Strategy and validation for product managers

Top-down sizing narrows an industry total. Bottom-up sizing combines customer counts with revenue per customer. For example, 50,000 target companies at $1,000 annually gives a $50M annual opportunity; having 200 current customers does not establish how much of it you can capture. Use both approaches to cross-check assumptions where reliable data is available.

The required opportunity depends on costs, margins, distribution, and funding expectations. Capturing 20% of a $50M annual TAM would mean $10M in annual revenue, but neither that share nor profitability follows from the market size. Build SOM from reachable customers and capacity instead of applying an arbitrary 1% share.

The survey asks how users would feel if they could no longer use the product. A 40% very-disappointed share is a heuristic associated with Sean Ellis, not proof of product-market fit. Superhuman reported improving from 33% after segmentation to 58% over three quarters. Its example surveyed people who used the core product at least twice in the previous two weeks; choose an audience appropriate to your own product. Source: https://review.firstround.com/how-superhuman-built-an-engine-to-find-product-market-fit/

No. Review the survey alongside retention, repeated use, willingness to pay, and referrals. The PMF Score Calculator combines selected signals in its own model. A composite score does not turn them into a validated probability that the product has found a market.

ROI = (Total Benefit − Investment) / Investment × 100 over a stated horizon. Estimate incremental financial benefits after ongoing costs and include implementation time. Payback estimates when those benefits recover the investment. Customer acquisition payback benchmarks are not feature-investment benchmarks.

Compare payback with cash needs, risk, and how long benefits are credible. A $500K investment with $20K plus $30K in net monthly financial benefits pays back in 10 benefit months; add implementation time when measuring from the initial investment. That is a worked example, not an automatic recommendation to build.

Talk to prospective users about their needs, then test a prototype or limited version and observe use, repeat visits, and payment behavior. Survey responses and observed behavior both have limits. Ask the Sean Ellis question only of users with enough experience to answer; no fixed response count proves product-market fit.

Yes. Estimate the number of relevant customers from reliable public records, directories, or your own research, then apply a justified revenue-per-customer assumption. State coverage gaps and avoid counting overlapping groups twice. Industry reports can still provide a useful cross-check.