The evidence that changed what we could responsibly offer

In 2018 I approved an eight-hour runtime headline derived from low-mode testing. The main use image beside it showed high-mode work. Under the defined test conditions, high-mode runtime was 3.2 hours. A review of thirty customer jobs found twelve needing six hours in that mode. The presentation encouraged a combination of duration and task that the evidence did not support.

We withdrew the unqualified headline and spent ₹37 lakh on repositioning and field-material repair. The case does not describe a battery defect, a recall or an invented customer loss. It describes the meaning of the product offer I had approved. The low-mode result could remain useful if presented with its actual condition; it could not stand unqualified for the high-mode task illustrated beside it.

I worked with engineering and commercial colleagues to rebuild the account around mode and intended use. Engineering retained testing, safety and release decisions. Product leadership had to make the proposition intelligible before a customer relied on the main impression. A later qualification was not an adequate substitute for that coherence.

The twelve jobs did not establish that every customer needed six high-mode hours. They showed a material need within the reviewed group that the broad headline could mislead. Other jobs could have different requirements, and runtime would still depend on defined conditions rather than become a universal assurance.

The outcome was a narrower, more accurate proposition and an explicit correction cost. I retain my approval in the story because the error was not merely a graphic designer's placement choice. I had allowed the product's most attractive duration to travel away from the mode that produced it. The case changed how I review the relationship between a specification, an image and the task the customer is being invited to imagine.

A 2022 study examined forty nonferrous workstations where customers wanted portable task lighting. Twenty-six needed an assessed clamp option and fourteen needed a bench stand. The magnetic base alone could not serve the settings we had observed. A simpler universal-base story would have made the range easier to present while leaving the placement question unresolved.

We funded a ₹1.1 crore programme and retained two bounded options. My recommendation concerned the product proposition and the customer situations it would address. Qualified engineering owners retained attachment suitability, safety and release decisions. The programme did not certify a clamp for every surface or invite users to improvise a mounting arrangement.

The study changed what we treated as central to the offer. Placement was not a minor accessory decision after the lamp had already been chosen. It affected whether the customer could position the light appropriately for the task. A feature that worked in the demonstration setting could not be assumed to travel unchanged into every workstation.

I worked with commercial colleagues to explain the two choices without presenting either as universally superior. The twenty-six and fourteen counts described this reviewed population, not the entire industrial market. A customer outside those settings would still require assessment rather than a forced fit into the nearest option.

The reviewed outcome is a programme and a clearer range choice, not a measured improvement in every customer's work. It also retains the cost of complexity: the enterprise had more than one placement proposition to explain and support. I considered that preferable to a simpler claim whose apparent universality depended on overlooking the environment in which the product would actually be used.

In 2023 we followed thirty-six prototype handover tasks. Fourteen users misread the brightness setting retained after restart. The product remembered a state, but the next person did not necessarily understand the state being inherited. The finding concerned the interaction between product behaviour and the user's interpretation, not simply whether the prototype could store a setting.

We funded a ₹68 lakh physical-mode-indicator programme. Engineering director Nikhil Tevqami retained control design, safety assessment and release authority. My role was to make the handover problem a product priority and ensure that the proposition accounted for the person arriving after the original user.

I did not want the response reduced to telling users to remember more. A portable product can move between colleagues and tasks. The account needed to examine how its current mode would be understood in that ordinary transition. At the same time, the study did not establish that a particular indicator would remove every misunderstanding. The revised work still required assessment.

The fourteen observations came from the defined prototype tasks. The case claims no accident, actual operating loss or universal rate of user error. It also does not provide instructions for modifying a lamp or overriding its controls. Responsible technical owners remained involved in deciding what the product could properly do.

The outcome was a funded design direction with the handover condition made explicit. I use the case to show why product leadership should examine the sequence around a feature, including the next person and the next task. A retained setting can be useful, but its value depends partly on whether the person using the product can recognise it. The feature's existence alone did not settle that experience.

A 2024 application study involved twenty-eight teams. Nineteen needed to compare colour-related appearance, while nine focused on geometry and shadow in their work. A brightness-only hierarchy would have put the range on one simple ladder. The study suggested that the product question should first distinguish the tasks customers were trying to perform.

We funded a ₹1.6 crore programme for two bounded application propositions. The decision did not establish universal colour accuracy, reliable defect detection or suitability for any regulated inspection process. Technical owners retained the assessment of optical performance and intended use. Our lamps remained within ordinary non-hazardous industrial settings, without medical, emergency or explosion-proof claims.

My contribution was to keep the task distinction inside the range discussion. More output could be relevant without being the only feature through which a customer should choose. A person comparing appearance and a person examining shape may need different questions answered before selecting a product. The product account should make those questions visible rather than imply that moving up a brightness tier resolves every need.

The nineteen and nine counts described the studied teams. They did not establish a permanent market segmentation or prove that the two groups never had overlapping work. The propositions would need to remain bounded by evidence and be revised if further learning challenged the initial account.

The reviewed result was a funded direction and a more task-specific way to explain the range. It contains no invented quality improvement or customer productivity gain. I wanted the enterprise to understand why it was carrying more than one proposition and which evidence each required. A clear range should simplify the customer's decision by preserving relevant differences, not by making every difference disappear behind one headline specification.

A 2025 qualified-service pilot received forty-eight lamps. Thirty underwent lens-only repair and returned to service after qualified acceptance. Eighteen required further housing or electrical assessment. The programme cost ₹56 lakh. I wanted the product's service proposition to describe that actual route and preserve the group whose outcome remained unresolved.

The pilot showed that a lens-only intervention could be appropriate for the thirty accepted units. It did not establish that every returned lamp was repairable in the same way or that customers should undertake the work themselves. Authorised specialists retained diagnosis, repair and return-to-service decisions. No procedural repair instructions are offered in this profile.

My responsibility was to connect the ownership experience with the capability the enterprise could support. A customer should understand what happens when a product needs attention, including assessment and the possibility that the required work is different from the simplest available repair. The existence of a service option should not become a universal lifetime promise.

I kept the eighteen units in the account instead of reporting only the thirty successful returns to service. Their further assessment mattered to the scope of the offer. The programme also did not establish a measured carbon benefit, financial saving or extended life for every lamp. Those claims would need their own evidence beyond the repair count.

The reviewed outcome includes completed qualified repairs and unfinished work. That combination is useful for product judgement. It shows an ownership proposition becoming concrete without making it broader than the demonstrated capability. I would rather explain the supported service route and its remaining limits than let the attractive word repairable imply that the customer no longer needs qualified assessment or that every future case will end the same way.