Jason M Lemkin on Building a B2B Sales Team, Sales-Product Tension, and the VP of Free

Literature notes on the Lenny’s Podcast conversation (recorded 2024). Jason M Lemkin founded SaaStr and co-founded EchoSign (grown to $100M ARR, sold to Adobe). The conversation is a tactical manual for building a B2B sales organisation from the founder’s first close to a scaled org, then turns to the product–sales relationship and a contrarian defence of free.

Four questions [Adler frame]

Q1 — What is the whole about? How a founder who does not consider themselves a salesperson builds a sales team: when to hire, whom to hire, how to pay them, how to scale the org, and how to keep product and sales in productive tension rather than open war. A coda argues that freemium and long free trials build more durable businesses than the 2023-era tightening of trials and price increases.

Q2 — How is it argued? Almost entirely from operating and investing pattern-experience — Lemkin draws on EchoSign, his portfolio companies, and a decade of SaaStr conversations. The mode is the disconfirming anecdote: a single counter-case (“they fired the sales team and went from $5M to $1M”) carries the rule. Heuristics are stated as near-laws (“rules of eight”, “would I buy from them?”, “hire someone whose last product was harder to sell”) then stress-tested against scale.

Q3 — Is it true? Internally consistent and grounded in named cases (Pipedrive, Intacct, Salesforce, Slack, Canva). The reasoning is anecdote-driven, so [?] the rules generalise less cleanly than stated — “interview 30, hire 2” is a vivid ratio, not a measured one. The strongest, most general claim — compounding is the only thing that matters in B2B; anything that anti-compounds is a permanent cost — is asserted, not demonstrated, but coheres with the named billion-ARR examples.

Q4 — What of it? A field-ready sequence for founders crossing from founder-led to rep-led sales, and a sharp frame — the VP of Free — for the structural blind spot in PLG orgs: no one is incentivised to defend the non-converting free base. The compounding lens reframes growth tactics as either building or eroding a balance that determines whether a company reaches durable scale.

Glossary

OTE (on-target earnings): total expected compensation for a sales rep at quota — base salary plus bonus/commission. Conventionally split 50/50 base to bonus in SaaS [§ Comp and quota].
AE (account executive): the seasoned rep who works and closes leads. SDR (sales development representative): entry-level opener who prospects and screens inbound, then passes qualified leads to an AE [§ Roles and titles].
Carrying a bag: a sales leader personally holding a quota and closing deals, not only managing. Lemkin’s screen for whether a VP of Sales still cares about the craft [§ VP of Sales].
A+ middler: Lemkin’s term for the founder’s natural sales strength — fluent in the product and the customer’s problem (the middle of a sales conversation) even when weak at opening (outbound) and closing (asking for money) [§ Founder as middler].
Rules of eight: the span-of-control heuristic for building a sales org — roughly one manager per eight ICs, one director per eight managers, and so on [§ Scaling the org].
VP of Free: the (almost always absent) role that champions the non-converting free user base — distinct from a VP of Growth, who is incentivised to monetise that base [§ VP of Free]. See VP of Free.
Compounding / anti-compounding: revenue that recurs and grows because customers stay and expand; anti-compounding moves (dark patterns, unearned price rises, forced annual contracts) raise short-term revenue while permanently damaging the relationship that produces recurrence [§ Compounding].
Full-stack AE: the (per Lemkin, largely mythical) rep expected to prospect, manage their own outbound list, and close. In practice reps specialise into openers and closers [§ Roles and titles].
B2D (business-to-developer): selling technical products to engineers; a distinct motion where most business-process reps fail [§ Harder-to-sell heuristic].

When and whom to hire [§ When to hire]

The decision is when, not if — almost every B2B company eventually needs sales; even Canva (self-serve to ~$500M revenue) and Slack (self-serve at launch, majority enterprise by IPO) ended up with sales teams. The trigger is twofold: the founder has closed the first ~10 customers personally, and sales is consuming more than 20% of the founder’s calendar. Lemkin’s standing rule for a founder-CEO: 20% of time in sales, 20% in recruiting — cross either threshold without leverage and the calendar dies.

Be honest about the motion the first 10–20 unaffiliated customers require. If they put a credit card in with no human contact, the DNA is self-serve — invest in viral loops and growth, not reps. If they need calls, security answers, and onboarding help, that is a sales motion, and producty founders who flee it because “sales is icky” risk collapse [?anecdote: a portfolio company fired its sales team and fell from $5M to <$1M in 20 months].

The founder as middler [§ Founder as middler]

Founders fear they cannot sell, but selling has three parts — open (outbound, cold email), middle (demo, answer questions, talk workflow), close (ask for the next step and the money). Founders are A+ middlers by day one: no one knows the product and the customer’s problem better. Early adopters love talking to the CEO. The growth task is narrow — learn to ask for a next step at the end of every conversation. Reps, by contrast, are weak in the middle: past five basic questions they cannot explain how the database works or how to do payroll in an obscure jurisdiction, which is why early reps need founder/SE support.

Hiring the first reps [§ First reps]

Hire two, not one — a single first rep is a coin flip; two let you A/B-test humans. Expect to interview ~30: roughly 20 break your brain (no prep, didn’t read the site), ~8 are okay-but-no, and 1–2 are magicians. The one durable filter is “would I buy my product from them?” — not the LinkedIn, not the logo (Twilio, Cloudflare, Snowflake), not the matching deal size. Lemkin’s own first rep was living in his brother’s garage, mid-pack on paper, and closed the company’s first five-, six-, and seven-figure deals over a decade, because he could explain the problem and the solution credibly.

Constraints on the filter: the first reps still need a couple of years of B2B experience and enough maturity to be trusted with precious early leads — there is no time to babysit or train. Leads are the scarce resource: a “perfect” hire from Snowflake who fumbles two of your three monthly leads can kill the company. Look for “pirates and romantics” — a quirky rep with a few extra IQ points who has, for no logical reason, fallen in love with a feature-poor early product.

The interview’s core test is “sell me this app” (not “sell me this pen”): give a real assignment, let them watch the explainer video and webinars, have them demo the product back. ~98 of 100 candidates will not even watch a public explainer video before a late-stage interview. The exercise both filters and connects directly to the “would I buy from them?” criterion. Lemkin notes the wider hiring-discipline collapse since early 2020 — reference checks and demo assignments were dropped in the talent scramble and have not returned, which fails both company and candidate. Corollary he repeats: anytime a hire fails, it is the founder’s fault — you hired them; be kind.

VP of Sales: timing and the bag [§ VP of Sales]

Do not hire a VP of Sales until two reps are consistently hitting quota. A VP’s job is to take you from rep 3 to rep 300 — to scale a motion that already repeats. Hire one before the motion exists and you are asking one person to find product–market fit, be rep one, be rep two, and scale all at once: “approaching 100% chance of failure,” gone in ~8 months, with ~$2M of the round burned because they never understood the product. The founder must first be the (bad) head of sales — manage the first reps, sit in the field — before handing off.

Most first VPs should be a stretch (a former director/senior director), not someone on their third VP gig — veterans on their third tour often no longer want to do the work. Screen hard for carrying a bag: in the current market Lemkin sees too many VP candidates who only want to manage, are in zero deals months in, and cannot answer “how’s the Airbnb deal going?” His first screening question is literally whether they still want to sell. The diagnostic Colombo question for any VP (sales or product): “What would you do in your first 14 days?” If the answer is not “go meet ~20 customers” but “build process / set up Salesforce / plan territories,” do not hire — for any senior customer-facing role. Even at Salesforce’s scale, Benioff still flies to customers and sits at Davos taking 50–100 meetings a day; “sales is just dashboards” is true only at $4B–$10B+, beyond most listeners’ stage.

Comp and quota [§ Comp and quota]

Founders over-think the OTE number and under-think the cash commitment. Pay market — a rep wanting $150K OTE at 50/50 is a ~$70K base, ~$6K/month, so a two-month experiment is ~$20K of cash at risk. (“You don’t have $20K to test sales? Then don’t do sales.”) Negotiating $150K down to $130K is irrelevant; what matters is whether the rep can close 3–5× their take-home (3× SMB, 4× mid-market, 5× enterprise) — and you want them rich, because a rep paid $150K who closes $450–600K is accretive unless marketing costs are broken (not sales’ fault).

Early ramp: let a new rep keep 100% of what they close for one quarter (max) — unprofitable but it puts food on the stove and points on the board without crushing pressure. Run it longer and mediocre reps lean into it. A well-built plan rarely needs revisiting for years; over-tweaking just burns cash plugging a leaky boat. Prefer fewer, better reps early — two reps closing $1M each beats twenty struggling at $100K (domain knowledge, culture, leads concentrated on the best closers). Capacity planning — needing N reps at $500K each to add $50M — is real only at scale.

The single highest-leverage hiring heuristic: hire someone whose last product was harder to sell. A rep coming from an easier product arrives without the skills your product demands; a rep from a slightly harder product feels “weight released from their feet” (Sam Blond, first SDR, came from Intacct selling cloud financials — the least-trusted thing to put in the cloud — and was instantly #1). Markers of “harder”: more technical / B2D, more competitive (a strong rep from the #4 player entering at #2 is gold), more complex business process. Beware reps who can only reduce a complex product to a slogan (“we’re the Gong for X”) and lose the $600K deal when asked how the integration actually works.

Scaling the org [§ Scaling the org]

Sales has diseconomies — roughly half the company is in sales at $10M, $50M, $100M; public companies carry higher CAC than startups because of market penetration. Build on rules of eight: 8 SDRs → 1 manager, 8 AEs → 1 director, 8 directors → 1 VP. SDR spans briefly pushed to 12 but have reverted to 8. A stretch VP scales only if they can hire managers above the rep level — promote ~50% from within, hire ~50% from outside; the org cracks when a first-time VP can only promote junior reps into management. There is no real senior-IC track in sales (unlike product/engineering) — a structural gap that pushes good closers into management they neither want nor are suited for, and tempts founders to make the #1 AE the head of sales (different skills entirely).

Product–sales tension [§ Product–sales tension]

The best B2B orgs put the head of product deep in sales. A VP of Product who truly owns the roadmap and has gravitas becomes a “mini-CEO” in big deals — able to say “we can launch on this tomorrow and tweak the UI in 90 days,” a commitment no AE or VP of Sales can make. Mid-pack PMs working on pixel colour should stay out of deals; the bar is fluency in the entire product plus authority to commit.

Managing sales’ feature demands: give the head of sales a fixed quarterly engineering budget — ~10% of story points / the pie chart — that is theirs, period. This forces force-ranking across the whole team instead of the loudest deal winning. Mid-quarter changes are allowed but carry explicit cost (work already in flight). Two parallel weekly rituals make it work: sales does a whiteboard force-rank of requests (the output reliably differs from where the meeting started — Monday’s must-have HubSpot integration yields to Wednesday’s SAP request), and product spends ~15 minutes confirming it is building the right sales priorities (where an engineer may reveal “I can build that integration in a day”). IC-to-IC roadmap conversations should happen (a reason to be in an office) but decisions escalate to VP level — empower the IC to recommend, not decide, or one stressful VP conversation becomes a hundred. Stress between product and sales is healthy; its absence means you are not in enough deals — but it needs enough structure not to break the org.

On product owning P&L / revenue KPIs: Lemkin wants the VP of Product aligned to the bookings number, but is wary of weaponising the average PM with revenue quotas on experimental features — “ask me in a year.” His cautionary parallel: customer success was reoriented from retention to revenue across the industry in ~2023 and it destroyed relationships [?anecdote: a beloved public SaaS vendor demanded $50K upfront from a $299/month customer or shut-off that week — then walked it back, which is almost worse]. Big companies are better than startups at letting initiatives fail without firing people, which is what makes anyone willing to staff the new thing.

Freemium, trials, and the year of the customer [§ Freemium and trials]

Lemkin is contrarian on the 2023 tightening of trials and contracts:

The 14-day trial is an artefact, not a finding — Salesforce moved 30→14 days so its SMB reps could close deals that month, with no evidence it helped customers or usage. Audit inherited metrics for whether they are customer-centric.
Long/infinite trials built epic companies — Slack, Canva, and (until recently) Zoom were content to wait years to convert.
Annual contracts for SMB is bad VC advice — it looks great on a spreadsheet, but individuals and small buyers want to pay monthly on a card; forcing $240 up front for a newly discovered product makes them bounce. Big companies with procurement want annual — let customers pay how they want.
Earn your price increases. Everyone raised prices in 2023; almost no one added the value (a genuinely better edition) that historically justified it — they “just sent an email.” One year’s pass, then earn it.
Free products are better software — a product that cannot offer a free edition cuts corners because human onboarding bridges the gap between buying and deployment; investing engineering in a free edition makes the whole product (even the enterprise tier) better.

The VP of Free [§ VP of Free]

The organising idea of the coda, and the episode’s novel concept. In a PLG company, who champions the non-converting free base? Growth is bonused on conversion, marketing on immediate monetisation, customer success has been weaponised into revenue collection — so no one defends the long tail. Using Lenny’s newsletter as the live case: ~600K readers, perhaps ~1% convert; the 590,000 who do not are advocates, community, and future converters. Over-monetise them and the community (the flywheel) collapses. In most orgs the VP of Free is, by default, the founder working with product — and Lemkin’s point is that this brief should be held deliberately, not left to fall through the gaps. See VP of Free.

Compounding [§ Compounding]

The unifying claim: in B2B, nothing matters except that revenue compounds — billion-ARR companies (Cloudflare, others) exist only because revenue recurs and grows. Matt Mullenweg told Lemkin the thing he most underestimated over 20 years at Automattic was the power of compounding. Customer-hostile moves (the 2023 dark patterns, unearned price rises, weaponised CS) anti-compound — they buy short-term revenue at a permanent cost to the relationship that produces recurrence. For a low-end/PLG product the master metric is churn: top-decile churn or Lemkin is out as an investor, because 3–4%/month SMB churn is “almost unsolvable.” Relatedly, sales orgs over-study won deals; the tribal lore should weight lost deals more heavily. The closing exhortation — “make this the year of the customer; ship three great things” — frames product as the team best placed to rebuild the customer relationship after a hostile stretch. This lens connects to Jeff Weinstein‘s compounding-relationship philosophy at Stripe Atlas.

Lightning round [§ Lightning round]

Favourite show: The Terminal List (and Top Gun: Maverick, watched three times “from a product perspective” — high quality, no wasted calories, Cruise delaying release years rather than ship it early).
Favourite interview question: the 14-day Colombo question (above).
Favourite tools: Opus Clip (AI video clipping — lets one person do what a clip team would, at minimum-viable quality); a OnePlus folding phone (replaces ~50% of his iPad use).
Life motto: “Be kind.” His Twitter username carries it as a reminder. Be kind especially when things do not work out — and remember every failed hire is the founder’s failure.
On reaching people: the most important people read email; a genuinely good, specific, no-ask-for-coffee pitch gets opened and sometimes answered.