The Data Detective Method
An investigative approach to strategic clarity. Built across a decade of structural disruption. Designed to produce decisions that hold under commercial pressure, not reports that sit on shelves.
A different way of thinking about the problem.
Most strategy work begins with an answer. A framework, a model, a familiar conclusion drawn from previous engagements. The work then becomes the construction of an argument that supports the answer already implicit in the approach.
The investigative approach reverses the order. The work begins with a question and a hypothesis (an informed view of what the evidence is likely to show) and tests that hypothesis against whatever data the situation makes available. The data is rarely clean. The hypothesis is rarely correct on the first pass. The investigation revises both as it proceeds, and stops when the answer is clear enough to act on.
What this produces is not a more polished version of a familiar conclusion. It is a different kind of finding: one that can survive contact with the actual commercial pressure the organisation is facing, because it was built from the same evidence the pressure is producing.
The method, in five parts.
The method was refined across ten years of live work inside structural disruption. It is not a framework picked off a shelf. It is what The Tenth Floor is.
Every engagement begins from the evidence, not from a pre-formed view of what the answer should be. No off-the-shelf frameworks. No prior industry conclusions assumed to apply. The first job is to identify the question that, if answered well, would change what the organisation does next. Until that question is named precisely, the investigation has not begun.
This distinguishes the method from familiar consulting practice. Most engagements begin by selecting a framework and applying it. The Data Detective approach begins by examining the evidence and letting the framework, where one is needed, emerge from the structure of the problem.
The data does not need to be clean before the work begins. Most organisations hold significant evidence they have never properly interrogated: forum conversations, customer service logs, search behaviour, transaction patterns, sales call recordings. The value is in the investigation, not in waiting for perfect conditions that never arrive.
This has commercial consequences. An engagement can begin from the data the organisation already has, without commissioning new research, without rebuilding measurement infrastructure first, and often without the timeline most strategy work assumes.
A pattern that describes what happened is a useful record. A pattern that explains why it happened is actionable. The difference determines whether a finding produces a decision or a discussion. Most analytical work stops at description. The method requires going further, to causation, mechanism, and the underlying behaviour that explains the surface pattern.
The test is practical: can the organisation use the finding to change what it does, and would the change be the right one if the finding is correct? A pattern that fails that test is not yet finished work.
The output of every engagement is a specific, evidence-backed response to the question being resolved, fast enough to act on, and connected directly to the decision at hand. The deliverable is not a deck. It is a finding, with the evidence underneath it, and a recommendation that can be implemented inside the organisation’s actual operating reality.
An answer is shorter than a report. It is also harder to produce. The discipline of resolving the question, rather than describing the territory around it, is where the method does the work that other approaches do not.
The measure of a good engagement is that the next question can be answered without us. At the end of every engagement, the team that worked with The Tenth Floor can apply the same approach to the next question, independently. The frameworks, the dashboards, the measurement strategy, the working method: all of it transfers.
Consulting spend that accumulates without transferring capability is spend that has to be repeated. Method that transfers compounds. The first engagement resolves the question that prompted it. The second, run by the client’s own team, resolves the next one without external help.
Fixed scope. Fixed fee. Everything transfers.
Every engagement is fixed scope and fixed fee. No dependency created by design. The engagement is built from the beginning so that it ends. Everything produced (the findings, the dashboards, the method, the measurement instrument) belongs to the client’s team from the moment the work concludes. We are here until you don’t need us, and that is deliberately how we build the work.
The most reliable test of a method is not its theoretical coherence. It is what happens when it is applied to real problems, in real conditions, with real commercial consequences on the other side. Across twelve engagements, four continents, and a decade of disruption, that test has been run consistently.
A method developed inside the discipline of marketing strategy.
The Data Detective approach was first developed inside marketing strategy work, where the consequences of getting the question wrong are visible in commercial results within months. That pressure was the development environment. The method was refined across more than a decade of live engagements with senior leaders facing decisions where the inherited rules had stopped producing reliable answers.
The same approach now applies across strategic decisions outside marketing: operating model decisions, talent strategy, product positioning, organisational design. The principles do not change. The investigation becomes specific to the question.
Subhendu Mukherjee teaches related thinking at the University of Miami’s graduate marketing programme, and has lectured at Hyper Island, Google Squared, and Nanyang Technological University Singapore.
Fuski is a separate venture built by the same team. It applies aspects of the Data Detective method to a productised AI-assisted analysis tool. The Tenth Floor and Fuski are operationally distinct; mentioned here for completeness, not as a co-branded offering.
How information is handled.
For organisations sharing proprietary data, the terms are explicit. The work is done inside your environment. Data does not leave it — not because compliance requires this, but because the engagement should happen inside your world, not imported into ours. That is a structural commitment about what a consulting relationship should look like, not a list of protocols designed to limit liability.
- Every connection is verified before use.
- All data stays encrypted in transit.
- Every action is logged and visible to your team at all times.
- Access ends, and can be revoked, at any point you choose.
- Your IT and security team is involved throughout, not informed after the fact.
- AI tools used inside engagements are evaluated for data handling, retention, and training-data policies before use. Tools that train on client data are not used.
Curious whether the method fits the problem?
A first conversation with the leadership team is the most direct way to find out.