Strategy is a branch of pricing — Jean-Manuel Izaret on the inversion
Why pricing is upstream of strategy, why capturing all the value is the wrong goal, and how the cost/value/competition triangle actually gets used.
Jean-Manuel Izaret — JMI to BCG colleagues — started his career on a chemical plant where improving the yield by 0.5% after 30 years of improvement counted as a breakthrough. He moved to consulting because he wanted to work on growth rather than cost-cutting, and built BCG's pricing practice over the next 25 years across luxury, industrials, SaaS, and now AI. The thesis he opens with — *strategy is just a branch of pricing* — sounds like a partner-of-pricing's joke at the partner-of-strategy's expense. It's not. It's the spine of his book and the editorial frame of this conversation.
The pricing episodes on SoF have built a triptych. Per Sjöfors made the economic case (1% price up = 11.3% profit lift). Tim Smith made the structural case (pricing as a quantitative discipline between commodity and one-off). JMI brings the consulting-partner view that the other two pointed toward: pricing as upstream of strategy itself, with Salesforce and Uber as the worked examples and the cost-value-competition triangle as the practitioner’s frame. The discipline that separates a strategist from an extractor is the 50-60% value-share self-limit.
The inversion — pricing as strategy
JMI builds the claim through two cases the audience has lived through.
Strategy is just a branch of pricing. That in order to have a good strategy, you need to first have good pricing.
Salesforce didn’t invent CRM software — Siebel had been doing it for a decade. What Salesforce did was reprice it: per-user, per-month, with a minimum commitment, instead of a multi-million-dollar perpetual licence paid up-front. The entire SaaS industry that followed is downstream of that pricing-unit decision. Uber didn’t invent the taxi; it repriced the ride. In both cases, the strategy was the pricing model, and asking “what’s our strategy?” was the same question as “what are we charging for, and in what units?” JMI’s broader read: pricing and strategy share three inputs — your economics, customer value, competitor moves — and companies treating them as separate functions make worse decisions on both. The line is operationally useful: if your pricing unit is wrong, the strategy on top of it is shaped by gravity, not choice.
Capture 60%, leave 40%
The framework that separates a strategist from an extractor sits in the value-share number.
If you capture all of the value, usually you make your customers indifferent between you and the others.
The robot example is the cleanest illustration. You sell a robot for $10K plus $1K of incremental value to the customer. Price it at $11K and the customer is mathematically indifferent — same total cost of ownership either way. Price it at $10,500 and the customer pockets $500 of upside, which is the economic reason to choose you again next quarter. JMI’s empirical claim from decades of pricing projects: the optimum value-share even for genuinely unique products sits between 50% and 60%. Above that and you’re an extractor; the relationship doesn’t compound. Pharma is his negative example — trying to capture maximum value in every transaction has produced a goodwill deficit harder to repair than the foregone margin ever was. The 60% rule is the discipline that prevents the deal-by-deal optimum from destroying the customer-by-customer compound. It’s also why most companies overestimate elasticity and underprice — optimising for share, not for the compounding relationship.
The cost-value-competition triangle
The practitioner’s working frame for every pricing decision.
I should price based on my own economics. I should price based on what the customers value. And I should price based on competitors and what they are charging.
Three inputs; most companies anchor on one and ignore the other two. JMI’s frame is that the right weighting shifts with market structure. In commoditised markets (oil, securities, bulk chemicals), prices converge to a single market-clearing number and the economics-based anchor dominates. In differentiated markets (software, luxury, branded consumer goods), value-based pricing matters because the comparison set is messy and willingness-to-pay varies. In oligopolies with three to five players (PC OEMs procuring hard drives via reverse auction, JMI’s worked case), game-theoretic dynamics dominate — and counter-intuitively the price is set by the number-three player, because numbers one and two need to leave them space or watch the number-three player price-war the market down. Most companies that get pricing wrong have built their organisation around one of the three anchors and lost the triangulation that good pricing requires.
The AI-pricing signal
The forward-looking call that lands with finance executives is mechanical.
When AI's models are priced per user, they are here to help the user. They're not here to replace the user. When they are priced per task or per outcome, that's when they are here to go for our jobs.
The Copilot-per-seat pricing model is the labs telling you AI is positioned as a productivity assistant — fewer users means less revenue, so per-user pricing is self-defeating for any vendor genuinely betting on human replacement. The day the pricing model shifts to per-task or per-outcome is the day the labs believe they can capture the economic value of the output rather than a productivity tax on the human. JMI’s broader thesis — the pricing unit is the strategy — gets its cleanest 2026 test case here. Watch what the labs charge for, not what they say. The Microsoft-vs-Intuit story that closes the episode is the dramatised version: when Microsoft made Office Accounting free, Brad Smith at Intuit reframed the question from “how do we match the price” to “who would buy an accounting software for free” — and the answer (early-stage startups) led Intuit to make Simple Starts free for that segment. Microsoft exited the category three years later.
What to listen for
The full episode is the most credentialed pricing conversation in the SoF back-catalogue: JMI on cultural fairness varying by geography, dynamic pricing as fine when expected (airlines) and terrible when surprising (Uber surge), inflation as a 15-year catch-up, and Brad Smith reframing Microsoft Money out of the category. His three-word descriptor is Intellectual. Surprises. Change. Listen at /podcast/ep-030-jean-manuel-izaret; for the other two pricing conversations, see Per Sjöfors and Tim Smith, or /topics/pricing.
Related questions
- What does JMI mean by 'strategy is a branch of pricing'?
- The textbook chain runs strategy → product → pricing, with pricing as the last decision in the line. JMI inverts it. Salesforce.com built its entire business on a new pricing model — per-user, per-month, with annual commitments — not a new product. Uber did the same thing: same physical service as a taxi, repriced per-ride with a different unit. In both cases, pricing wasn't downstream of the strategy; pricing was the strategy. The reason this matters operationally is that the inputs to good pricing — your own economics, what customers value, what competitors charge — are also the inputs to good strategy. JMI's claim is that if you've answered the pricing question well, you've answered the strategy question. The joke is delivered with a straight face but earns the laugh: at BCG they sell strategy, and the senior pricing partner says strategy is just a branch of his discipline.
- Why does JMI argue that capturing all the value is usually the wrong goal?
- Pricing theory often says you should price up to the customer's willingness-to-pay — capture all the surplus. JMI's empirical view is that the optimum value-share, even for genuinely unique products, sits between 50% and 60%. The mechanic: if you capture 100% of the value, your customer is mathematically indifferent between you and the next-best alternative — they have no economic reason to choose you again. If you leave 40-50% of the value with the customer, you've given them a reason to come back, you've built trust that compounds, and you've protected the relationship from a competitor who hasn't been quite as greedy. The pharma industry is JMI's negative example: trying to extract maximum value in every transaction has produced a structural goodwill deficit with patients and regulators that's harder to repair than the foregone margin ever was.
- What is the cost / value / competition triangle?
- JMI's working frame for every pricing decision. Price based on (a) your own economics — your costs, your margin requirements, your scale curve — because in commoditised markets ignoring this gets you killed. Price based on (b) what the customer values — the comparison they're making, the alternative they'd buy, the upside the product unlocks for them — because in differentiated markets ignoring this leaves money on the table. Price based on (c) what your competitors are charging — because in oligopolies with few players, signalling and game-theoretic dynamics dominate raw demand elasticity. Most companies anchor on one of the three and ignore the other two. The pricing partner's view is that good pricing triangulates among all three, and that the weight you put on each shifts with the market structure you're operating in. JMI's book Game Changer codifies this as a set of distinct 'games' a business can be playing at any time.
- What is JMI's AI-pricing observation, and why does it matter?
- JMI says watch the pricing model, not the marketing. When AI products are priced per user — like Microsoft Copilot — the labs are positioning AI as a productivity assistant. If they were trying to replace users, per-user pricing would be self-defeating: fewer users means less revenue. When AI products shift to per-task or per-outcome pricing, that's the signal that the labs believe they can replace the human in the loop and capture the full economic value of the output rather than a productivity tax on the human. Today, almost everything is per-user or per-token. The structural shift to outcome-based AI pricing will be the visible market signal that the underlying technology has crossed from assistive to replacement. It's a much cleaner read than vendor positioning, and it lines up with the broader 'strategy is pricing' thesis — the pricing unit is the strategy.
Updates
- Editorial pass under the v2 podcast-summary guideline.