This week Microsoft, Alphabet and Amazon all report earnings, giving us the first clue as to whether the cloud is truly recession-proof.
Recent CIO surveys suggest that enterprises are reluctant to cut costs on security and cloud in the face of a recession, but sometimes there’s a gap between survey responses and budget realities. As the Wall Street Journal suggested: “Even the cloud can get rained on.” And so far, each of the cloud bellwethers is reporting lots of rain, as it were.
Even so, we shouldn’t get caught up in the initial results, as it’s to be expected that even “safe” spending like cloud and security will take a near-term hit as enterprises, grappling with a difficult macro economy, hit pause. The real question is whether that pause will last. For things like cloud computing, the answer is almost certainly “no.”
But first, the bad news about cloud earnings
Alphabet and Microsoft both reported lower-than-expected cloud earnings this week. For Alphabet, Google Cloud revenues came in at $6.28 billion, growing 37% year over year, down from 44% growth in Q1. Note that Google Cloud, along with Microsoft, isn’t easily compared with AWS revenue, as each of the three companies includes very different things in their respective “cloud” categories.
Microsoft grew its Azure revenue by 40%, down from 46% the quarter before. Amazon reports results on Thursday, July 28.
SEE: AWS Lambda, a serverless computing framework: A cheat sheet (free PDF) (TechRepublic)
Some of that apparent growth deceleration actually isn’t. As cloud revenues get bigger, the growth rate necessarily goes down, even as overall revenues and demand go up. But let’s assume that a healthy chunk of the lower growth is due to lower demand.
That’s OK: It’s normal. Given the macro environment, it would be bizarre if we didn’t see at least a brief pause as enterprises take stock of the macro situation and determine how they want to achieve their micro-goals like digital transformation. On the Alphabet earnings call, Alphabet CFO Ruth Porat said as much, citing “uncertainty in the global economic environment” as a reason for their earnings drop.
And yet, as I write this shortly after the markets close, pretty much every cloud company’s stock is up. Why? Because things weren’t worse, and because the future is bright.
And now, the good news about cloud earnings
In a Credit Suisse research note on Microsoft Azure, available to its investor clients, the equity research firm spells out the prospects for growth:
Following the Global Financial Crisis, the focus of change was then at the applications layer — with the adoption of SaaS. We believe (1) the focus of change now being on the infrastructure and platform layers of the stack, combined with (2) delayed delivery to enterprises of on-premises hardware due supply chain disruptions and (3) potentially tightening IT budgets both accelerating “renting over buying,” will result in (A) greater near-term demand resiliency for IaaS and PaaS during economic uncertainty — similar to the resiliency of penetration-driven SaaS growth during the European Debt Crisis directly on the heels of the Global Financial Crisis — and (B) sustained, robust growth for the rest of the decade.
Otherwise put, whatever the short-term stumbles of earnings for the cloud providers, the longer-term prospects are very bright. Yes, customers may skrimp in the short-term, but one key way they’ll do that is through reserved instances to lock in favorable pricing. Indeed, the cloud was made precisely for a moment like this, when uncertainty favors pay-as-you-go, elastic consumption and pricing models. Even companies that one might not otherwise associate with cloud, like SAP, have shown strong cloud growth (up 34% and now the company’s largest revenue stream).
If the pandemic put digital transformation into mach one, a global recession just might kick it into mach 10. Not because the cloud is some sexy, cool way to do business, but precisely because it has become the most reliable way to make IT investments in uncertain times.
Disclosure: I work for MongoDB, but the views expressed herein are mine.