Most new market segments look good on a spreadsheet. The TAM is large. The customer problem is real. But roughly 20 to 30 percent of entry attempts fail, usually not because the market is too small, but because conversion rates and willingness to pay are untested assumptions, not evidence.
The teams that win on segment entry do one thing differently. They work with incomplete data on both sides. Customer interviews give a wish list. Sales history gives a demand curve. Neither one alone is sufficient. Neither one is perfect. But together, they constrain the range enough to build an assumption set worth defending.
Three specific questions make this constraint visible.
Question one: What does conversion look like in this segment versus your core?
Your sales data from the core market is incomplete. It reflects the customers you already have, not the ones you could get. The new segment’s public signals are incomplete too. You can research competitor pricing and win rates, but you can’t see the actual closure numbers.
Start with what you know. Pull license conversion ratios from your historical data. Customer interviews, when you run them, will tell you what they wish they could buy. Build a scoring matrix combining what your sales data tells you about demand curves and what customer conversations tell you about pain weight. Weight the matrix by revenue potential, development cost, and market direction signals from competitor releases and adoption trends. Resource-load the outcome. The point is not to predict perfectly. The point is to make the reasoning visible enough that someone can challenge the numbers instead of the instinct.
Question two: What are buyers in this segment actually willing to pay?
Willingness to pay is not the TAM. It is the addressable portion of the TAM that survives your pricing and delivery model. An industrial SaaS company building its first bottom-up revenue model faced the same problem. Leadership had revenue projections that disagreed depending on who was in the room. We pulled their global asset base, applied historical license conversion ratios by project phase and product line, and generated an ARR forecast by geography. The model survived an external strategic review. It became defensible not because it was right, but because every assumption traced back to how the business actually converted opportunities.
Build one model that can hold up when someone from outside pushes on it. Not ten slides of assumptions.
Question three: Do you have the capacity to execute without hollowing out the core?
This is where the real constraint sits. If you redirect 10 to 20 percent of sales and product time to a new segment, something in the core market slows down. The segment bet is not free.
Two stakeholders at a product company measured the same launch with different metrics. The product team tracked active users. The business sponsor tracked revenue per user. Both numbers were moving up, but the roadmap decisions they implied were opposite. Once both read the same number, the backlog resolved. Not contentious. Just visible. When capacity questions are visible, the math either closes or it doesn’t.
Write the assumptions. Get them challenged. If the three questions point in different directions, that is useful information. It means either the conversion story is better than the willingness to pay, or the capacity exists but the demand signals are unclear. Any of those answers is better than betting on instinct.
