Key Principle
A "customer" is not one person. It decomposes into the End User (who uses the product and usually becomes the champion) and the Decision-Making Unit (who controls the purchase). Steps 3-5 form the sharpest phase of the Holy Grail of Specificity: the entrepreneur profiles the end user (Step 3), sizes the market in dollars (Step 4), then instantiates a single real individual as the Persona (Step 5) — the North Star for every downstream decision.
The Persona must be a real person, not a composite. Composites allow each team member to project different assumptions; a real individual is falsifiable. When the team cannot answer a question from the fact sheet, they call the actual person and ask.
Why This Matters
Focus is fractal. Just as multi-market hedging fails at the market level (Step 2), broad end-user targeting fails at the customer level. Divergent priorities within a broad profile pull the feature set in multiple directions, dissipating resources. The corrective question: "Why would this specific end user want to use my product?" If the answer differs across the profile, the profile is too broad. (Step 3)
TAM sizing grounds the beachhead in economic reality. A TAM below $5M/year is too small to reach cash-flow positive; above $1B/year signals insufficient focus. The sweet spot is $20M-$100M/year — large enough to sustain the venture, small enough to dominate. (Step 4)
The Persona's purchasing priority hierarchy is the single most important element. It maps to what keeps the Persona awake at night — what will get them fired or promoted. Stated preferences often diverge from revealed behavior; validation through primary research is required. (Step 5)
Good Examples
Chuck Karroll — Mechanical Water Filtration (Step 5). The team assumed the end user was the data center manager. Primary research revealed it was the facilities manager. Chuck's purchasing priorities mapped directly to his performance review: (1) Reliability — downtime triggers CEO-level escalation; (2) Growth — supporting business unit targets; (3) Costs — partially forgiven if reliability and growth are handled; (4) Environmental/"Greenness" — nice-to-have extra credit, not a purchasing driver. The team's original value proposition was environmental friendliness. Primary research overturned it entirely.
Ed Champ — SensAble (Step 5). Selected despite being ~10 years older than the End User Profile because he understood the rational, emotional, and social dimensions of the end users. His fears (job loss from acquisition, shipping substandard product, losing design intent to engineers) and priorities (time to market, ability to express design intent, design intent surviving engineering handoff) provided two complementary lenses for product decisions. The Persona need not be a demographic clone — selection criteria are domain expertise, empathy with end users, and accessibility for ongoing questions.
OnDemandKorea — B2C TAM (Step 4). Funnel narrowing: 2.5M Koreans in US, narrowed to 1.2M already consuming content on 89 competing sites, filtered to 400K matching the End User Profile (60% female, 55% target age), converted at $15/year = $6M beachhead TAM. Each filter used observed behavior or third-party data, preventing projection from aspiration.
Counterpoints
The Overly Broad Profile (Step 3). Moscow ride-sharing case: "male and female, 17-40" with occupations spanning students to middle management and "medium or high social level." People across that range do not share goals, aspirations, or fears. The profile must be narrow enough that a team member reading it can picture a specific human.
Optimism Bias in TAM (Step 4). Entrepreneurs systematically inflate TAM. The goal is a conservative, defensible number the founder trusts — not a number that impresses. Inflated TAM attracts capital on false premises, leading to under-delivery and misallocated resources against a phantom market.
Multiple Personas (Step 5). Startups lack resources to serve multiple customer archetypes. Entrepreneurs who cite large companies' multiple personas confuse what is possible at scale with what is viable at startup stage. Multiple personas fragment focus, diluting product-market fit across all of them — the startup ends up mediocre for everyone instead of indispensable for one. Exception: two-sided markets get one Persona per side.
Key Quotes
"Often in entrepreneurship, your success is determined as much by what you do not do as by what you do." — On Anti-Personas: explicitly defining customers you decide not to serve. (Step 5)
"If I had only one end user to represent our End User Profile, who would it be?" — The guiding question for Persona selection. (Step 5)
"The deepest pain signals are identity-level frustrations, not feature gaps." — On SensAble designers whose artistic vision was distorted by tools built for engineers. (Step 3)
"A market opportunity = a specific end user + one or a handful of applications." — The segmentation principle that feeds the End User Profile. (Step 1)
Rules of Thumb
End User Profile checklist. Demographics (gender, age, income, location) + psychographics (motivations, fears, heroes, lifestyle, media, purchase motivation) + identity (what makes them special and identifiable). If the profile reads like a statistical abstraction, it is too vague. (Step 3)
Founder-as-End-User advantage. If a founder matches the End User Profile, assumption-based design is eliminated. If no founder fits, hire a target end user onto the executive team. (Step 3)
TAM formula. TAM = (End Users matching Profile) x (Annual Revenue per End User). Use dollars/year, not user count — revenue forces validation of both demand and willingness-to-pay. (Step 4)
Dual-analysis method for TAM. Bottom-up ("counting noses") via primary research + top-down via secondary research. Top-down alone systematically overestimates because it lacks customer-level specificity. (Step 4)
TAM sweet spot. <$5M = too small. ~$5M = viable only with fast capture and 90%+ gross margins. $20M-$100M = target range. >$1B = insufficient focus. (Step 4)
Proxy ratio technique. When direct nose-counting is incomplete, derive a density ratio from known data (e.g., designers per 1,000 employees), validate against companies with direct counts, then extrapolate. (Step 4)
Persona is a real person. Not a composite. Give them a real name (or alias), include a photo or drawing, post the fact sheet on the wall. Some companies make a cardboard cutout. The visual presence prevents drift toward abstract market thinking. (Step 5)
Purchasing priorities = incentive structure. Understand the Persona's compensation and performance-review metrics and you can predict their purchasing hierarchy. (Step 5)
Fears and priorities are complementary lenses. Fears reveal defensive motivations; priorities reveal offensive ones. A product addressing both becomes indispensable — it removes pain and enables ambition simultaneously. (Step 5)
Anti-Personas sharpen focus. Create personas representing customers you explicitly decide not to serve. This enables efficient redirection and prevents resource dilution. (Step 5)
TAM is iterative. It is a living number refined by later steps, not a fixed output. (Step 4)
Related References
- The 24-Step Framework — Core Philosophy — The 24-step framework and core philosophy
- Market Segmentation and Beachhead Selection — Steps 1-2: Market segmentation and beachhead selection
- Use Case, Product Spec, and Value Proposition — Steps 6-8: Use case, product spec, and value proposition