The issue of digital transformation is dominating the strategic planning landscape right now, and the quest to speed up the collection and analysis of data is accounting for a sizeable portion of these discussions.

For all organisations, the success or failure of their digital transformation journeys ties directly to the effectiveness of their IT service delivery environments.

However, it is clear that the scale and complexities of next-generation applications are placing considerable stress on existing datacenter facilities, IT architectures, network capacity, and operational practices.

And a significant challenge that must be addressed is the fact that internal datacenters are often in the wrong locations for critical new workloads. For example, when it comes to transformational workloads that are mobile and customer facing, organisations need datacenter assets near large telecommunications hubs and population centers, not within their corporate headquarters.

And for transformational workloads that are required to provide data and insight with low latency to a local set of users or devices, organisations need to deploy standard, low-overhead local clouds in micro-datacenters.

As such, enterprises need to develop datacenter diversification strategies that minimise the impact of the technical debt associated with ageing internal datacenters while enabling quicker responses to business changes.

Ultimately, agile development requires agile IT. And that, in turn, requires agile datacenters.

It is for this reason that forward-looking organisations must adopt a vision that is based on agile IT architectures and smarter datacenter facilities solutions that provide a sustainable foundation for new service innovation.

To this end, IT departments must transform themselves to become the primary service providers and facilitators of secure, cost-effective, and resilient resources for new digital business services.

This transformation includes changing procurement, budgeting, and funding practices to effectively leverage the growing range of pay-as-you-go/use options for datacenter facilities, IT resources, and advanced cloud-based data/application services.

It also includes changing IT provisioning and management practices to leverage a wide range of datacenter and cloud services that will contribute to all digital transformation initiatives.

Most datacenters built before 2005 were optimised to support organisations’ systems of record, focusing primarily on managing internal business operations and sharing information with employees.

Following the start of the mobile explosion in 2010 and the growing focus on customer-delivered content, organisations began to spend more of their money on systems of engagement that created new user experiences and systems of insight (i.e., big data analytics).

This led to a greater focus on client-facing and analytical applications rather than back-office systems such as enterprise resource planning (ERP), where the primary focus remains consolidation, optimisation, and operational efficiency.

But in the next three years, enterprises need to prepare for another major shift in the workloads that will drive major datacenter expansions, with artificial intelligence, machine learning, and augmented reality all set to gain in prominence.

Many of these next-generation workloads will emphasise the delivery of context to further enhance the customer experience and will include a growing range of cognitive and machine-learning solutions that support rapidly evolving services based around the internet of things.

In financial services, initial implementations of blockchain technologies will also begin to affect datacenter planning decisions related to upgrading systems of record in the next three years, while most other industries will need to begin assessing their impact by 2020.

Unlike the “mobile engagement” wave, which almost exclusively shifted datacenter expansion to larger, centralised service-provider datacentres, this next wave of workload expansions will have a more widespread impact.

Some workloads, like deep learning, will drive a greater need for access to elastic compute pools. And rather than directly deal with the capacity planning issues posed by these apps, many organisations will “rent” even more IT resources from cloud service providers, transferring the task of building and running elastic datacenters to a shrinking group of industry experts.

However, to gain the maximum return from many cognitive workloads, enterprises will also require greater placement of compute and data resources in corporate headquarters and at edge locations, a development that may come as a shock to those organisations that planned to shut down all or most of their own datacenters.

While many businesses will conclude that increasingly complex performance and latency demands make existing facilities unattractive, the success of their digital transformation journeys will rely on them increasing investment in new, modular micro-datacentres that better meet performance and operational efficiency targets.

The columnist is group vice-president and regional managing director for the Middle East, Africa and Turkey at global ICT market intelligence and advisory firm International Data Corporation (IDC). He can be contacted via Twitter @JyotiIDC. Content for this week’s feature leverages global, regional, and local research studies undertaken by IDC.