How We Think About Innovation

The Two Directions of Invention

There are two natural paths to innovation:

Problem-first — You notice something that annoys you, and you fix it. The Wiffle ball (1953) and Liquid Paper (1956) both came from someone encountering a frustration and solving it directly.
Solution-first — A new capability enters the world, and you work backwards to find the problems it can now solve. Carbon dating is a classic example—people wanted to date ancient objects, but the method wasn’t possible until scientists understood radioactive decay. The capability came first; the application followed.

Persistence and the Forty Attempts

WD-40 stands for “Water Displacement, 40th Attempt”—taken directly from the lab notebook. A three-person company called the Rocket Chemical Corporation tried 40 formulations before finding one that worked. They eventually renamed themselves the WD-40 Company.

Persistence may be the most underappreciated ingredient of innovation.

Learned Helplessness: The Innovator’s Invisible Enemy

When you encounter a problem repeatedly and for long enough, you gradually stop noticing it. Great inventors resist this tendency—ordinary inconveniences continue to bother them.

Mary Anderson (c. 1913) — Drivers would pull over in the rain, wipe the windshield with a rag, drive another mile, and repeat. Everyone did it. Anderson said it was ridiculous and invented the windshield wiper. She was roundly criticized. Within ten years, wipers were standard on every car.
Joseph Gayetty (1857) — Invented toilet paper. Before that, people simply didn’t know what they were missing.

Gayetty’s original packaging: “Many people have wooed their own destruction, physical and mental, by neglecting to pay attention to ordinary matters.”

Reject Either/Or Thinking

Amazon tries to reject false tradeoffs—the idea that you must choose between two good things rather than find a way to achieve both.

Reduced customer contacts per unit by 85% over 7–8 years—not by cutting corners on service, but by eliminating the reasons customers needed to reach out, while building robust self-service tools.
Embassy Suites consistently wins customer service awards despite being almost entirely self-service. When friction is removed, the experience is rated as excellent.

Maximize the Rate of Experimentation

If running an experiment is expensive, only a handful can happen each year. Someone at the top has to decide which ideas are worth trying—and the most inventive people become frustrated.

Low-cost experiments let teams try things independently, without institutional permission at every turn.
“Customers Who Bought This Also Bought” went through dozens of experiments. Adding product images seemed obviously helpful—but data showed images actually made the feature perform worse.
The point: you have prior guesses; you validate them with real information.

Customer Obsession vs. Competitor Focus

The right question isn’t what will change in the next ten years? It’s what won’t?
For Amazon: customers will always want selection, low prices, and convenience. Anchoring strategy to stable customer needs means it doesn’t have to be rebuilt every time the competitive landscape shifts.
Small, autonomous teams with clear objectives and specific, actionable metrics—not distant outputs like revenue—do the unglamorous, grinding, highly technical work that compounds dramatically over time.

The Fixed-Cost Advantage

Much of Amazon’s best customer experience is a fixed cost. Rich product information costs roughly the same whether one million or 47 million customers read it.

Instant Order Update — Tells you if you’ve already purchased an item. In the short run, it measurably reduced sales. Amazon kept it anyway, betting that preventing accidental repurchases would build trust worth more than the marginal transaction it occasionally prevents.

Working Forward from What Changes

When something becomes dramatically cheaper, the right question isn’t how to pocket the savings—it’s what you can now do that was previously impossible.

Disk space costs ~30x less than five years prior. That made Search Inside the Book possible—storing full-page images of hundreds of thousands of books and letting customers search complete text before buying.

Hire Builders

If I had to identify the single most important factor in building an innovative organization, it would be selecting the right people.

Hire people who like to build and invent at every level of granularity, not just at the whiteboard.
Real invention is the accumulation of solutions to a thousand unglamorous problems—integrating a GPS with a digital camera, securing both to a vehicle, figuring out the vehicle needs to be tall enough that passersby don’t notice the camera.
Anyone willing to work only at the level of abstraction will be unable to make progress in the real world.

Staying Heads Down

In 1997, when Barnes & Noble launched their online store, Forrester Research coined “Amazon.toast.” Amazon had $60M revenue and 125 employees vs. B&N’s $3B and 30,000 employees.

Be afraid—not of competitors, but of customers. They’re the only ones who will ever give you money.
You can listen to outside voices and look for kernels of truth, but you cannot let them set your strategy.

Invention always leads you down paths that will look, to outside observers, a little strange. That’s not a warning. It’s almost a guarantee.

Core Takeaway

Innovation, at its core, has always been: refusing to accept that what is must be what remains.