LLenny's Podcast
← All episodes
Madhavan Ramanujam27 July 2025

Pricing your AI product: Lessons from 400+ companies and 50 unicorns

7Frameworks
12Insights

Frameworks in this episode

Marketing2 steps

Beautifully Simple Pricing

In your early days, price so a customer can repeat it back and it tells a value story.

Read →
Sales6 steps

Mastering B2B Negotiations Through Value

Extract full value from every deal via gives-and-gets, value selling, and disciplined negotiation tactics.

Read →
Strategy2 steps

Stop Churn Before It Happens

To stop churn, acquire customers who won't leave — churn is an acquisition problem, not a save problem.

Read →
Innovation3 steps

The 20/80 Willingness-to-Pay Axiom

20% of what you build drives 80% of willingness to pay — and it's usually the easiest 20% to build.

Read →
Finance4 steps

The Attribution-Autonomy Pricing Matrix

Pick your AI pricing model by plotting value attribution against product autonomy, then climb toward outcome-based.

Read →
Sales3 steps

The POC as Business Case

Frame every pilot as co-creating an ROI model, charge for it smartly, and never anchor the commercial deal.

Read →
Strategy4 steps

The Two-Engine Profitable Growth Architect

Master market share and wallet share together — equal attention, not equal effort — to avoid the single-engine trap.

Read →

Insights & moments

The myth-busts, hot takes, explainers, and tools worth keeping.

Myth Buster· 1

Myth Buster56:30

Stop Churn Before It Happens: Attract Customers Who Won't Leave

Madhavan's counterintuitive churn axiom: to stop churn you must attract customers who won't leave in the first place. Trying to stop churn once a customer says they want to go is too late and reactive, at best you buy another six months. Instead, study which customer types stay longest and focus acquisition dollars on getting more of them.

  • The best way to stop churn is to attract customers who won't leave
  • Fighting churn once a customer wants to leave is too late and reactive
  • Retention offers usually just delay departure by ~6 months
  • Analyze which customer characteristics correlate with staying longer
  • Redirect acquisition spend toward those low-churn customer types

“to stop churn, you need to attract customers who won't leave. That sounds counterintuitive, but that's the best way to actually stop churn.”

Madhavan Ramanujam · 56:30

“most companies would try to stop churn when someone actually says I want to go. It is too late and you're being reactive.”

Madhavan Ramanujam · 56:30
#churn#retention#acquisition#counterintuitive

Hot Take· 2

Hot Take37:00

The Popular IDE Startups Undermonetized at $20/Month

Asked whether the popular IDE startups are in trouble, Madhavan says some of them are, for sure. By anchoring at $20/month while delivering huge value, they trained customers to expect more for less and may face churn and non-enduring revenue. The fix, if any, is retrofitting higher-priced, more sophisticated products, but throwing out a cheap price purely to grab market share is a trap.

  • Some popular IDE/AI coding startups undermonetized and may run out of road
  • Fast revenue growth isn't the same as enduring, profitable, low-churn revenue
  • Anchoring on $20/month while delivering 10x productivity trains customers to expect more for less
  • Some companies try to undo it with higher-priced, more sophisticated products
  • Chasing market share with a cheap price and no expand strategy is the danger

“Some of them for sure have.”

Madhavan Ramanujam · 37:00

“If you just threw out a $20 product hoping to just, you know, accelerate your market share, you're in trouble.”

Madhavan Ramanujam · 38:30
#ai#pricing#hot-take#coding-tools
Hot Take55:00

The 20/80 Axiom: Redefine MVP as 'Most Valuable Product'

Madhavan's 20/80 axiom: 20% of what you build drives 80% of willingness to pay, and ironically that 20% is often the easiest thing to build. Founders build it, give it away almost for free, then chase their tails building the 80% that only drives 20% of willingness to pay. He argues MVP should mean 'most valuable product,' not 'minimum viable product.'

  • 20% of what you build drives 80% of willingness to pay
  • That high-value 20% is often the easiest thing to build
  • Founders give the valuable 20% away almost free and chase the low-value 80%
  • Truly understanding what drives willingness to pay is critical
  • Redefine MVP: 'most valuable product' rather than 'minimum viable product'

“20% of what you build drives 80% of the willingness to pay. But the irony is that that 20% is the easiest thing to build…”

Madhavan Ramanujam · 55:00

“It shouldn't be minimum viable product. It should be the most valuable product”

Madhavan Ramanujam · 55:30
#pricing#mvp#product#willingness-to-pay

Explainer· 1

Explainer27:30

Why AI Pricing Is Different: Monetize From Day One

Madhavan explains that AI founders must tackle monetization at the seed/pre-seed stage, unlike prior SaaS companies. Two forces drive this: real cost dynamics to navigate and, more importantly, value capture. Because AI products deliver so much value, under-pricing trains customers to expect more for less, and AI that taps labor budgets is charging against a far bigger pool than software budgets.

  • AI founders must handle monetization from day one, not defer it like old SaaS playbooks
  • Two reasons: for the first time there are cost dynamics to navigate, plus the value-capture problem
  • Agentic AI taps labor budgets, which are ~10x larger than software budgets
  • The model has shifted from paying for access (software) to paying for work delivered
  • AI can finally solve the attribution problem, unlocking real pricing power

“And if you don't capture that from day one, then you're training your customers to expect more for less.”

Madhavan Ramanujam · 28:00

“labor budgets are 10x compared to software budgets.”

Madhavan Ramanujam · 28:30
#ai#pricing#monetization#value-capture

Story· 3

Story13:00

How Superhuman Sold a $30/Month Email App

Madhavan uses Superhuman as an example of pricing that tells a value story. Competing against free email, Rahul's team set a simple $30/month price but framed it as a dollar a day to get back four hours of productivity a week. Reframed as the price of a weekly latte for four hours back, the price no longer looks high.

  • Superhuman charged $30/month while competing against free email products
  • They framed it as $1/day to get back four hours of productivity per week
  • The pitch: the price of a latte in a week to get four hours back
  • Value contextualization applies to budget products too (e.g. Subway $5 footlong)
  • A simple price paired with a value story removes friction

“you pay a dollar a day for actually getting four hours of productivity back in the week and then suddenly the pricing doesn't look too…”

Madhavan Ramanujam · 13:30

“it's like the price for a latte in a week to actually get four hours back.”

Madhavan Ramanujam · 13:30
#pricing#storytelling#value-story#saas
Story22:30

The $500K Third-Option Hack That 4x'd a Deal

A founder believed a customer's budget was ~$100K but felt his product could command $500K, and lacked the courage to ask. Madhavan coached him to present options: $100K plus 10% of incremental value, or a $500K fixed price. The conversation shifted to value, the $500K got negotiated to $400K, and the founder 4x'd the deal versus where he would have landed.

  • Founder feared asking for $500K when the perceived budget was $100K
  • Solution: offer $100K + 10% of incremental value, OR $500K fixed
  • Options move the conversation from price to how value is generated and measured
  • Buyers often pay the fixed premium to avoid outcome-based uncertainty
  • The $500K got negotiated to $400K, roughly 4x the original expectation

“go in with a 100K plus 10% on any incremental value that you bring or it's a 500k fixed.”

Madhavan Ramanujam · 23:00

“in this specific situation that 500k got negotiated to 400k and they just forex the deal compared to where they would be.”

Madhavan Ramanujam · 23:30
#negotiation#pricing#b2b#sales
Story42:30

Intercom's Finn: 99 Cents Per AI-Resolved Ticket

Madhavan cites Intercom's Finn as a clean example of outcome-based pricing. Finn charges based on an AI resolution: if the AI resolves a support ticket independently, they charge; if a human has to intervene, they don't. He calls it two chapters in one, beautifully simple pricing plus an outcome-based model, contrasting a $20 agent seat with a 99-cent-per-resolution charge.

  • Finn charges per AI resolution, only when the AI resolves the ticket with no human in the loop
  • If a human intervention is needed, they don't charge
  • It's a shift from seat-based (per support agent) to outcome-based pricing
  • Reportedly 99 cents per support ticket resolved via AI
  • It combines beautifully simple pricing with an outcome-based model

“So if an AI is able to you know resolve the ticket completely independently without a human in the loop then they charge for it.…”

Madhavan Ramanujam · 42:30

“That is uh two chapters in one. Beautifully simple pricing and an outcome based pricing model.”

Madhavan Ramanujam · 49:30
#ai#outcome-based-pricing#intercom#case-study

Tool· 2

Tool1:04:00

Products He Loves: Delphi (Lennybot) and Granola

Madhavan names two products he recently loves. Delphi, the digital mind representation behind the Lennybot, which he thinks is the future of how thought leadership is created and consumed, and plans to build a 'Delphi' of himself. Second is Granola for meeting notes, which his team has started using for capturing, organizing, and querying meeting content.

  • Delphi powers the Lennybot as a 'digital mind representation'
  • He sees AI thought-leadership clones as the future of consuming ideas
  • He plans to create a Delphi of himself, ideally before his book launches
  • Granola is his go-to for taking, organizing, and querying meeting notes
  • Granola has been the most-mentioned product on the podcast recently

“granola. We we love that product.”

Madhavan Ramanujam · 1:05:30
#tools#ai-products#delphi#granola
Tool1:01:30

Three Books Madhavan Recommends Most

In the lightning round, Madhavan shares the books he recommends most: Business Model Canvas by Alex Osterwalder, a strategic-business-model classic; Thinking, Fast and Slow, for the behavioral and customer-psychology angle that matters in both B2C and B2B; and Contagious by Jonah Berger, which distills how to make messages go viral into a usable framework he applies in his own outreach.

  • Business Model Canvas by Alex Osterwalder, ties strategy to business-model thinking
  • Thinking, Fast and Slow, understanding customer psychology in B2C and B2B
  • Pricing is as much behavioral as it is numbers, especially in human B2B negotiations
  • Contagious by Jonah Berger, a framework for making messages viral
  • He has applied Contagious's principles in his own outreach

“the first one that comes to mind is Business Model Canvas by Alex Ostraaler is a classic and one of my favorite books”

Madhavan Ramanujam · 1:01:30
#books#recommendations#psychology#marketing

Takeaway· 3

Takeaway31:30

Reframe the PoC as a Business Case, and Charge for It

Most founders frame a PoC as a proof of technical functionality; Madhavan says that's wrong. The entire goal of a PoC is to co-create a business case and ROI model with the customer. You should also charge for it, smartly, to filter tire kickers from serious buyers, while making clear the PoC price is not an anchor for the commercial deal.

  • The PoC's goal is to create a business case, not prove product functionality
  • Frame it as a 30-day pilot to co-create an ROI model with the customer
  • Charge for the PoC to isolate serious buyers from tire kickers
  • Make clear the PoC price is not a reflection of the actual commercial deal
  • If pushed for a number, contextualize on value or give a budgetary range, not a point price

“The P should be framed as the entire goal of the P is to create a business case. Period. Full stop.”

Madhavan Ramanujam · 32:00

“you start isolating people who are just tire kickers versus serious buyers.”

Madhavan Ramanujam · 33:00
#ai#poc#sales#roi#b2b
Takeaway43:30

AI Companies Can Capture 25-50% of Value vs 10-20% for SaaS

In classic SaaS, capturing 10-20% of the value delivered was considered great. Madhavan argues AI companies can capture 25-50% because the work is autonomous, done by the AI with no humans in the loop, and the value is attributable. Roughly 5% of companies are in a true outcome-based model today, and he believes that will reach 25% within three years.

  • Classic SaaS rule of thumb: capturing 10-20% of value delivered is great
  • AI can capture 25-50% because it's autonomous and attributable
  • The best outcome-based companies recover 25-50% of the value they create
  • ~5% of companies are in a true outcome-based pricing model today
  • Madhavan predicts that 5% will move to 25% within the next three years

“in the classic SAS situation we used to say if you can charge 10 to 20% of the value that's actually great but in AI…”

Madhavan Ramanujam · 43:30

“this is also my belief that in the next 3 years that 5% number will move to 25%.”

Madhavan Ramanujam · 44:00
#ai#outcome-based-pricing#value-capture#benchmarks
Takeaway52:30

Pricing Power: Your Reluctance to Raise Prices Is Emotional

Madhavan invokes Warren Buffett's line that a company is defined by its pricing power, and if you need a prayer session to do a 10% price increase, you have a terrible business. His price-paralysis axiom follows: reluctance to raise prices is usually internal and emotional, not external and logical. Founders should revisit price points and pass through 3-5% annual increases as a value exchange.

  • Buffett: a company is really defined by its pricing power
  • If a 10% price increase requires a prayer session, you have a terrible business
  • Price paralysis: reluctance to raise prices is internal and emotional, not logical
  • Prices for what you consume rise ~3-5% a year, so revisit your price points
  • Raise prices strategically, framed as a value exchange, to limit churn

“the true definition of a company is a pricing power and if you have a prayer session for doing a 10% price increase you have…”

Madhavan Ramanujam · 53:00

“your reluctance to do a price increase is often internal and emotional and it's not external and logical.”

Madhavan Ramanujam · 56:00
#pricing-power#price-increases#buffett#psychology