Are we doomed to repeat the mistakes of the past or can we learn from them and prioritise integration as we embrace cloud-based IT? The majority of IT departments have grown up struggling with and around self-contained silos of information with every new application.

These silos have typically been assigned their own server (most probably a virtual one these days) plus their own data store – often a self-contained database on yet another SQL server. Integration with other applications and data stores may be little more than an afterthought. Worse still, responsibility for this important task may be devolved out to line of business managers ill-equipped to understand the wider implications of getting integration wrong or motivated to do anything about it.

A much-repeated pattern, the end result of these silos is an ever growing collection of isolated information strung together by disparate attempts at integration which may, in some cases, further add to the problem. Living with all these silos is hard and keeping the information within them current and consistent is an ongoing battle with major implications when it comes to accuracy and data centre efficiency.

Unfortunately, we are seeing this same practice being carried over into the cloud with a distinct danger that companies end up creating yet more silos of information. Silos that will be even harder to maintain, to find, manage and integrate with others than those they struggle with already.

It’s particularly worrying when it comes to public cloud applications that are finding their way into the enterprise. Easy access to public cloud applications provide line-of-business managers with a quick and easy solution to all manner of ills – and with no need to involve IT.

Again, lip service may be paid to integration by the service providers, but that will mostly be centred on pulling information from other apps rather than working with them. Moreover, in most cases, hardly any thought or guidance will be available as to how customers should bridge the physical divide between the cloud and the corporate datacentre.

Fortunately, and contrary to what many might claim, we aren’t genetically disposed to repeat the mistakes of the past. We can learn from them and leave the silo generation behind, simply by understanding the dangers upfront and putting integration at the top of the agenda whenever a new development is planned. It needs to be made clear to everyone involved, both what their responsibilities are and what the implications of ignoring this part of the application equation can mean longer term.

Those commissioning new applications, for example, need to understand that they can’t just conjure up a new application in isolation. Any new application needs to work with and share information with those run by other business units and they should understand that there are positive benefits to be gained from doing so.

Equally IT departments need to understand the need for expert guidance and look at the integration tools they make available to ensure that they are both accessible and easy to use. More than that, those tools need to have an adaptive and fluid approach to integration to be able to bridge the cloud divide and facilitate the integration of all new developments, to avoid silos, whether applications are run on-premise or in the cloud.