In one sentence
The best way for a startup to become powerful is not to extract more value from its current position, but to transform itself into a more central, valuable, and hard-to-displace part of the customer’s world. This is legitimate only when the transformation genuinely improves the customer’s experience.
Overview
Graham’s central office-hours question is: What would make this company more powerful? Asking how to make more money usually produces incremental improvements; asking how to gain structural power can reveal opportunities that increase value by orders of magnitude.
He treats power as a set of possible transformations: own the customer relationship instead of supplying a component; make money or tokens flow through the company; create an app-store-like ecosystem; introduce network effects; generalize a product into a marketplace; go full-stack; follow unexpected user behavior; help customers make money; acquire customers and data early; or escape slow, dominant intermediaries by approaching the market from the side.
The constraint is decisive: none of these strategies works merely because it benefits the startup. Customers must become better off. A startup’s initial weakness is therefore productive: it can become powerful only by making customers’ lives better.
Core ideas
Power is structural, not merely financial
Revenue optimization tends to yield small improvements. Structural power comes from becoming central to transactions, relationships, standards, ecosystems, data, or customer workflows.
Own the relationship and the flow
A component supplier is vulnerable when another company controls the customer. Startups should ask whether they can own the outside relationship or make money, payments, or other valuable tokens flow through them.
Turn products into platforms or marketplaces
An app store, API, shared standard, or marketplace lets others build value on top of the startup. Their efforts then increase the startup’s value. Even unexpected network effects—such as user comparisons, shared data, or agent-to-agent transactions—can transform a service into a platform.
Go full-stack when intermediaries hold you back
Instead of selling technology to companies in an industry, a startup may use its own technology to provide the end product directly. It can also gradually absorb the hardest work its customer performs, eventually reaching the customer’s customer.
Treat misuse as market information
Users who employ a product for an unintended purpose may be revealing a stronger demand than the intended use case. PayPal’s shift from handheld-device security to online payments illustrates how a peripheral feature can become the real business.
Help customers make money
Products that increase customers’ earnings are adopted quickly and can command substantial prices. This creates a double advantage: faster adoption and stronger willingness to pay.
Play the long game
Startups can exploit the short-term focus of competitors, executives, and quarterly planning. Early generosity, low prices, or favorable terms may be rational when they compound through user growth, retention, or customer trajectories.
Create more value than you capture
Trying to extract every possible dollar generally produces limited gains. Creating a much larger pool of value can produce far greater returns. Open source and extensibility illustrate how giving away control can establish standards, trust, distribution, and ecosystems.
Practical takeaways
- Ask in every strategy discussion: “What would make this company more powerful?” Then look beyond incremental revenue tactics.
- Map who owns the customer relationship, who controls the transaction, and where the money, data, or tokens flow.
- Look for a platform, marketplace, API, standard, or ecosystem opportunity—even if the original product does not obviously have network effects.
- Examine peripheral features and unintended uses. Unexpected behavior may identify the strongest product.
- Ask whether the company could sell to customers earlier in their life cycle, especially fast-growing startups that decide quickly and may remain customers as they expand.
- Try to acquire data upstream, before it becomes fragmented or expensive to obtain.
- If an industry is controlled by entrenched intermediaries, avoid frontal competition when possible. Become valuable in an adjacent dimension that makes them less relevant.
- Simplify fear-driven complexity. Graham’s recurring advice is often effectively: remove an unnecessary constraint and be more ambitious.
Caveats and counterpoints
- Power-building strategies can be dangerous when they are pursued for the company’s benefit rather than the customer’s. Network effects, lock-in, and control are not inherently valuable if they worsen the product.
- Selling cheaply can accelerate growth, but pricing below value may destroy useful customer signals; growth then stops revealing whether the product is genuinely wanted.
- Going full-stack requires capabilities, capital, and operational discipline that a component startup may not possess. It can also create direct competition with former customers.
- Selling to early-stage startups works best when churn is low and the product naturally expands with the customer. Otherwise, small customers may remain too small to support the business.
- Open source, APIs, and extensibility trade control for adoption and experimentation. They can enlarge the ecosystem but may also enable competitors or make monetization harder.
- The essay is a founder-oriented heuristic, not a universal operating plan. It emphasizes strategic possibility more than execution costs, regulation, market power, or organizational limits.
Questions worth revisiting
- Where is the startup currently dependent on another company’s customer relationship, distribution, standard, or approval?
- What could users build on top of the product if access were easier or more open?
- Are users already using the product in ways the company did not intend?
- Could the product help customers earn money, save substantially more than it costs, or grow faster?
- What would the customer’s ideal world look like, and could the startup become one of its central components?
- Is the company targeting customers too late—after they become slow, bureaucratic, and expensive to sell to?
- Which apparently peripheral feature could become the main business?
- What ambitious action is the company avoiding because of fear rather than because of a real constraint?
Return to this when…
Return to this essay when a startup is optimizing conversion rates, pricing, or sales efficiency but feels strategically boxed in. Its most useful purpose is idea generation: identify ways to move closer to the customer, control a more important layer, compound through other users’ efforts, or escape dependence on entrenched intermediaries—without losing sight of whether customers are genuinely better off.