If you need proof that the federal government no longer drives the technology market, look no further than cloud computing.
Many federal market eyes are focused on the Pentagon’s forthcoming JEDI cloud computing procurement. It has a potential value of $1 billion per year for 10 years. Consider this: Microsoft, the number two commercial cloud provider behind Amazon Web Services, sold $10 billion worth of cloud services just last quarter.
In crafting its solicitation, the Defense Department didn’t want its cloud business to represent too large of a portion of the winning bidder’s business. There’s no danger of that. Analysts think Amazon tops Microsoft in pure cloud services revenue. It does around $8 billion per quarter in revenue, but it doesn’t have the Office 365 applications revenue that Microsoft counts as part of its cloud business. Regardless, when the Pentagon makes its award to one of the two companies — possibly as early as next month — the win will be in prestige more than in financial impact.
Not that prestige lacks a halo effect. No doubt many customers with high security requirements will think “Well, if it’s good enough for the Pentagon, it’s good enough for me.”
An irony in the two-way competition is that in many ways, Microsoft enabled Amazon in the first place. By producing one of the early technologies that enabled desktop computing, then with historic shrewdness making it the nearly universal operating system, Microsoft helped establish the PC as an appliance as ubiquitous as the telephone. From there it was a short hop to consumer utility of the Internet, on which Amazon changed the world in its own way.
Because of the so-called gate requirements, the Defense Department filtered out IBM and Oracle, two of the pioneers in enterprise, mainframe computing. Oracle, as Federal News Network has reported extensively, has mounted a vigorous protest of the solicitation, but has been rebuffed in court. It may not be able to challenge the award since, technically, it will not have been a bidder.
The JEDI deal has gotten more complicated in recent days, now that four Republican members of Congress called on the Trump administration to get on with it.
That’s okay, as long as they don’t even appear to try and influence source selection. House Armed Services Committee Ranking Member Mac Thornberry (R-Texas) understands this, I’d presume. Political people, in theory anyhow, shouldn’t drive acquisition decisions. It’s happened, to be sure, but usually it’s when things got out of hand, like the original, way-back Air Force tanker leasing plan that was famously grounded by then-Sen. John McCain. In the JEDI case the members felt the solicitation has dragged on long enough.
For his part, President Donald Trump said he was considering looking into the deal, after hearing complaints from other vendors. That’s not a good idea. He’s right in that some leading companies have complained about the way the Pentagon set up the deal. But the president’s known dislike of Amazon could inject politics into an acquisition. Even if JEDI is harebrained, that wouldn’t help it.
Among the complaints is the Pentagon’s insistence on a single award. You could argue the Pentagon ought to have a long term, multiple-award, big-cloud-tent vehicle. Given the non-guaranteed ceiling on JEDI and the robust cloud requirements across DoD, I predict JEDI’s DoD-wide relevance could fade anyhow. Here are a couple of reasons why:
Talk to Army research people and they’ll tell you how much they need “edge” computing. In a contested situation, who knows whether broadband connectivity back to a cloud would even be available?
Cloud services aren’t all equal. Each commercial cloud provider has hundreds of highly specific (to it) software services beyond bare infrastructure. No single provider is likely to meet all of the requirements of a department.
JEDI will happen, but it’ll be far from the last word in DoD cloud computing.