Start from the application, not the industry
Industry classification is a convenient label and a poor predictor of demand. Two companies sharing a classification code can have completely different processes, while the same piece of equipment often serves several unrelated sectors.
A more durable starting point is the application: what the product physically does, to what, at what scale, under what conditions. From there the question becomes which organisations run that process — a question with a much more specific answer.
- What process does the product participate in?
- What inputs, outputs, capacities or tolerances define fit?
- What existing setup must be present for the product to be usable?
- What conditions rule an organisation out?
Layer firmographics on top
Once the application is defined, firmographics narrow it to organisations that can actually transact: size, geography, operating model, ownership of the relevant process rather than outsourcing it.
Keep exclusions explicit. A written exclusion is the part of a market definition that survives handover to the next person; an unwritten one is rediscovered expensively.
Identify the buying roles separately
The organisation is not the buyer. In most product purchases, at least three roles participate: someone who feels the problem, someone who validates the specification, and someone who controls the budget.
Write a separate value statement for each role. The technical validator and the budget holder are rarely persuaded by the same sentence, and a single message aimed at the average of the two persuades neither.
Attach a purchase trigger
A market definition without a trigger produces campaigns with no reason to arrive today. Triggers are events that move a latent need into an active one: capacity change, equipment failure, regulation, expansion, a new site, a contract win.
Where you cannot name a plausible trigger for a segment, treat that segment as lower priority rather than assuming demand exists.
Test the definition before you scale it
A market definition is a hypothesis. Test it on a small, tightly defined audience before broadening. The useful signal early on is reply quality and objection content, not open rate.
Record what you learn against the definition itself, so the next campaign inherits the correction rather than repeating the mistake.