The disturbances missing from a convincing demonstration

## Look beyond the events on the screen

The display was clear because it contained events the instrument had captured. That observation sounds obvious, yet it is easy to forget during a demonstration. A viewer sees the successful output, hears the explanation and begins to imagine the product in use. The events that never reached the display are absent from that experience. Unless the evaluation preserves an independent account of what was presented to the instrument, visual confidence can grow without anyone confronting what the product failed to observe.

Our fictional bench exercise involved eighty defined brief disturbances. The first configuration captured sixty-one and missed nineteen. The displayed examples were useful for explaining the feature, but they were not the whole result. I asked the team to keep the eighty-event exercise visible and investigate the nineteen misses. The management question was not whether the screen looked convincing. It was whether the evidence supported the use for which the feature was being proposed, and what limitations needed to remain attached to that proposition.

There is nothing inherently wrong with choosing clear examples for a demonstration. People need an understandable introduction before they can examine a technical evaluation in depth. The problem arises when selected examples become the evidence for a broader conclusion. A presentation can explain how a feature works while a separate assessment establishes its limits. Those purposes should be identified, so that a successful teaching example is not quietly treated as a complete qualification result by an executive, salesperson or prospective customer.

The distinction matters particularly when the unobserved event is brief. The absence of an event from a display does not establish that no event occurred. Nor does it automatically explain why it was missed. Qualified measurement engineers need to investigate the relevant behaviour under the defined conditions. My role is not to prescribe a technical correction from the boardroom. It is to ensure that the missing evidence is treated as a product question, with the appropriate expertise and a clear account of its significance for the proposed claim.

I also ask what the exercise represents. Eighty defined events are an evaluation population, not every possible customer condition. Their selection may be appropriate for a particular development question without supporting a universal statement. The team should be able to explain why those events were included, what the exercise leaves outside its scope and how the result will inform the next decision. That explanation helps management use the test responsibly instead of attaching an expansive meaning to a precise-looking count.

## Preserve the exercise while improving the instrument

The redesign and evaluation cost ₹73 lakh. A later run captured seventy-three of the eighty events, leaving seven misses. That was a material improvement within the exercise. It was also incomplete. Both facts needed to survive in the account. If the team had quietly removed difficult events, changed the denominator or presented only the new successful examples, the comparison would have become less informative even if the resulting slide looked more favourable. Improvement should not require forgetting what the original question was.

Preserving an exercise does not mean it can never change. Investigation may reveal a flawed test, an irrelevant condition or a need for a broader assessment. When that happens, the reason for the change should be explicit and its effect on comparability understood. A corrected evaluation may be the right next step. What I resist is a silent change that makes the product appear to have improved without letting the reader distinguish engineering progress from a different test population. The management account should explain the basis on which a comparison is being made.

The seven remaining misses were not automatically equivalent to the earlier nineteen. Their causes and significance required specialist assessment. A simple count could show a narrower set of unresolved observations without proving that the residual risk was proportionately smaller in every use. I wanted the engineers to explain the findings in terms relevant to the product decision. Management should not infer a complete consequence assessment from arithmetic alone, particularly when the events may matter differently to the customer's intended activity.

The cost also needed a careful description. ₹73 lakh was development and evaluation expenditure. It was not automatically recovered value, avoided customer loss or evidence of future revenue. The work could support a more useful product proposition while leaving commercial outcomes unproved. Keeping the categories separate made the decision clearer: the company had spent money, obtained a stronger bounded result and retained unresolved work. That account was less dramatic than a transformation claim, but it was something directors could examine without first undoing several hidden assumptions.

There is a reasonable concern that exposing every limitation will make a product impossible to sell. I would distinguish an understandable account of material limits from an indiscriminate dump of development detail. Customers need a proposition that is both usable and supported. They do not need every internal discussion, but they should not be encouraged to rely on a capability that the company has not established. Commercial and technical colleagues can work together on that explanation without asking either group to abandon its responsibility.

## Explain a useful but bounded result

After the later exercise, the supported statement concerned the defined bench conditions and the observed results. It did not establish complete coverage in every customer environment. The company could decide to continue engineering, narrow the proposed use or obtain further evidence for a particular application. Those were active choices, not signs that the programme had produced nothing. A bounded result can be commercially useful if the boundary fits a real customer need and the relevant professional assessment supports the offer.

The technical leader should help identify that fit. I would ask which decision the customer intends to make, what the instrument must observe for that decision and which conditions require further investigation. The answer may differ across applications. A feature appropriate for one purpose may not support another, even when the hardware and display are unchanged. The sales conversation should therefore preserve the assessed use rather than treat every adjacent opportunity as covered by the same successful demonstration. This is a reason to connect product knowledge with commercial judgement early.

Directors can examine the discipline without becoming designers of the instrument. They can ask whether evaluation summaries retain unsuccessful observations, whether comparisons use a coherent basis and whether management distinguishes a bench result from customer evidence. They can also scrutinise the expense and the next decision that further work is intended to enable. A board should avoid demanding a perfect result merely because a residual limit is uncomfortable. It should instead expect management to explain the limit and the commitments being proposed in light of it.

The most useful question I now ask during a polished product demonstration is what happened outside the examples I am being shown. Sometimes the answer reveals a deliberate and appropriate selection for teaching. Sometimes it reveals an evaluation boundary that needs to be understood. Sometimes it reveals unresolved behaviour that changes the proposition. In each case, the question improves the conversation without dismissing the work. The screen can show something valuable; leadership requires ensuring that its clarity does not stand in for the events and conditions the screen cannot show.