capacity. Before supporting a commitment on that basis, I asked what configuration evidence sat beneath the commercial total. The validated request set contained 240 customer configurations: 156 standard and eighty-four bespoke. Those were requests, not firm orders. The mix did not by itself establish which products the factory should be prepared to supply, in what quantities or at what time.

Supporters wanted to move early so the business would be ready if the opportunity developed. Waiting for every detail could lose momentum and leave customers facing a slow response. I accepted that concern. My objection was to treating the proposed sales value as though it already answered the manufacturing question. A monetary total can combine configurations with very different implications. The product team needed to make those differences visible before asking operations and capital owners to rely on the forecast. It could not manufacture certainty by repeating the largest number in the proposal.

I declined to treat that account as sufficient grounds for the capacity decision and supported CNY 12 million configuration research. The 156 standard requests and eighty-four bespoke requests remained separate, with the assumed 1000 units retained as part of the proposal. Sales and commercial owners kept responsibility for customer commitments; engineering and factory colleagues assessed the implications of the requested options. The research allocation did not approve factory capital, announce booked revenue or establish that every bespoke request would become a supported variant. It provided a defined next step for testing the range proposition.

The choice risked a slower response to an opportunity that might prove valuable. I would include that risk in the leadership account rather than make further investigation sound costless. The result recorded here is the withheld recommendation and the funded configuration work, not a later investment success or a lost-order claim. The case made me more precise about what a product forecast contributes to a capacity discussion. The commercial ambition can be worth pursuing while the factory still needs a clearer account of the items, options and customer decisions behind it. The CPO should connect those questions without assuming authority over the final capital decision.