Suddenly it was there in my mailbox: “How do you feel about Software Testing as a Service, a.k.a. STaaS and could you blog about that?”. A request for a blogpost, hmmm, that’s different. Especially since I haven’t had much to do with the subject (yet), other than reading the different articles and offerings on Cloud, SaaS and the sporadic services that are offered as ‘hosted testing software’. But I like a challenge and I’m also interested in the topic so I set out on my information-gathering quest, there’s no such thing in my world as giving an opinion without knowing the background of the subject under discussion.
Sure I got a feeling, but is that feeling realistic and based on founded grounds or is it just based on (false) assumptions?
First of all: is STaaS any different than SaaS? Well, simply said: NO. Bottomline is that SaaS is, for those of you who didn’t already know, software that is offered as an online service. So if looked at it black-and-white; STaaS is nothing more than offering software that can enable software testing activities offered as an online service.
When I look at it this way, I say: “Good initiative”. Organizations that don’t have testing as a core business or don’t have the infrastructure (or recources) to host that kind of software, this would be a great solution. Think about the possibilities: imagine a big library-like site where you can choose-and-click the software testing software you need and then make use of it as long as you like without having the pains of administration.
Alas, STaaS as term, is used for more types of ‘products’ or ‘services’. So with the explanation above, I won’t be giving a complete answer. There are at least two more perspectives.
The first one is the infrastructure for testing ones software itself (the space/ hardware), for which I find the STaaS term isn’t appropriate. This should be more like ‘Serviced Test Environment’ or something like that. You can think of ‘virtual environments’ and so on. For example: when more testruns are done where older versions have to be ‘frozen’, this is ideal, because you can temporarily upscale your infrastructural capacity without having to buy the whole hardware package yourself. When not needed anymore the virtuals are just ‘deleted’ and you don’t have the hardware still in your server room and thus no maintenance and administration costs.
The second one is the service of Software Testing itself (activity). I find, that there is nothing new here and certainly not worth naming ‘STaaS’. For one, the meaning of aaS is that of product-like nature and not activity-like nature. Secondly; when one hires people for doing their testwork one can call it ‘Consultancy’ or if one hires people who test their software off-site, than it’s called, among others, ‘outsourcing’, ‘offshoring’ or even ‘testfactory’.
I do – however- think there’s still some development in these kinds of offerings. I think there are still services to be offered that will fit much more to the business needs than just ‘hiring people’ or ‘letting the product be tested by a company’. All of these shouldn’t be named ‘STaaS’ though.
So, my conclusion (now), is that STaaS as term isn’t necessary at all, it’s redundant and confusing. SaaS will do just fine. The ideas behind the different explanations of STaaS are however worth developing or investigating, but should be named in a way it’s clear what is meant by it. I think organizations will benefit from products which will enable high qualitative software activities without having to be burdened by the administrative and organizational hassle. I think this kind of products, services and offerings are still scarcely out of the egg and need some serious development, but it has an important role in the way software testing will be done in the future.