Crunching the Number Crunchers

Last February The Gartner Group (a US firm of technology analysts) published a report showing that SaaS was more costly in the long run for large organisations. Now McKinsey (another firm of analysts) has published another report coming to similar conclusions.

Gartner’s logic makes more sense that McKinsey’s, and claims that despite initial higher costs for purchase and installation, after three years the cost would be cheaper than paying the monthly SaaS subscriptions, especially if you can spread those initial costs over a large number of users.

Well, Yes and No.

If you plug all the numbers you know about into a spreadsheet you may come to the same conclusion, but these mathematical models don’t cater for the real world, for example:-

Risk
The assumption is that you install a CRM system and everybody lives happily ever after. Well of course, if that CRM system is Really Simple Systems then this is not a fairy tale! However, as somebody we’ve all forgotten about once said “Stuff happens”. Stuff like the department being closed down, or sold, or downsized. Or a new technology coming along and making it all obsolete. Or a new CEO saying that he wants to use the same system he had in his last company. Or suddenly realising that this new CRM system doesn’t really do what you wanted it to do. Or it being so over designed and complicated that the sales people stop using it after six months. Or the vendor’s support falling off a cliff. Chuck a pre-paid for CRM system out after two years, let alone one, and that was an expensive mistake. Plus right now most companies’ cost of capital is a lot higher than bank base rates, and they are happy to pay a small premium for keeping that cash in house.

Unforeseen Costs
Then there are all the costs that never quite seem to be budgeted for. Like hours unscrambling laptops after the data sync failed. The operating system upgrade that revealed some bugs in the application, which had to be upgraded but the upgrade necessitated a database upgrade, which then caused another application to crash, occassionaly. Plus a maintenance price hike because the vendor is going through a hard time at the minute (pace, SAP & Microsoft!). Plus all those extra payroll costs that governments will be springing on us in two years’ time to pay for the current economic mess.

And finally, although no SaaS vendor seems to want to admit this, we can all bet that those monthly subscriptions will come down in price as the application area continues to be commoditised, and in three years’ time a larger enterprise really should be able to negotiate a price saving.

So even for large companies, I think SaaS works out cheaper, and it certainly does for small ones. The really good thing about monthly payments is that if you change your mind, you can leave. Once you’ve paid for a traditional install, there’s no way to get that money back when things change – and they will!

So according to Gartner & McKinsey, traditionally purchased software works out cheaper in the long run, providing the world stands still. In the long run, said Keynes, we’re all dead.

Thunder in the Clouds

I’m sure you’ve read about the IBM’s Open Cloud Manifesto, Microsoft’s reaction to it and then the press reaction to both. If you haven’t, see Paul Greenberg’s summary. Even the worthy Economist covered the story, so it must be big news.

“Open” initiatives are always launched by a consortium of second tier vendors who want to challenge the first tier vendors’ grip on the market. Unix v Open VMS, CORBA, the Open Software Foundation, we’ve all been here before. The market leaders will have their own “open” systems that are open as long as you use their products, whether their platform is Amazon’s, Google’s, SAP’s or Salesforce.com. And there is no way any of those vendors will want to make it easy for users to use products outside of the fold. Not surprisingly, the loudest noises are coming from Microsoft, the vendor with the most to lose from the whole concept of cloud computing – not just the application revenue, but the whole stack of client and server operating systems, middleware and databases.

When playing the “Open” card, vendors have to tread a fine line between making their systems sufficiently “open” so that customers are attracted to them, but not so “open” so that they can leave easily. Think lobster pot and you get the idea.

As always, the market will decide, despite different vendors’ attempts to set the agenda. Thank goodness!