Last week, blogger Aidan Finn posted an excerpt from a Microsoft email announcing 13 percent price hikes in the European region. Reaction from the trade press was immediate, predicting that this action presaged ongoing price hikes.
Barb Darrow of Fortune, in her piece titled "Surprise! The cost of cloud is about to rise" said "Companies spoiled by round after round of price cuts on cloud computing services over the past year or so had better gird themselves for change." David Linthicum, SVP at Cloud Technology Partners, in his blog titled "The era of ever-cheaper cloud services is over" went on to say:
I suspect we'll see some more price increases in 2015 and 2016, for the simple fact that cloud providers are under pressure to make actual money, not simply gain market share. The margins are likely razor thin, so price hikes are the only way to increase cash flow now that the demand is on a path of steady growth. As cloud providers lock in customers, they are bound to raise prices, much like large enterprise software providers have done for enterprise licenses.
On the other hand, in a research piece, 451 analyst Owen Rogers, who focuses on cloud pricing and economics, proffered a different reason: exchange rates. His analysis says:
We believe this price increase is a 'blip' caused by currency fluctuations, and cloud services (particularly IaaS) will continue to decline in price overall over the next few years. We anticipate many cloud providers will absorb currency fluctuations in their pricing, such that gross margins on some services are temporarily reduced but a competitive price is maintained.
[Related: Amazon opens up about AWS revenues]
And Finn himself put forward currency exchange rates as the reason for the price change:
Microsoft announced, by email, tonight that pricing for Azure in the Euro Zone will increase by 13 percent. This is not surprising; The Euro has tanked thanks to Greece over the last 6 months.
I agree with both Rogers and Finn that the root cause for the Microsoft price increase is the change over time in the Euro/Dollar exchange rate, as seen in this chart from oanda.com:
As can be readily seen, the Euro has dropped precipitously over the past year, making it likely that Microsoft's Azure offering has run into stiff headwinds in terms of its cloud revenues and how they translate into its US results, which probably accounts for its pricing change.
A sign of things to come?
The question is, of course, does Microsoft's action foreshadow a general reversal of the historic downward trend of cloud prices across the industry? Should current and potential users reexamine their assumptions regarding likely costs and the associated application deployment decisions? Moreover, the crucial question, as noted by Darrow, is whether Microsoft's move will be matched by other providers, especially AWS, the dominant cloud service provider.
From my perspective, Microsoft's action is not the initial move in a shift toward higher cloud prices, for the following reasons:
- This is an emerging market, with users still making initial choices in what promises to be long-term vendor choices. Raising prices in a nascent market is likely to put off potential customers, especially ones conditioned to ever-lower prices as a core feature of the market. So Microsoft's move is unlikely to foreshadow a general trend.
- Amazon is unlikely to increase its prices. Amazon takes a long view regarding its business, and uses ever-lower prices as a core part of its strategy. Raising prices in response to disadvantageous exchange rates does not align with that strategy, so Amazon probably won't be hiking its prices. In any case, as I discussed when I analyzed AWS financials after Amazon broke them out in April, AWS has high enough margins that it can absorb a currency exchange hit.
- To say that cloud prices have reached an irreducible minimum and that CSPs will now, inevitably, turn toward raising prices, is not accurate. First, assuming Moore's Law continues at least into the mid-term (35 years), the input costs for basic cloud computing will continue to drop. While some providers might think this provides the opportunity increased margins or, for those who are suffering financially, to raise prices sufficiently to move toward profitability, there are some CSPs who eschew this approach and reduce prices to align with internal cost reductions (e.g., AWS and Google). David Mytton, a frequent commentator on cloud pricing, noted in his blog that core cloud services like compute, storage and network are commodity offerings and tend to be highly competitive and track closely to cost inputs -- and therefore it's unlikely that Microsoft's move presages a general trend. David's post is highly perspicacious and should be read on this topic.
Finn notes in his piece that Microsoft's move is quizzical. To a certain extent, the exchange rate change shouldn't have much of an effect, according to Finn, given that Azure receives revenues and incurs costs in Euros within Europe. So the two should cancel one another out, with only the margin remitted to Microsoft in the U.S. subject to exchange rate issues.
I'm not so sure about that, given that a large part of the revenue is tied to capital investment, which incurred the cost in the past at the higher Euro exchange rate; today's revenue has to be applied against those costs, which, when the revenue is worth much less means the greater part of the revenue needs to have the exchange rate applied bottom line, the exchange rate change affects more than just the difference between cost and European revenue.
Timing is everything?
However, I think there's something much odder about Microsoft's decision to increase prices. Raising prices, which is likely to direct potential customers to other providers, is, on the face of it, a poor short-term decision. Since the increase positions Microsoft more higher-priced than the current market leader, it only makes sense if you expected Amazon to follow suit and raise its prices, which, as noted above, is unlikely.
The decision to raise prices is doubly odd given that this is a nascent market. Customers making choices today are likely to stick with them for years, so surrendering the lifetime value of many customers to avoid a relatively small loss today is perplexing.
It's also perplexing because Microsoft has worked its guts out to achieve its position as second-largest CSP and the only member of Gartner's IaaS leading quadrant other than AWS. Raising prices is likely to cede revenue to AWS, which is Azure's primary competition. Traditionally, one of the ways a less-established market entrant competes with the dominant market player is to offer lower prices. This move, which positions Azure as more costly than AWS, flies in the face of that approach.
Finally, I find the decision to raise Azure prices puzzling because it forfeits one of Microsoft's undoubted advantages vis-a-vis Amazon: deeper pockets. Microsoft has virtually unlimited capital resources compared to Amazon, so why would it not take advantage of them? In this scenario, far from raising prices, Microsoft would lower Azure prices to make Amazon absorb lower margins, which would hurt the dominant company in the industry more than Microsoft. That approach is clearly what Google has pursued, and it's surprising that Microsoft hasn't executed the same playbook.
If I had to guess, I would speculate that the Azure team undertook this price hike as part of the overall Microsoft price increases; packaged software (which also saw price increases) makes up the vast preponderance of Microsoft revenues and, while under no particular competitive pressures, is also not growing very rapidly. So, for the purposes of earnings growth, it is important for Microsoft to adjust European prices, and the Azure group needed to increase its prices in parallel.
It will be interesting to see how this price increase plays out. But for sure I think it's safe to say that this move by Microsoft should not be viewed as "the end of the beginning" for cloud computing, and the turning point after which cloud prices go up, not down.