I keep fumbling market sizing because I either over-engineer a tree or skip a sanity check under the timer. I’m trying to lock a simple, first‑principles flow I can run reliably in 8–10 minutes and defend without sounding cagey.
Here’s my current draft: clarify scope in 15–30 seconds, decide top‑down or bottom‑up, pick a clean base (population, households, installed base, or firm count), then stack two or three levers max like penetration, frequency, and price. I do one quick triangulation path if time permits and end with a sanity check against a known adjacent market. I’m also forcing myself to say assumptions as ranges, not single points, and collapse to a midpoint.
For folks who’ve done this a lot: what guardrails keep your math honest and your narrative tight under a timer? If you had to set default anchors for common categories like consumer services, marketplaces, or SaaS, what are your go‑tos? Also curious how you defend assumptions quickly without turning it into a debate.
stop worshipping frameworks. it’s just base x behavior x price, done fast and clean. pick one anchor you can defend, speak ranges, round aggressively, and do a single sanity check against a market everyone knows. if you’re off by 2x but consistent, you’ll pass. elegance loses to speed and clarity in these things.
top‑down unless the product is extremely niche, then bottom‑up. don’t stack five levers; two or three max. never quote something to the third decimal—interviewers smell fake precision. say your assumption, give a one‑line why, move on. then sanity‑check per capita spend. practice with a kitchen timer, not a powerpoint.
i’ve been using a tiny mantra: who buys, how often, how much. i say it out loud so i don’t drift. then i sanity check vs a known category i remember (groceries, rideshares). feels basic but keeps me from spiraling.
i timebox steps: 1 min clarify, 6 min math, 2 min check, 1 min wrap. when i stick to it, i don’t panic. also i round to easy numbers so i don’t get lost in calc’s.
Two principles keep these tight. First, define the economic buyer and the unit of analysis in the opening sentence. That reduces branching later. Second, restrict yourself to two controllable levers plus price, then narrate why each lever is the right driver for the category. Defend assumptions with order-of-magnitude references, not trivia. For example, compare per-household spend to a neighboring category the interviewer likely knows. Close by stating the implication in business terms: even at the low end of your range, the market supports X scale or Y unit economics. That converts arithmetic into judgment, which is what they’re evaluating.
You’ve got a solid flow already! Keep it lean, speak ranges, and end with one crisp implication. Practice a few reps and you’ll feel the rhythm. You’re closer than you think!
I bombed a fintech sizing by building a gorgeous tree that ate seven minutes before I touched numbers. Fixed it by opening with one line on scope, then ran households x adoption x ARPU, and did a quick per‑cap sanity check against credit card interchange. The interviewer actually nodded when I said, “Even at the low end, this supports a seed‑stage GTM.” It wasn’t perfect, but it sounded decisive and tied to business reality. The shift was keeping it to three levers and speaking ranges, not points.
Allocate time intentionally: 60 seconds to lock scope and unit, six minutes for math, 90 seconds to triangulate, 30 seconds to implications. Use stable anchors like households, firm counts, or device install base. Convert to spend through penetration and frequency, then multiply by a realistic ARPU. Sanity-check by comparing per-capita spend to adjacent categories and ensuring your result isn’t exceeding known benchmarks. Round numbers early to keep arithmetic clean and state ranges to avoid false precision. Finish with a business takeaway instead of lingering on decimals.
When choosing top-down versus bottom-up, use data availability and concentration as a rule: fragmented consumer categories favor top-down, concentrated B2B niches favor bottom-up via account counts. Pick one alternative path for triangulation, not three. If your two paths disagree by more than 2x, identify the lever causing variance and reset that single assumption. Close by normalizing the outcome per buyer or per day to see if it passes a reality check. That quick normalization catches most order-of-magnitude mistakes.