Notes: Todd Jackson on PMF Levels

Four questions [Adler frame]

Q1. What is it about? A four-level framework for diagnosing B2B product-market fit and a set of accompanying practices — dollar-driven customer discovery, the four Ps pivot model, the marginal customer concept — designed to help early sales-led B2B founders increase their odds of reaching extreme PMF.

Q2. How is it argued? Inductive: Todd draws on First Round’s portfolio data (hundreds of startups), his own operator history (Gmail, Facebook newsfeed, Dropbox VP Product), and named company case studies (Vanta, Lattice, Looker, Plaid, Ironclad, Persona). The argument is empirical-observational rather than theoretical; the framework presents recurring patterns, not first-principles derivations.

Q3. Is it true? The four-level taxonomy is a useful diagnostic, but the level definitions and benchmarks are approximate — Todd repeatedly qualifies them as “rough” and “wide bars.” The $100 vending machine frame is memorable but is really a restatement of unit economics sufficiency, not a novel insight. The four Ps model is a useful decomposition; the insight that founders should try Persona/Problem before Product is well-supported by the case studies. The framework is explicitly scoped to sales-led B2B — Todd acknowledges consumer product development has more “alchemy” and is out of scope.

Q4. What of it? The most actionable takeaways for a B2B founder: (1) treat PMF as a continuous spectrum, not a binary; (2) qualify customers by economic commitment, not enthusiasm; (3) when stuck, change Persona or Problem before rebuilding the Product; (4) the marginal customer should be getting easier, not harder, as a health signal.

Glossary

Extreme PMF — widespread demand for a product that satisfies a critical need and can be delivered repeatably and efficiently to each new customer. Todd’s precision: all three dimensions (satisfaction, demand, efficiency) must be strong.

Nascent PMF (L1) — pre-seed stage; three to five paying customers; demand comes from founder’s network; product may be largely manual; efficiency metrics not applicable.

Developing PMF (L2) — seed/Series A; five to 25 paying customers; first scalable demand channel emerging; 500K–$5M ARR; 10% cold-conversion rate benchmark; some retention data available.

Strong PMF (L3) — Series A/B; 25+ customers; word of mouth operating; sales cycle shortening; customers advocate unprompted; repeatability demonstrated.

Extreme PMF (L4) — category-defining; customers would fight to keep the product; embedded in critical workflows; competitors reference you to explain their category.

Marginal customer — the next incremental customer you would acquire. In a healthy PMF trajectory, each marginal customer should be easier and cheaper to acquire and serve than the previous one.

Efficiency (dimension) — economic viability of customer acquisition and delivery; captures that satisfaction + demand without a working economic model is not real PMF (WeWork, Casper are named counter-examples).

Dollar-driven discovery — customer discovery anchored to economic behaviour, not opinions or preferences. Key probe: what did you pay for the last solution? Who controls that budget?

Four Ps — the four pivot levers: Persona (target segment), Problem (job being solved), Promise (positioning/messaging), Product (core functionality). Ordered by increasing cost to change.

$100 vending machine — hypothetical: a machine that dispenses $100 for $1 input. Maximal demand and satisfaction, zero efficiency. Illustrates that economic viability is a necessary PMF dimension, not just a later concern.

Friend zone — Rick Song (Persona founder) framing: a customer who uses, recommends, and talks enthusiastically about your product but will not pay for it, and would not be severely disrupted if it disappeared. Enthusiasm ≠ demand.

Regretted churn — involuntary churn from customers who wanted to stay; excludes deliberate downgrade or non-ICP customers. Todd uses this to separate satisfaction signal from efficiency signal.

The three dimensions in sequence

Todd’s key structural claim: the three dimensions of PMF (satisfaction, demand, efficiency) are not pursued simultaneously. The correct sequencing:

Investing in efficiency before satisfaction is established is a mistake. Investing in demand before the product can satisfy it burns credibility. The vending machine example illustrates what happens when demand and satisfaction are present but efficiency is absent.

The four Ps in practice

Three named examples:

Lattice — kept Persona (HR heads), changed Problem (OKR management → performance management), changed Promise (positioning), changed Product. Mid-2010s timing aligned with the pendulum swinging back to performance management as a priority. Jack Altman sold the pivot with Figma mock-ups before rebuilding.

Plaid — kept Product (bank account connectivity infrastructure), changed all other three Ps. Pivoted from consumer budgeting app to developer API. The product element they kept was the hard-to-replicate technical layer they had already built.

Vanta — changed all four Ps iteratively. Started with smart speaker (B2B Alexa), moved to dropshipping, then landed on compliance automation after customer discovery revealed universal hatred of SOC 2 questionnaire work.

Pattern: Persona and Problem pivots are frequently cheaper than they appear because the customer relationships already built can be leveraged to validate the new direction (Lattice example).

Dollar-driven discovery

The “don’t get friend-zoned” discipline applied to discovery. Standard customer discovery fails because people over-state enthusiasm and over-state willingness-to-pay in hypothetical framing. Todd’s correctives:

1. Anchor to past economic behaviour: “What did you spend on the previous solution to this problem?”
2. Identify the budget owner: who controls the line item this product would replace?
3. Test urgency: if the problem disappeared tomorrow, what would you do?
4. The Persona “talk” (Rick Song / Ironclad): sit the customer down one-on-one and ask explicitly whether the product is a necessity, whether competitors at half the price would cause a switch, whether their company would survive without it. The goal is to surface friend-zone status before it becomes a long-term liability.

Vanta as L1 case study [§ Level One]

Christina Cacioppo’s Vanta journey is Todd’s canonical L1 illustration:

• Multiple failed pivots (B2B smart speaker, dropshipping) before finding the SOC 2 compliance problem
• Customer discovery method: asking security engineers “what do you hate most about your job relating to security?”
• First product was entirely manual — she filed the SOC 2 herself (posing as the AI system)
• The critical insight: compliance certification was not just a pain, it was a revenue unlock — customers could land enterprise deals once certified
• Vanta changed all four Ps to arrive at its eventual form

Key L1 principle illustrated: efficiency is irrelevant if you are delivering transformative satisfaction. The manual spreadsheet was the MVP.

Looker as L1/L2 case study [§ Levels One and Two]

Lloyd Tabb’s Looker illustrates a different path: spending longer at L1 to nail a repeatable delivery model, then flying through L2.

• First five customers required 20–40 hours of hands-on founder consulting before sale
• Insight: customers cannot evaluate Looker without seeing their own data modelled in it — the demo must be personalised
• This became “forward deploy” — the Looker sales methodology
• Once the model was repeatable, they achieved 75% close rate and zero churn in L2
• Demand channels added: SDR cold outreach, AWS Redshift partnerships, look-and-tell customer events in San Francisco

The L1/L2 lesson: investing in delivery quality before demand scaling can pay off in dramatically higher L2 efficiency.

See also

• Todd Jackson on Product-Market Fit Levels, Customer Discovery, and the Four Ps
• Todd Jackson
• PMF Levels
• Product Market Fit Engine