I once supported a portfolio account because its categories were clear and its arithmetic was correct. There were nine brands, each with a recognisable place in the presentation. The account made the business look broader, and I allowed that impression to support a claim about diversification. What I had not examined closely enough was the route through which institutions bought them. Six shared one procurement route. The names differed; an important part of their commercial dependence did not. The mistake was in the inference I asked the count to carry.
This fictional 2008 case at Vaurenne Therapeutics matters to me because it was not a problem with obviously careless work. We had described real distinctions within the invented portfolio. The weakness appeared when those distinctions were treated as an answer to another question: how independently the activities exposed the business to its commercial environment. Correcting the research and classification cost C$620,000. It also required me to acknowledge that colleagues had relied on an account I had endorsed, rather than present the later discovery as someone else's omission.
## The count answers a narrower question
A product count can be useful. It tells a reader something about the number of items being discussed and can help organise a complicated business. The difficulty begins when the unit of counting is silently exchanged for a unit of strategic independence. A brand name does not by itself establish a separate purchasing decision, route to market or source of demand. Those relationships need their own examination. The number may remain completely accurate while the conclusion drawn from it becomes much less secure.
In our review, Camille Dervelune challenged the connection between the nine names and the breadth we claimed. Her question did not require a belief that the shared procurement route would fail. It asked whether our account had made the dependence visible enough for management to decide how much of it the business was prepared to carry. That is a more useful discussion than treating concentration as either automatically unacceptable or irrelevant because the products belong to different categories. The relationship matters through its consequences for the particular choice.
The correction also needed a clear boundary around my competence. Commercial concentration is not a judgement about whether a medicine is appropriate, effective or safe. Qualified professionals own those assessments. A strategist can identify a shared buying route and ask what it means for the business without converting the observation into a clinical conclusion. Keeping that boundary explicit makes the commercial question more credible. It prevents a necessary challenge to the portfolio story from acquiring authority over subjects the evidence does not address.
## Keep the product difference and the common route
Colleagues responsible for the brands objected to reducing their work to a single dependency measure. Their concern was reasonable. Different products can have different histories, needs and operating contexts, and the common purchasing route does not erase those distinctions. The response should not replace nine separate boxes with one box and claim that the new simplification is the truth. We needed an account in which product differences and shared commercial exposure could coexist, because each might matter to a different decision.
I found it helpful to ask what management was trying to decide before choosing the dominant description. If the question concerned the variety of activities requiring professional support, the product distinctions could be central. If the question concerned reliance on an institutional purchasing route, the shared connection deserved more weight. A portfolio does not have to produce one universal classification for every purpose. It does need to make the purpose of the classification visible enough that a reader does not use it to answer a question it was not designed to address.
That approach takes more explanation than a simple diversification claim. It may also make a presentation less visually balanced. Neither consequence is a reason to retain an unsupported inference. The aim is not to make a portfolio chart as complicated as the business itself; that would defeat its purpose. The aim is to preserve the few relationships that would materially change the choice. In this case, the shared route was one of those relationships. Omitting it made the simplicity misleading rather than merely concise.
The correction had a personal dimension as well. I could not ask colleagues to preserve nuance in their product accounts while describing my own role as an uninterrupted sequence of good judgement. I had preferred a clear breadth story and had not tested the dependence beneath it. The research expense was part of the result, and so was the need to explain why the earlier recommendation had changed. Retaining those facts makes the case more useful to another leader than a version in which the strategist appears only at the moment of insight.
## Let the correction change the next decision
An acknowledged mistake matters most when it changes the questions asked later. In the fictional 2023 acquisition assessment, new therapeutic-area labels again offered an appealing breadth story. The proposed portfolio carried C$84 million in trailing annual revenue, with 45% tied to a familiar institutional purchasing group. The earlier correction made me more attentive to that connection. It did not settle the acquisition decision by analogy. The opportunity could still have been attractive for reasons other than diversification, and the corporate development argument deserved a proper hearing.
We ultimately declined after C$1.7 million in diligence expense. I would not describe that as proof that concentration should always prevent a transaction, nor claim a quantified loss avoided. The value of the earlier lesson was that it made the question more exact: what did this addition change, what remained shared, and why would the institution choose the resulting exposure? A business might answer those questions and proceed. The discipline lies in the quality of the account, not in making refusal the default sign of strategic seriousness.
For a leadership website, the distinction is equally useful. A list of brands, countries or transactions can establish scale but does not explain judgement. I would rather show one inference I revised, the evidence that changed it and the later decision it helped me examine. All details here are invented for that purpose. A real account would require verified definitions and permission to discuss the relationships. The enduring question is modest: before calling the portfolio broader, can I explain which dependence has actually become different?