Two Engines, One Price: Madhavan Ramanujam on Monetizing AI Products from Day One

Madhavan Ramanujam, formerly managing partner at Simon-Kucher and now running his own early-stage AI fund, says his new book Scaling Innovation is a sequel to Monetizing Innovation, written eight years earlier. The first book was about building products people will pay for. Entrepreneurs kept coming back with the next question: how do you turn a great product into a great business that grows fast and profitably?

His core thesis is that an enduring business must master two engines at once, market share and wallet share. That means being good at acquisition, monetization and retention together: winning customers, earning money on them up front, and continuing to earn while they refer others. Many companies pick a single-engine strategy instead, growing at all costs, monetizing early but missing acquisition, or serving a small loyal base without doing either well.

Ramanujam is careful to say this does not mean equal effort at every point in time. It means equal attention and a conscious view of the trade-offs. His analogy is an aircraft: you would not choose to fly it on one engine, so why run a company that way?

Ramanujam maps six common traps onto three founder archetypes. The disruptor risks landing without being able to expand, having given the farm away, and also confuses market share won with market share held by neglecting retention and upsell. These are the cost of being too eager to acquire.

The money maker can nickel-and-dime customers with exotic pricing levers and hidden fees. They can also fall into the price premium paradox, where pricing high is mistaken for signaling value but ends up choking acquisition. The community builder, by contrast, focuses so much on the loyal foundation that they miss the frontier of new customer types, and trains the best customers to expect more for less.

The profitable growth architect, he says, is a disruptor, a money maker and a community builder at the same time, using strategies that avoid each archetype's blind spots.

One of the startup-phase strategies he highlights is beautifully simple pricing. His test is practical: ask early prospects or customers to explain your pricing back to you as if they were selling it on your behalf. If they cannot, your pricing is not simple, and the friction will show up in sales conversations.

Simple pricing must also tell a value story. He credits Superhuman, which competed with free email, for pricing at $30 a month and framing it as about a dollar a day to get four hours back each week. He also cites Subway's $5 footlong as a value story that is not limited to premium products.

The book includes a ten-point checklist for judging whether pricing is simple enough. Later in the conversation, Ramanujam calls Intercom's Fin, which charges 99 cents per AI resolution, "two chapters in one": beautifully simple and outcome-based at once.

In the scaleup phase, Ramanujam says, mastering negotiation is among the most important strategies, especially in B2B where humans still negotiate with humans. It rests on three things: gives and gets, value selling, and negotiation strategy. Every concession should be traded for something, because a concession with nothing in return tells the other side they can keep pushing. One of his favorite B2B gets is a recurring value audit in which the customer's own team assesses the product's value, creating their business case and giving you pricing power at renegotiation.

Value selling has three parts. First, create needs rather than just discover them, for example by getting a prospect to say that three weeks of dashboard prep could be instant. Second, build affirmation loops by checking regularly that the customer sees the value being shown. Third, build an ROI model with the customer from the start, because no one believes a model you cook up after a pilot.

A good ROI model has three buckets: incremental gains on the KPIs the customer tracks, hard cost savings such as headcount or license costs, and the often-overlooked opportunity cost, meaning what a team does with the hours you free up. If the customer agrees on the inputs, they are unlikely to dispute the output. That is why he says a pilot should be framed as building a business case.

The third negotiation pillar is strategy, starting with options. If you present one product at one price, the conversation is only about price. With good, better and best at $100K, $200K and $300K, a budget-conscious buyer might say they like the $100K price but want the $200K features, and you can ask what about those features matters to them, which moves the discussion back to value.

Options can also be pricing models. Ramanujam describes a founder who believed the budget was around $100K but whose product might justify $500K. The coaching was to offer either $100K plus 10% of incremental value, or $500K fixed. Most buyers avoid outcome-based pricing, so the $500K becomes the premium for certainty, and in this case it was negotiated to $400K, four times the expected deal.

He adds two more tactics: anchoring high, since starting high means ending higher, and tapering concessions. Weak negotiators give 5%, then 10% more, then 15% more, signaling there is always another discount. Strong ones taper: fifteen, then five, then two, which tells the other side the negotiation is ending. Lenny notes this is a much less scary way to ask for a bigger number than simply doubling your ask.

Ramanujam says AI pricing differs from earlier SaaS for two reasons. There are real cost dynamics to manage, and there is far more value to capture. An agentic product that taps labor budgets, which he says are about ten times software budgets, will under-monetize if it uses old playbooks, and customers will be trained to expect more for less. AI also finally helps solve attribution: a manufacturer can say throughput rose 10% and scrap fell 5%, which gives vendors real pricing power.

The questions founders ask most are how to charge, which now matters more than how much, and how to run commercial discussions during early pilots. The model has shifted from paying for access to paying for work delivered. He tells founders to reframe a POC entirely: its goal is to create a business case, with integration and functionality fit as consequences, for example a 30-day pilot for co-creating an ROI model.

Should you charge for a POC? Yes, but smartly. A price tag filters tire kickers who would burn 30, 60 or 90 days and never buy, but you must make clear it is not the commercial price, or a $10K pilot anchors a $120K annual deal. If a buyer still demands a number, contextualize it on value, for example pricing at one-tenth of the ROI, or give a range such as $500K to $1M that the business case will narrow.

Asked about popular AI coding tools, Ramanujam says some have under-monetized and will run out of runway if their fast revenue growth is not enduring. If they trained customers to expect a $20-a-month price while delivering large value, they are in trouble, and the way out is usually higher-priced, more sophisticated products. His point is that it is a trade-off between more customers and making money, which is why both shares matter.

His 2x2 has attribution and autonomy as axes. Low on both, seat or subscription pricing fits. High attribution but humans in the loop, as with a coding assistant like Cursor, calls for a hybrid of seats plus consumption credits. High autonomy but low attribution, as with back-end infrastructure, suits usage-based pricing, where usage proxies value.

The golden quadrant is high autonomy and high attribution, which means outcome-based pricing. Examples are Intercom's Fin, charging only for AI resolutions handled without a human, and Chargeflow, taking up to 25% of recovered chargebacks. Ramanujam says about 5% of companies are truly outcome-based today, with hybrid the most popular, and he expects the 5% to reach 25% within three years. Classic SaaS captured 10% to 20% of value, he says, whereas outcome-based AI can capture 25% to 50%.

He cautions against rushing there: if you cannot prove attribution you will fail. Pick the archetype that fits today, then plan the path, which means building dashboards and value audits that track the customer's KPIs and building agentic workflows that take humans out of the loop. Coding may move from seats to hybrid and eventually toward hiring something like an AI developer, with outcome pricing.

As companies grow into multiple products, Ramanujam says packaging becomes the next big issue: platform plus add-ons, good-better-best versions, per-use-case products or fully modular offerings. On changing pricing, he says teams once revisited their pricing model every two years, and with AI that interval has probably halved. Keep the pricing model stable unless attribution or autonomy changes, but revisit price points regularly, since prices rise 3% to 5% a year for everything else.

He quotes Warren Buffett's view that pricing power defines a company: if you need a prayer session to raise prices 10%, you have a terrible business. His axiom for this is price paralysis: reluctance to raise prices is internal and emotional, not external and logical. The book has 42 branded axioms, a count he jokes is a Hitchhiker's Guide reference, and Sierra's founder Clay helped with the naming.

His other favorites are the 20/80 axiom, where 20% of what you build drives 80% of willingness to pay, yet is the easiest part to build and is often given away free. He suggests MVP should mean the most valuable product. He also cites stopping churn before it happens by acquiring customers who will not leave, and landing so that you leave room to expand by gating on features and usage.

Asked for the lesson founders most underestimate, Ramanujam returns to his thesis. Many say they think about wallet share but have actually postponed it, consciously or not. The contrarian point is that the effort need not be equal, since some stages call for market share dominance and others for wallet share, but the attention should be.

He says acquisition, monetization and retention are correlated, so no single lever is isolated. Land helps acquisition and expand helps wallet share, while a pricing model should let you acquire faster, recover value and be simple enough that customers understand it and stay. He urges readers to treat price before product, and a great business and a great product as two halves you need both of.

On his move into investing, he and co-GP Josh Bloom, who together worked with more than 500 companies at Simon-Kucher, now invest in 20 to 25 early-stage AI companies in their first fund. Instead of charging fees for service, they work on monetization with founders in exchange for a place on the cap table.