Pricing Your AI Product:
Lessons from 400+ Companies

Madhavan Ramanujam

Senior Partner, Simon-Kucher & Partners
Author, Monetizing Innovation

JUL 27 2025

Core Concept

Price is a Measure of Value

"Like liter is a measure of volume, price is a measure of value. Do people actually want your product and would they actually buy it?"

72% of innovations fail, because pricing was an afterthought
Product market fit is incomplete without product market pricing fit
Price before product, period, you don't choose if you have a pricing conversation, only when

Framework

5 Ways to Have the Willingness-to-Pay Conversation

1. Relative Framing

Index competitors at 100. "Where do we land vs. Salesforce on value? On price?" People are absolutely meaningless, relatively super smart.

2. Acceptable / Expensive / Prohibitive

Three-question sequence after pitching value. Acceptable = growth price. Expensive = value price. Prohibitive = laugh-you-out-the-room price. Look for cliffs in demand curves at psychological thresholds.

3. Purchase Probability (1–5 Scale)

Rating of 5 = only 30–50% chance they buy. Rating of 3 or below = never buys. Use at scale to build a demand curve and find the optimal price point.

4. Most & Least Questions

Show subsets of 6 features. Ask: "Most important (must-have, will pay)" and "Least important (don't need, won't pay)." Rotate subsets to rank all features. This reveals your top-20% that drives 80% of willingness to pay.

5. Trade-Off / Shopping Exercises

Put customers through real buying scenarios with different feature + price combos. Reveals mental models and price elasticity. Best for late-stage, pre-launch precision.

The Porsche Cayenne rule Every single feature was battle-tested with customers for willingness to pay before anything hit the factory floor. Result: Cayenne became >50% of Porsche's profit.

Deep Dive

Segmentation & Packaging Playbook

Segmentation: Productize, Don't Position

Same person = Charles III or Ozzy Osbourne, demographics mislead. Segment by needs, value, and WTP
One size fits none, every market has heterogeneous needs; find the segments and build for each
Start with one segment, nail it, then expand, don't launch five products at once
Dynamic segmentation: the same customer orders pizza Friday night and a healthy salad Tuesday lunch, understand when they switch

The water example Same water: fountain (free) → bottle ($2) → sparkling ($2.50) → minibar ($5). Same product, four packages, four segments, four prices. That's packaging done right.

Leaders, Fillers & Killers

Leader (50%+ want it)

Must-have feature. The Big Mac in the Happy Meal. Build the entire package around this.

Filler (nice-to-have)

Adds perceived value to the bundle. French fries + Coke. People wouldn't buy alone but accept bundled for marginal price increase.

Killer (10–20% want it)

Kills the bundle if included, coffee with a burger depreciates WTP for everyone. Sell as an add-on to the niche who wants it badly.

Tactics

How You Charge > How Much You Charge

Subscription wins when usage is predictable month-to-month, or value is ongoing but usage is episodic (LifeLock)
Usage-based wins when customers want low commitment, fairness, or variable bills track variable value (Michelin per-mile, Segment monthly tracked users)
Hybrid wins when you need both: HubSpot fixed base + overage. Incentivizes growth without bill shock
Value matrix: two-axis pricing (seats × departments) that rewards wall-to-wall adoption with better per-seat price, incentives drive PLG behaviors automatically

The break-even test Offer the same economics in 3 different model structures. Rational humans pick "indifferent." Real humans always pick one. Their choice tells you which model feels right.

Revisit every 6–12 months Or whenever you add a new plan, new features, or market dynamics shift. Pricing is not a one-time event.

Contrarian

Pricing Myths That Kill Products

✗ Build the product first, figure out pricing later INSTEAD → ✓ Price before product. You're going to have a pricing conversation with the market regardless, the only variable is when. Do it before you build, not after.

✗ Segment by demographics and personas INSTEAD → ✓ Charles III and Ozzy Osbourne share every demographic. Segment by needs, value, and willingness to pay, then productize to each segment, don't just position to them.

✗ Price low to drive growth, "penetration strategy" INSTEAD → ✓ Penetration only works if your entire cost structure supports it, like Amazon. Most startups that "price low to grow" just destroy margin and attract the wrong customers.

✗ More features in the entry plan = more conversions INSTEAD → ✓ Don't give the farm away at the entry level. If 60–70% land on your cheapest tier, you've misconfigured the packaging. Use the compromise effect: a decoy top tier makes the middle tier irresistible.

Based on Madhavan Ramanujam's episode on Lenny's Podcast. All ideas on this page are from the episode.Watch on YouTube Follow @madhavansf on X

GO DEEPER IN THE EPISODE

04:30 The core thesis of Scaling Innovation 09:20 Common traps founders fall into 12:06 Beautifully simple pricing 15:00 Mastering negotiations 26:51 Other strategies for effective pricing and monetization 27:35 How AI pricing is different 31:33 Handling POCs 36:25 The importance of mastering monetization 38:58 Choosing the right AI pricing model 43:13 Current trends in AI pricing 44:48 Strategizing for outcome-based models 50:23 Packaging strategies for scaling 51:37 Adapting pricing strategies over time 53:40 Key axioms for pricing success 58:00 Takeaways for founders 01:01:33 Lightning round and final thoughts

Every sketch here began as a transcript.

Send us yours and we’ll draw it. $1 a transcript.

Sketch mine →

0:00