I have always been impressed by the scientific rigor of the taxonomy that biologists have put in places for classifying living entities: kingdom, order, family, species etc... Just for fun, as I was stuck in a delayed flight back to Seattle, I tried to apply this taxonomy to SaaS. Here is the result...
This of course has not scientific value and must be taken more with a "tongue in cheek" spirit, but there are nevertheless 2 key points I want to make with this analogy:
Let me elaborate a little bit.
Many Species of SaaS
Although there is a common set of "DNA" shared by all SaaS applications, there are many possible DNA differences creating distinct branches in the SaaS tree. These different DNA differences can be technical (multitenant, isolated...), monetization attributes (ad funded, usage based...), target audience attributes (consumers, line of business...). Every change in DNA will result in a different specie. This is why there are no agreed upon definitions for SaaS, unless of course you take a more general definition, i.e. defining the class (not the specie). A legitimate question now could be: who is at the top of the food chain, i.e. is there a superior specie? I would reply that it depends on the habitat, which brings us to my second point.
Habitat dictates which species survive
It is very clear in Nature that species evolved based on the conditions of their habitat or more precisely, the one who were more in tune with their habitat survived: big furry animals in the poles, liquid-efficient lizards in the desert, simple micro organisms next to underwater volcanoes, you get the point... Similarly it is clear to me that SaaS applications need to be of many shapes or forms based on their habitats, hence the many continua we introduced in our papers.
At the SIIA ondand conference these last 2 days, I have been hearing a multitude of CEO/CIO/COO/VP of ... chest-banging like silver back gorillas (which btw are of the same Order than Homo Sapiens, so I suppose it is not too surprising :)) about how good they were at embracing multi tenancy, how they would degrade their offering (and valuation) by offering an hybrid model, how their incremental cost of an additional customer is close to zero etc... Although all good stuff, based on this analogy, they were describing a specific specie of SaaS. Which means that if the "habitat" they are selling into is not ready or friendly to that specie (e.g rejects data colocation, requires the ability to "migrate" from cloud to on-premise, needs identity federation) that specie of SaaS might not survive.
Don't get me worng, I agree that the "pure" SaaS DNA may be the most capital efficient (through maximun economy of scale, minimization of support and all the good things we describe in a previous whitepaper), and therefore the darling of the VC community, but please remember that the "pure" SaaS DNA specie will thrive only in specific habitats. The good news is that these habitats represent huge market opportunities and therefore one can safely commit to the "pure" play, but there is also life in many "less glamorous" (read less efficient) habitats... (just ask the cockroach :)
I suppose my point can be summarized with "one size cannot fit all". It is, with little doubt, a good strategy to aim for the most efficient versions of SaaS (multitenant etc...) but if you want to dominate multiple habitats (SMBs, large enterprises, consumers, ad funded, usage based...) you should better be prepare to mutate.