There’s quite a bit of confusion about what is what in cloud computing. This is in part due to the absence of a standard set of definitions for the three types of offerings provided in the cloud, namely infrastructure, platform, and software as a service. But I also think there’s confusion about what the “cloud” actually is. As usual with any term that is elevated to “the next big thing” in the market, everyone jumps into the fray, either to claim credit for the wagon or at least jump on the wagon before it leaves the station. This is the moment when the marketing machine is often seen to take over to redefine and repurpose everything to make it part of this new momentum. We’ve seen this trend before when suddenly all enterprise software became SOA and all hardware turned “green.” Now everything runs “in the cloud”. So as not to get caught in the undertow but to catch and ride the actual wave to your advantage, one ought to know the difference between types of currents and be able to read past the surface ripples caused by the hype.
Cloud computing actually is a well-defined term and its meaning is close to what is implied. Cloud has been a metaphorical term and visual for depicting a network for a long time, especially the network outside the enterprise. The telecommunication infrastructure, which evolved into the backbone of the Internet, has always been drawn as a cloud and this is what basically the term cloud refers to, the network outside the internal network of an enterprise. But gradually, most enterprises grew beyond a private network operating on physically separated, leased lines and have moved into virtual private networks (VPN) on a shared infrastructure that is also shared by the Internet. This is what is implied by a “private” cloud, more or less. But increasingly, enterprises are also connected to the Internet, which today is the whole web of public (unsecure) and private (secure) interconnected networks of varying bandwidths (typically slower closer to the end-nodes which are the servers, computers and devices on the network). A “hybrid” cloud is one that uses both private and public networks to deliver information and functionality over the Internet. By this definition, most every business conducted on the internet uses a “hybrid” cloud.
So how does all this relate to the insurance? What are insurer preferences? By-and-large, insurers have opted for hybrid cloud environments for non-core services, such as CRM and email, where computing capacity is a combination of dedicated private server and shared public capacity. But when it comes to critical core systems, such as policy adminstration, the preference has been for private clouds with dedicated private servers that maximize security and privacy (see blog Securing the Insurance Cloud).
For more information on the topic, download the white paper Evaluating Core Systems in the Cloud.