Reuters

Amazon's strategic advantage isn't really its inventory, its recommendation engine, or its consumer technology. The advantage is in its infrastructure.

In the global race to build a fast-as-possible service for delivering the things you order online, no one in the world is further along than Amazon, and it will be very hard for even a deep-pocketed competitor like Walmart to catch up as long as Bezos has permission from investors to spend every dollar his company earns on new fulfillment centers.

Amazon Prime, the subscription service launched in 2005 at $79 that makes most two-day shipping free, is a clever way to monetize this infrastructure advantage. Amazon can afford to sell its most popular stuff near cost if tens of millions of its customers are paying $79 on top of it, no matter what. Prime has served a couple other purposes, as well. For consumers, it's an entertainment bundle, giving us free access to Kindle Owners’ Lending Library, Amazon Video, and (allegedly) a forthcoming digital radio service. For investors, Prime represents a key lever for generating profits in the future. Many of the analysts I spoke to for my business column last year on Amazon said they didn't think it could raise prices dramatically on most of its merchandise. Instead, they said Amazon could always raise the price of Prime on its most passionate customers and add hundreds of millions of dollars to its bottom line just like that.

Last week, that's precisely what Amazon did, announcing a $20 price hike that will bring the annual cost of Prime to $99. Is that a garish raise? On the one hand, it's a 25 percent hike. On the other hand, $99 is just $5 more than the inflation-adjusted price of Amazon Prime when it debuted for $79 in 2005. For regular Amazon shoppers, buying Amazon Prime is still obviously worth it.

Prime is just one part of Amazon's bundle strategy. Last year the company continued to roll out its Prime Fresh service, which, for a $299 annual fee, sends fresh food orders to some city's doorsteps in a matter of hours. 

The power of memberships isn't just that they represent dependable revenue for Amazon in the topsy-turvy world of retail. It's also that they're sticky for customers. Couch potatoes have a hard enough time canceling their $90-a-month gym memberships, thanks to status quo bias and general laziness. It's even harder to justify canceling a $8.25-a-month membership that gets you free fast shipping to the biggest online store, a great digital video offering, and more, just because the price went up by less than $2 a month.

But if millions of the people do abandon Amazon Prime over a $20 price hike, that would be the worst news of all. It would show that the razor-thin-margin company, having successfully bought the trust of Wall Street, had failed to build a similarly strong allegiance with its customers, millions of whom would have left the service over an extra $1.85 per month. How would investors continue to justify a quasi-infinite price-earnings ratio for a company whose best customers refuse to pay another $0.40 per week?

In short, I expect the Prime price hike to work. Amazon better hope so, too.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.