Despite the growth of new infrastructure trends like cloud computing, colocation and hyperscale data centers (DCs), the reality is that a vast majority of organizations still rely on in-house DCs for their core operational needs. It would therefore not be fair to assume that third party data center management will completely replace in-house DCs in the near future. At the same time, the same companies are also beginning to realize that scalability and business agility are critical success factors in today’s digitally connected world.
By holding on to their in-house DCs, businesses potentially end up limiting access to a rich set of services that can lead to greater business agility, scale business processes easily and build an IT ecosystem that can drive business innovation. In many cases, apprehensions around security and control have kept CIOs away from using third-party resources (DCs or public clouds), especially when it comes to safeguarding business critical applications and protecting personal data. For example, the financial services industry needs to store and process large amounts of customer transactions. While this may be a great opportunity to drive efficiencies by moving huge financial datasets to hyperscale DCs, the perceived security risk does not allow companies to implement this kind of migration, and they continue to rely on in-house DCs.
This implies that the most critical parts of the business do not get any of the strategic business benefits from dynamic, hyperscale environments (clouds and hyperscale DCs). Ironically, peripheral or non mission-critical systems like HRMS, small departmental applications and web portals are the first to migrate outside the firewall.
Recent analyst viewpoints seem to be very much in line with this logic. According to analysts like Gartner, the traditional in-house DC model is undergoing dramatic change, and we should expect a strong decline in the use of traditional DCs over the next 5-7 years. According to Gartner, by 2025, 0nly 20% of enterprises would have traditional DCs, as compared to 90% in 2018. (So, 70% of companies would have moved completely away from data center management).
A key factor that pushes the move to outsourced DCs is the strong backing by the technology community globally towards cloud-based applications and enterprise-class data center storage solutions. With cloud becoming a large market today, we’re seeing greater and faster technology innovation within the cloud ecosystem compared to the on-premise software landscape. Every major enterprise software vendor (SAP, Microsoft, Oracle, Salesforce, etc.) has made significant investments in building cloud development communities. Gaps in security, user experience and performance are being addressed at a much faster pace within the cloud development communities, with the expectation that they would be on par with on-premise systems in the future. This would strengthen the case to move out of in-house DCs and into completely outsourced infrastructure for even enterprise applications.
The other important question is – who would drive the transition within the industry? Would large enterprises be the first to consider switching off their on-premise DCs, given the opportunity to significant savings (since their data volumes and compute needs are also very large)? Our guess is that the change would be driven initially by the mid-sized companies, rather than the large enterprises. This is because mid-sized companies already use a higher percentage of cloud and external DCs for their IT needs, but still tend to hang on to their in-house DCs for critical systems. With smaller IT budgets, such organizations actually have greater motivation to migrate completely to third-party infrastructure. The CAPEX savings from using cloud services and hyperscale DCs allow them to invest in ‘enterprise-class’ security (endpoint, network, data and application), governance tools and robust data center storage solutions. However, in the absence of pre-defined practices and benchmarks, mid-size company CIOs will need to make sound trade offs between the potential security risk from sending critical data outside the firewall and being able to invest in stronger tools and platforms in order to minimize such risks.
The in house DC model therefore doesn’t seem like a strategically viable model in the near term. Building and maintaining their own DC is likely to be very low on the CIO’s list of priorities going into the future. CIOs will need to reassess their technology strategy and move away from in-house data center management if they want to stay competitive in a highly connected and hyperscale future.