But it is the ambiguity of this “cloud deployment model,” offered by SAP in its announcement, that is troubling to some industry watchers and cloud purists. As pointed out by ZDNet’s Dennis Howlett, there was no official mention of “multitenant software” in the press announcement. True multitenancy is an important distinction to SaaS purists and persnickety analysts, though apparently less important to customers and certainly of no care to end users who just want to get their jobs done.
Tom is referring to a bunch of ambiguities I raised in the discussion around SAP BusinessOne being repurposed for subscription based pricing. He is quite right to say that end users don’t care. They just want to do ‘stuff.’ However, multi-tenancy matters a great deal to providers and for good reason. Ultimately, it also matters to buyers. I’ll explain a few good reasons.
Be aware though there are plenty of debates around this topic and SaaS/cloud vendors often use it as one of the weapons they wield as part of the sales process. The problem comes when you start asking about flavours of multi-tenancy. Then there are the naysayers who often get it hopelessly wrong.
One of the principle reasons that Naomi Bloom insists on true SaaS (multi-tenancy) is because it has a direct impact on operational cost. She cites many other examples. Many of the vendor cases I see talk about direct operational cost running at 5% of revenue. That is a low number. One vendor I know has it down to 2% and wants to get to 1%. Think about it – you’re generating £10 million in revenue. The difference between 1-5% is £400,000. That’s a sizeable chunk of change. One of the principle reasons SAP re-architected SAP Business ByDesign (BYD) was because their ‘mega-tenant’ architecture could not operate efficiently and at a cost point with which SAP could live. From everything I have been told, BusinessOne is not fundamentally different from earlier iterations of BYD. The difference is that the cost problem has been schlepped over into the hands of third party operators.
Another good reason for multi-tenancy is that upgrades become a thing of the past. Everyone is (more or less) on the same version of the software. It is rare to see customers get more than one release out of step. It is much easier to maintain one set of software than umpteen variations. Anyone who remembers the days when we had to support dozens of UNIX versions knows what I mean.
Then there is the question of add on modules or complementary applications. A smart vendor will use multi-tenancy as the basis for creating a platform from which developers can build new applications and modules and to which third party stand alone systems can be easily attached. Even in the SME world we see a lot of that happening. Check out any of the Xero, FreeAgent, Freshbooks or Kashflow sites to see what I mean. The platform in turn provides third party developers the chance to reach much larger markets than are possible with on-premise systems. In the on-premise world, we see a LOT of duplicated effort which disappears in the multi-tenant world.
There are many other reasons (see Naomi’s post) but I’ve just highlighted a few. So – do customers care? Sure they do. Why? Because they get access to a much richer set of applications than is otherwise possible and often at highly competitive prices. Anyone who thinks otherwise is not living in a transformative world. They’re living in a world of older, tried and failing models.
‘Persnickety analysts’ will continue to make those calls. Why? They impact the likelihood that one vendor will be around longer than another. Apart from what I have stated above, the economics alone help ensure that. Another reason? Those of us who have direct influence on the buying decisions will point customers to the issues because they are important in the context of transformational change. That, above everything is where the real value of these systems lays.