IDC predicts the software-as-a-service (SaaS) market is set to grow more than 25 per cent annually for the next four years, coming to represent over a third of new business software purchases. But they also predicts that this rise of on-demand models will contribute to a $7bn (£4.5bn) decline in global software licensing revenue in 2010. Will this kill the VAR? Not necessarily.
Commercial Off The Shelf (COTS) or pre-configured/shrink-wrapped software implies a solution to a problem that will work out-of-the-box. It won't. It will always need to be configured. The only question is how far it can be configured by the user and how far it needs an expert integrator to do so. The degree to which it needs expert configuration determines the amount of professional services an integrator can expect.
Think what software is: It is the automation of common tasks by a series of related macros. But the 'tasks' that are automated in most desk-top software are just most commonly undertaken (the lowest common denominator). In Enterprise software, these tasks are automated to reflect best practice. Software therefore doesn't automate the way you currently do your stock control. It automates either the way most people do it or the way they should do it. In the former case you will have to configure the software to do it your way. In the latter you are going to have to change your processes to match the best practice the software implies you should adopt.
So how does the integrator fit in this?
In the client-server world, their main job is getting the system to work at all and keeping it up and safe. Then you have to get it to play nicely with the other systems. Configuring it to meet the client's needs, let alone changing business processes to take advantage of the built-in best practice never seems to get a look-in.
With hosted aps, a lot of the integration work is taken away by plug-ins. Certainly the system sizing, availability and maintenance is. This, then, frees the SI up for more value-added consultancy services, doesn't it? Only if the SI is able to deliver these and the client is willing to pay for them.
This is certainly the opinion of a recent CRN OpEd: Channel not ready to sell cloud "Consultancy is, I think, the key word." it concludes
In a recent survey of over 600 IT decision-makers across Europe (mostly large enterprises), we discovered that they were twice as likely to use consultancy services as not. And this was consistent across all countries, company sizes and industries (except IT companies who think they know it all already). Consultants are used roughly 60:40 for business (requirement definition & RFP etc) and technical (implementation) roles. Mostly this is to plug skills gaps, but also to bring in best practice and an external perspective. And they're not just used for scarey new technologies, but, most often for basic network services, for instance.
So there is a definite market for value-added consultancy services that most SIs and VARs can deliver. In fact they already do, based on other research we have done, but do not necessarily charge market rates for it, preferring to give away the expertise to gain the trusted advisor status that will win the product sale.
Already there is little margin in most COTS software, to the point where it is common for partners to tell the customer to buy direct, but do the implementation for them. This, of course screws up the sales out data and partner program incentive schemes, but more on that another time. The question is whether the SI can charge their customer for what was not, previously, an accurately costed consultancy service.
If they can't, then, yes, the cloud will kill the SI and the world will be littered with botched DIY cloud implementations and even less faith in the ROI on IT investments than already exists.