When people think about pricing strategy, they think about how to price. What's the right number, what's the right model, what does the competitive landscape allow. But what's equally — if not more — important is what not to price. The most consequential pricing decisions aren't about setting the right dollar amount. They're about choosing which products in your portfolio carry the margin, which ones you deliberately suppress or give away, and how the priced and non-priced products work together to build a compounding business.
Pricing is never a single-product question. It's a product portfolio strategy question. And the companies that understand this don't just set prices. They architect a system where free or cheap products acquire the asset they need, and priced products capture value where their differentiation is strongest. The interplay between the two is the strategy.
Every great pricing model has two layers: a wedge — the cheap or free product that acquires a critical asset like data, users, or transaction volume — and a capture point — the priced product where the company is genuinely defensible and switching costs are highest. If you can look at a company's portfolio and clearly identify which products are wedges and which are capture points, you're reading their strategy as plainly as if they'd handed you the board deck.
Data Infrastructure: Price the Storage, Not the Intelligence
Consider what Snowflake and ClickHouse are doing with observability. Both companies have pushed aggressively into log and event data, pricing storage remarkably cheaply. On the surface, this looks like margin compression — a race to the bottom against dedicated observability vendors. But that misreads the move entirely. They aren't trying to win on storage economics. They're trying to win on data gravity.
Every terabyte of observability data that lands in your warehouse is another terabyte that doesn't leave. It joins your application data, your business metrics, your customer records. And once it's there, the priced product kicks in: analytics. Cross-referencing logs against revenue. Correlating infrastructure events with user behavior. Running queries that span domains no standalone observability tool can touch. The storage product — barely priced — exists to feed the analytics product, where the real margin lives.
This is portfolio thinking in action. Storage and analytics are different products serving different purposes in the business model. Pricing them the same way, or thinking about them independently, would miss the entire point. One is the cost of acquiring data gravity. The other is the return on it.
AWS: The Non-Priced Entrance, the Priced Exit
Amazon Web Services understood something profound early on: the hardest thing to move is data at rest. So they built a portfolio where the products that bring data in are free or nearly free, and the products that do something with that data carry the margin.
S3 ingress is free. Storage pricing is low and has only gotten cheaper over time. Anyone can dump anything into S3 with minimal friction and minimal cost. These aren't loss leaders in the traditional sense. They are non-priced products in a portfolio — their job isn't to generate revenue. Their job is to create the conditions under which the priced products become indispensable.
And those priced products are everywhere. Compute, data transfer, egress, Lambda, Athena, SageMaker — every service that acts on the data sitting in S3 pulls against a gravitational center that the non-priced products created. AWS doesn't need to charge for ingress because ingress is doing something far more valuable than generating revenue. It's building the asset that makes the entire rest of the portfolio work.
The decision about what not to price — ingress, basic storage — was as strategically important as any price Amazon ever set.
Game Consoles: The Non-Priced Hardware, the Priced Ecosystem
Sony and Microsoft have spent decades selling hardware at or below cost. The PlayStation 5, at launch, was estimated to cost more to manufacture than its retail price. For any traditional hardware business, this would be ruinous. For a portfolio business, it's the entire point.
The console isn't really a priced product. It's a customer acquisition mechanism that happens to have a price tag on it. The real priced products — game licensing, first-party titles, PlayStation Plus, Game Pass — only work at scale with a massive install base. The hardware exists to create that base.
What makes this example instructive is how clearly the portfolio logic dictates the pricing. No one at Sony believes their advantage is in manufacturing boxes. If they did, the console would carry margin. Instead, the margin structure tells you exactly what they believe: their defensible advantage lives in the ecosystem. The exclusive titles, the network effects, the recurring subscriptions. The hardware is the non-priced wedge. The ecosystem is the capture point. The portfolio design makes the strategy legible.
Stripe: Non-Priced Simplicity, Priced Infrastructure
Stripe launched with pricing that looked almost confrontational in its simplicity: 2.9% plus 30 cents per transaction. In payments, where pricing had traditionally been opaque and negotiated, this felt like margin compression. Legacy processors saw it as a company leaving money on the table.
They were right about the margin compression. They were wrong about which product they were looking at.
Stripe's core payments processing was never meant to be the priced product in the portfolio. It was the wedge — simple enough to integrate in an afternoon, transparent enough to require no sales conversation, and just cheap enough that developers embedded it into their applications at the infrastructure level without a procurement process. It acquired the asset Stripe actually needed: deep integration into thousands of companies' codebases.
The priced products came next. Billing. Invoicing. Revenue recognition. Treasury. Identity verification. Fraud detection. Each one layered on top of the payment integration, each one carrying better margins, and each one deepening the dependency. Stripe's portfolio tells you their strategy: payments is the non-priced wedge (it generates revenue, but it's not where the margin ambition lives), and embedded financial infrastructure is the capture point.
The decision to keep payment pricing simple and low wasn't a failure to capture value. It was a deliberate portfolio choice to defer margin to the products where Stripe's differentiation is hardest to replicate.
Google and Android: The Entirely Non-Priced Product
Google's decision to give away Android is perhaps the purest expression of portfolio pricing logic. There is no wedge pricing because there is no price at all. The entire operating system — a product that costs billions to develop and maintain — is a non-priced product in Google's portfolio.
This wasn't altruism or even a competitive response to iOS. It was a portfolio decision. The priced products in Google's business — search ads, display ads, YouTube ads, Maps monetization — need the largest possible surface area of user attention. Every Android phone is a Google search bar in someone's pocket. The more manufacturers who adopt Android, the more distribution the priced products get, at zero marginal cost to Google.
But Android's non-price does something beyond just acquiring users for Google. It eliminates the ability of anyone else to build a margin-bearing business at the OS layer. By choosing not to price Android, Google didn't just serve its own portfolio — it reshaped the competitive landscape so that the OS layer couldn't sustain margin for anyone. The decision about what not to price was simultaneously an acquisition strategy and a competitive weapon.
What the Portfolio Reveals
Across all these examples, the relationship between priced and non-priced products reveals three bets each company is making:
Where they believe durable differentiation lives. The priced product is a declaration: "This is where we are hard to replace." Snowflake is betting that cross-domain analytics is harder to replicate than storage. Stripe is betting that embedded financial infrastructure is stickier than payment processing. If they're wrong about this, the margin erodes. If they're right, it compounds.
What they're willing to commoditize — including in their own portfolio. The non-priced or underpriced product is a deliberate sacrifice. Google commoditized its own operating system. AWS commoditized its own storage. Sony commoditized its own hardware. In each case, the company looked at a product it could have charged more for and decided that the portfolio was better served by keeping it cheap. That's a harder decision than most companies realize, because it means actively refusing revenue on one product to create the conditions for margin on another.
How switching costs accumulate across the portfolio, not within a single product. The non-priced product creates the initial adoption. The priced product creates the dependency. And the switching cost lives in the relationship between the two — the data that landed in cheap storage and now powers expensive analytics, the payment integration that brought in the billing suite, the console that anchors a library of digital purchases. No single product in the portfolio generates lock-in on its own. The portfolio architecture does.
The Implication
If you're building a company and you're thinking about pricing as a product-level decision — "what should we charge for this thing we built?" — you're asking the wrong question. The right question is: which products in our portfolio should carry the margin, which ones should we deliberately undervalue or give away, and how do the non-priced products create the conditions for the priced products to compound?
The companies that get this right don't just set prices. They design portfolios where priced and non-priced products reinforce each other — where the free thing acquires the asset, and the paid thing captures the value, and the gap between the two is where the strategy lives.