Bigger – and better? How your IT infrastructure budget will change

I was reading in a report recently that the majority of firms intend to increase their overall investment in IT infrastructure over the next two years. Which got me wondering: will this be in-house infrastructure? And if not, where will the money go outside the organisation?

Bill will get richer

Office 365 – specifically the Exchange component, is already becoming the groupware platform of choice. Anyone who's run an on-premise Exchange installation will know, frankly, what a complete pain this is to do: unless you're a vast company you're only ever going to be scratching the surface of a hugely powerful platform, and the effort of supporting it is similarly disproportionate.

Use the Force

The other core exodus of funds that can only possibly go up is Salesforce.com. Ever looked at their revenue growth figures? In Q2 2015 revenues were up a mere 23 per cent; not so long before that they did four quarters of over 35 per cent increase in revenues.

Salesforce is one of those companies that you just can't help falling over wherever you turn, whether it's for the core CRM product or for the ever-increasing collection of other services like Service Cloud, Remedyforce ITSM or Radian6 social media management.

And the thing is, Salesforce isn't a cheap solution: if you want an inexpensive CRM package there are a thousand and one out there. Salesforce is one of those services for which companies seem keen to pay a premium because it does so much more than anything they could do in-house. And they're going to carry on pouring more and more money into it.

The internal kit

Regardless of where the organisation's applications live there's going to be some basic requirement for networking and infrastructure. Increasingly, desktops and laptops will give way to tablets and mobile handsets – which in many cases are more expensive than the traditional PCs they replace.

Cabled networks are, relatively speaking, inexpensive and largely going out of fashion. In their stead we're going to be spending more and more on WLAN products; the focus of the office infrastructure will be moving away from enabling computers to talk to servers and toward connecting users' endpoints to wherever the applications happen to live.

Application enablers

I've already mentioned doing your email in the cloud with Office 365, but as with the big lump of cash you're going to spend on your internal network for enablement of portable devices, another big wad will be going to cloud providers for enablement software and services.

Identity management is just starting to get big (not least because people are spotting services like Microsoft's Enterprise Mobility platform). Even organisations with no unified directory service are filling that gap by signing up with a cloud-based service so that they can gradually bolt on their services and make their disparate systems more closely linked over time.

The apps and the data

Companies are slow at upgrading applications. If you're using Office 2010 in your company this is nothing unusual, despite its successor having been released over two years ago – and this is only on the desktop.

At the server things are frequently slower. As of May 2015 some 11 million Windows Server 2003 installations remain, according to Microsoft, despite it going end-of-life in July 2015, and I'd be delighted to have a quid for every SQL Server 2005 licence that's still out there and probably will be when its support ends in April 2016.

Moving forward, the software rental model will grow both in terms of annual subscriptions à la Microsoft Enterprise Agreement but, more commonly, thanks to people simply buying Software as a Service (SaaS).

The rental model of SaaS is generally justified to the beancounters by telling them that instead of spending a boatload this year and then upgrading in three years' time, we're spending roughly a third of that boatload a year for evermore – an attractive cashflow option in many cases.

But the overall spend will be greater, of course, because in the past we simply haven't upgraded every three years on our on-premise service. Maybe we have on the desktop for mainstream applications, but for specialist apps it's common to be several versions behind.

And at the core the above SQL Server example is typical because the cost of the non-licence-related aspects of an upgrade (refreshing the kit it's sitting on, installing and testing the application itself, re-jigging stored procedures and automation packages, regression testing, and the like) are tangible.

In the cloud you have to do a lot less of this stuff - but instead of renewing every ten years, you're paying a net cost of two or three times that over time thanks to the ongoing annual fee.

Oh, and I mentioned data in the heading: right now we have drawers full of tapes holding our archive; in the future our archived files will sit on disk in the cloud, and because this capacity is infinite (unlike in our data centre, where we'd have to keep adding disk shelves or replacing disks with larger units) we'll keep adding more.

We'll pay more for this, because it's more reliable than having single copies of stuff on tapes – which don't have an infinite lifespan – on our premises which could flood or catch fire.

In short

So, then: in the future we're going to spend more money by moving our core apps to the cloud: it'll cost more but we'll get more in return and deem it justifiable.

We'll ramp up our spend on enabling technology and portable devices, which is far from cheap, and will upgrade frequently in a market where instead of power increasing slowly and prices coming down quickly, power increases quickly but with no cost reduction.

And we're going to rent our software and let someone else run the servers – and spend two or three times as much in some cases because deciding not to upgrade application X this year isn't an option. ®