David Howard, Unisys’ director of IT automation, recently explored the adoption of cloud services in the public sector with Informa Telecoms & Media’s Sheridan Nye. Sheridan offered her thoughts on the subject:
The G-Cloud is an ambitious initiative to create a new regime in public sector procurement. The key to accelerate adoption in 2013 is to attract enough participating suppliers to the CloudStore. Services must be practical, easy to find and also innovative where required. Over time, diversity of supply should mean a greater range of services are being offered.
In practical terms, many buyers are budget-constrained right now. Others are not at the right point of the buying cycle and some are waiting for evidence from early adopters, or adopting cloud-based services via traditional tendering. So the value of CloudStore transactions may take longer to grow than the Cabinet Office would have liked.
Security of sensitive data should always be the top priority, of course. The public cloud environment is clearly not suitable for all applications. Even Amazon Web Services and Google are not yet accepted as direct suppliers on the CloudStore, which shows the continuing concern about data center location.
Yet a surprising amount of data that the public sector collects has no material impact on anyone’s privacy or security. So the first task for buyers is to understand the implications of the type of data they collect. Following a data audit, they can allocate applications to public or private cloud environments and avoid incurring the cost of security where it’s not needed.
The ICT industry is pretty good at managing encryption and detecting malicious intrusions – IT teams across the country are doing this every day for their financial services customers. Over time, public cloud security and reliability will improve as they become important differentiators for suppliers. But the biggest threats are often quite mundane – lapses in routine activities like management of staff access, security patch upgrades and cutting corners in operational procedures.
Vendors therefore have an opportunity to provide services for identity management applications, compliance audits and so on. Using the CloudStore will reduce the need to update patches and to chase down outdated applications like IE 6.0 web browsers.
The fundamental aim of the G-Cloud is to shake up public sector procurement and perceived complacency at incumbent suppliers. That means cutting down multi-year integration projects into manageable, single-year and service-centric contracts that are more transparent. The invitation to SMEs is to ‘come and take on the big guys.’ To date, they’ve taken about 60% of CloudStore sales.
Implementing the G-Cloud has exposed all kinds of inconsistencies in traditional applications pricing. Big suppliers were probably happy to keep that out of sight. So it’s appropriate to encourage buyers to look beyond the usual suspects. On the other hand, service definitions on the CloudStore must be clear. They shouldn’t muddy the waters between a top-end, fully featured service from a major ICT company and a component of that service from an SME. It’s important that the major players are motivated to engage with the CloudStore as a conduit for large contracts.
Large suppliers still have many sweet spots in the public sector. For example, for specialist Cloud Services providers, there is a huge opportunity to assist customers in the migration of their legacy and mainframe environments to private and hybrid clouds. Also private cloud providers will increasingly position their offers less around “speeds and feeds” and more on pre-integration and features.
The benefits are widely accepted these days: lower and better-understood costs, the shift from CAPEX to OPEX, consolidated data centers, redistribution of power from sellers to buyers, greater transparency and liquidity in the market – the list goes on! ICT functions should enable public sector organisations to focus on their core activities, thus also resulting in reduced overheads.
There are however, certain risks to be considered, so in certain cases, it may be appropriate to migrate in stages. A certain amount of ‘toe dipping’ can be expected, although that’s not ideal for suppliers that prefer large scale to proof of concept projects.
To an extent, the telecom industry has already travelled this path. Mobile service companies have long outsourced their network operations to their equipment vendors, for example. Public sector organisations don’t exist to deliver ICT. If Cloud services enable them to deliver services at lower cost, then that direction seems to make sense.
In practice, few scenarios end up at their most extreme end-point. Arguably, the public sector may well be relying on more contractors than in-house ICT teams than it has in the past, but there won’t be any ‘big bang’ transition.