Startup research, done so the memo holds.
Bottom-up sizing, primary sources, AI research patterns, and the failure modes that produce confident but wrong outputs.
8 ARTICLES
Bottoms-up vs top-down TAM: which one investors actually trust.
Top-down TAM is faster to produce and easier to discredit. Bottoms-up TAM takes longer and survives diligence. Here is when each one is appropriate and how to build both.
B2B SaaS benchmarks investors actually use in 2026.
The benchmarks that defined 2021 do not apply in 2026. Here are the current numbers investors use to evaluate B2B SaaS at every stage from pre-seed to Series C.
How to do a competitive analysis for a startup idea.
Competitive analysis for a startup is not about building a features matrix. It is about understanding what the market currently pays, what buyers complain about, and where the gap is large enough to justify a new company.
TAM SAM SOM explained for startup founders.
TAM SAM SOM is a market sizing framework that most founders use incorrectly. The point is not to find the biggest TAM. The point is to build a defensible bottoms-up model that investors cannot poke holes in.
Market size is an argument, not a number.
TAM slides with nine-figure numbers impress nobody who has looked at one before. What investors actually want is the chain of reasoning that produces the number.
How to run customer discovery interviews that actually tell you something.
Most customer discovery interviews fail because founders ask about the future. The only questions that generate reliable information ask about the past. Here is how to structure a discovery conversation that produces evidence rather than enthusiasm.
How to do startup market research without paying for expensive reports.
Most of the market data you need to validate a startup idea is publicly available. The founders who do not find it are looking in the wrong places. Here is where to look.
Competitive analysis for founders: the parts most articles skip.
Most guides to competitive analysis tell you to build a feature matrix. That is the least useful part. The more useful parts are the ones that tell you whether you can win the economics, the sales cycle, and the switching battle.