On Dec. 1, 2023, Amazon.com (AMZN 3.43%) made a surprising announcement.

After refusing to patronize SpaceX rockets for more than a year -- awarding multibillion-dollar satellite launch contracts to essentially every space company in the market but SpaceX -- Amazon reversed course two weeks ago and awarded SpaceX three (of more than 90) contracts to launch Amazon Project Kuiper internet satellites atop SpaceX Falcon 9 rockets.

And that wasn't even the most surprising revelation from Amazon's press release.

After launching two test satellites for the Project Kuiper satellite internet system earlier this year, Amazon now says it is "preparing to start satellite manufacturing ahead of a full-scale deployment beginning in the first half of 2024." Furthermore, "we expect to have enough satellites deployed to begin early customer pilots in the second half of 2024."

A lesson from space history

Why was this surprising? Well, consider the example of SpaceX and its satellite internet constellation, Starlink.

Like Amazon's Project Kuiper, Starlink started with the launch of a pair of test satellites (dubbed Microsat 2a and Microsat 2b) in February 2018. SpaceX then spent the next 15 months testing these satellites in orbit -- testing their communication with ground stations and with each other, testing their upload and download speeds -- before deciding that all systems were "go," and it was time to begin mass production.

Its first launch of a full payload of operational Starlink satellites entered Earth's orbit in May 2019. SpaceX then continued launching satellites until, by October 2020, it had nearly 1,000 satellites in orbit. Only then did SpaceX decide it had reached critical mass and invite customers to sign up for a beta service with Starlink.

Add it all up: SpaceX needed 32 months to go from test satellites to beta service. SpaceX. A company with a reputation for moving fast and breaking things, whose unofficial middle name is "Rapid Unplanned Disassembly."

Amazon thinks differently

Logically, you'd probably expect it to take at least that long -- three or so years -- for Amazon.com, which was founded by Jeff Bezos, whose own space company, Blue Origin, has adopted the motto: "Gradatim Ferociter," or "step-by-step, ferociously."

But logic may not apply here.

Recall that Amazon launched its first test satellites for Project Kuiper in October 2023. Yet, now, Amazon says it will begin mass launching satellites within the next six months and begin "early customer pilots" -- beta testing -- in fewer than 12 months. (So, not 32 months but rather just 14 months total from testing to beta.)

That's an awfully fast timeline Amazon is marking out.

And yet, if Amazon is to be believed, it seems customers (and shareholders) can expect it to go from test satellites to generating beta service revenue in less than half the time it took SpaceX to do the same. That's either an incredibly optimistic timeline or means Amazon has spent the last two months testing its first pair of satellites and already concluded that everything is working so perfectly that it feels confident moving up its rollout.

Amazon seeks billions in internet profit

Confident or desperate or greedy -- or possibly some combination of all three.

It's worth pointing out that Amazon has good reason to rush its rollout. When it first received its license to build Project Kuiper, the Federal Communications Commission (FCC) conditioned its approval on Amazon's commitment to have half the constellation's 3,236 satellites in orbit by July 2026 -- so Amazon's working on a deadline here.

It's also worth pointing out that -- even for a company with a $1.5 trillion market cap, nearly $515 billion in revenue last year, and more than $12 billion in operating earnings -- Amazon has a big economic incentive to get Project Kuiper up and running. Again, from a standing start, SpaceX has built Starlink into a $3-billion-a-year business in just over five years. It's also doubling its Starlink revenues annually and targeting a 60% operating profit margin on those revenues.

The math on this isn't hard. $3 billion x 2 x 0.6 = Starlink potentially generating as much as $3.6 billion in operating profit in 2024, an amount equal to 30% of Amazon's own operating profit last year (according to data from S&P Global Market Intelligence). If Amazon can even come close to duplicating SpaceX's work with Starlink, satellite internet could become a major profit driver for Amazon.com in future years.

Whether it takes Amazon three years, or 32 months, or only 14 months to do this -- the race has officially begun.