Quarterback Tom Brady’s biggest weakness was exposed in the New England Patriots’ 20-18 loss to the Denver Broncos two weeks ago. While Brady is known for his clutch performance, he plays poorly when he gets hit frequently. It forces him to get rid of the ball more hastily and limits his ability to see the whole field.

You’ll find proof in his stats:

• For the year, he completed 64% of his passes, 36 for touchdowns against only seven interceptions.

• Against Denver in the 2015 season finale, he completed just 48% of his attempts with two interceptions and just one touchdown. The reason: the Denver defense hit him 23 times, 10 more than his previous season high.

Brady’s experience illustrates a common tension in business: speed vs. quality. Just as the faster Brady has to throw, the worse those throws tend to be, the faster a company can execute, the greater the likelihood that the product or service being rushed will suffer. The slower, more deliberate the execution, the higher the quality tends to be, though with the risk that a company may be late to market.

For most of the industrial age, this tension leaned toward quality. In the manufacturing of complex machines, even small failures can result in disaster. It’s why we don’t want Boeing to take shortcuts en route to delivering its airplanes.

In a post-industrial world, though, this line has shifted toward speed. Software and virtual services tend to be more forgiving of errors because they usually won’t kill or injure, and because fixing them doesn’t require retooling an assembly line. Deploy enough software engineers, designers, architects, and coders, and companies can fix these issues on the fly.

Releasing products and services while they’re still in their testing phase is called beta culture. (Beta, in the software world, describes a product development phase in which the application is feature complete and somewhat stable, but still retains errors, otherwise known as “bugs,” that require isolating and fixing.)

The tech companies that have oriented themselves to a beta culture have changed many aspects of their production, distribution and marketing. Software development teams run in agile mode with 2-4 week “sprints” to push out new features, as opposed to months-long cycles and more ambitious deliverables. Betas often are shipped to a limited number of users who are selected for their willingness to report bugs. Marketers may include the word “beta” in a logo to signal to the user that the software in question isn’t fully baked.

Chrome: just one example of Google’s beta products.

I mention all this because beta culture is exactly how investors should be thinking about Amazon, which has proven that the concept also has a place in redefining a business model.

You’ll find a few critical posts in this blog regarding Amazon. The company tends to be more opaque to outsiders. Its investments seem willy-nilly. If we look at its culture through a beta lens, though, it starts to make more sense.

Amazon competes in three massive markets: e-commerce/retail, IT infrastructure via its Amazon Web Services unit, or AWS for short, and entertainment via Amazon Prime. It does not enjoy the near monopolistic powers we associate with Google in search and Facebook in social media.

Amazon is well aware of the disruption it has imposed its markets and that the long arc of corporate evolution means that there is a very good chance that somewhere, someone is scheming The Next Big Thing to usurp it.

How does Amazon make sure it is well positioned to neutralize these threats? By applying beta culture to its business model and releasing what may at first seem like wacky products and services before its competition can. Consider the following:

• Amazon Dash (wifi-enabled buttons that allow consumers to buy individual products).

• Amazon Echo, a device that plays music, interacts with other connected devices, and includes a digital assistant.

• Amazon Fire phone, a smartphone optimized for shopping.

• Kindle e-readers

• Fire tablets

• Fire TV Stick, which turns any modern television into a hub for accessing Prime video.

• Delivery drones, still in development.

• Brick and mortar stores in the works.

Each of these innovations is designed to increase consumption of Amazon’s products and services by removing some friction from the process. None are revolutionary in the way that Amazon.com was when it first burst on the scene, but together they show a remarkable commitment to playing offense.

Put another way, can you imagine Walmart investing this way? I can’t, either.

Of course, this doesn’t come cheaply. From 2012-2014, Amazon’s spending on technology and content increased by an average of 48% per year, more than any other operating expense. Amazon notoriously runs at virtual breakeven as a result of plowing every dollar back into the business, which has investors wondering when Amazon will harvest the fruits of these occasionally crazy ideas.

The answer is impossible to predict unless you’re Jeff Bezos. Still, there is a method to Amazon’s madness. It starts with a hearty embrace of beta culture.