Can the Real Cloud Market Size Please Stand Up?

It seems like every week another sizing of the cloud market is published, and – maybe as to be expected – none of them seem to agree. Let’s have a look at who is saying what, and whether we are comparing apples to apples, or apples and oranges.

We will start by looking at SaaS. The most recent numbers from IDC claim that SaaS revenue will grow 5 times faster than traditional packaged software. This would mean little if traditional packaged software is expected to no longer grow (five times zero would still be zero). Joe McKendrick at ZDNet took IDC’s numbers and extrapolated from them that “very soon, a third of all software will be delivered via cloud.”

This seems to directly contradict Gartner numbers from just a month earlier. In June Gartner released a report stating that “Software as a service (SaaS) will have a role in the future of IT, but not the dominant future that was first thought.” In the same report, Gartner notes that some of the bad habits of the traditional software market – like massive shelf-ware as a result of the desire to get higher discounts by closing enterprise license agreements – have also entered the cloud market. In the press release about the report, Gartner predicts the SaaS market will reach $8.8 billion in 2010 while IDC talks about $13.1 billion SaaS revenue in 2009. I am sure there are some definition differences, but a $5 billion difference on something that they both call “WW SaaS revenue” seems statistically significant, or does it?

One explanation is that IDC specifically includes a large expected growth in SaaS management software, in addition to SaaS application software. The reason is the technology –specifically network technology is now ready for SaaS delivered functionality like remote monitoring and remote configuration management. This is significant because unlike you may expect, the market for “management software” is actually bigger than the market for “application software.” About 10 years ago, incidentally about the same time I moved jobs from ERP to system software, the cost of IT management surpassed the cost of IT applications. So there is a clearly a reason why everyone is so enthusiastic about cloud computing (running this stuff yourself is simply too expensive).

But back to the numbers (working for a software vendor you can understand why we have some interest in these). If SaaS revenue indeed exceeds traditional software within a few years, would that be relevant? Well, not really, because what makes up a SaaS fee is different from a traditional software fee. SaaS includes hosting (hardware), network, backup, new releases, support, change management etc. Software costs may actually be only 10 to 20% of the total. Ideally we should compare SaaS costs with the TCO (Total Cost of Ownership) of applications. This would be a good idea, if people actually knew the TCO of their applications. But that is another topic all together.

What a difference a day makes!
When we look at the total cloud market the differences were even more astounding. Within a day of each other Gartner estimated the Cloud market to be $148 billion by 2014. This basically contradicts IDC, which just two days earlier estimated $55 billion in 2014. John Treadway highlights one reason why the Gartner numbers are so high in in his post “Why Gartner’s Cloud Numbers Don’t Add Up (Again!)”. He points out that Gartner includes Google AdWords advertising revenue in their cloud market numbers.

So, at the high level, these numbers do not add up. But that is not such a big issue because if you take a close look at individual markets, you see huge differences in SaaS adoption. For example, most customers now use SaaS solutions for Project and Portfolio Management (PPM), while still relying on on-premise solutions for PC configuration, anti-virus and ERP. While PC configuration was mainly held back by technical limitations, ERP suffered more from the traditional attitude of typical ERP buyers. As usual, the technical limitations around the first are likely to be resolved faster, than the user culture issues around ERP.

In the end, trying to put a number on cloud adoption or the overall market opportunity for cloud can feel like we’re looking into a crystal ball. We can even say that the numbers are not of much help. Cloud is growing fast – on that point it seems industry analysts, vendors and end users all agree — but exactly how fast we will not know for sure until after it has happened. For many looking to rely on these numbers for strategic, operational or market planning, this may seem like an issue, but we should ask ourselves, is it really? Consider this: Two shoe salesmen flew into a third world country. After getting off the plane, both phoned their home offices. One said “I am coming back home, because nobody wears shoes here,” while the other salesman sent a telegram asking for more supplies and five colleagues to join him. He saw a huge market opportunity because nobody wears shoes there!). Same data, different conclusion.

It’s also wise to look at what we’re comparing. In the cloud market, it does not really matter whether cloud is bigger than traditional software (as we know it today) in 2014 or 2040. Instead, what does matter is that we’re focusing on investing more in functionality that does great things for the business rather than focusing on the gear that manages the plumbing.