It seems like every day a new statistic is published about growing cloud computing adoption and what it means for the future of IT shops and network administrators. Headlines might have you think that IT shops are all-out buying into the cloud -- despite a complete lack of protection against threats. We sat down with Stefan Ried, senior analyst at Forrester Research, to discuss UK cloud computing trends and set the story straight on adoption challenges. (Hint: The problem is not security!)
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
There are several forms of cloud computing -- Infrastructure as a Service (IaaS), Platform as a Service (PaaS), Software as a Service (SaaS). Is the adoption of any one of those in the UK different than the rest of the world?
Stefan Ried: The different cloud computing categories have different adoption trends. If you compare the UK, Germany and the US, in all of them the adoption of SaaS is leading the cloud movement. What is special in the UK is how people are using it.
We asked SMB and enterprise users in a recent survey if they are going to replace their old stuff with SaaS or if they are going to complement it. In the UK, 23% of all people we surveyed for customer relationship management plan to complement their existing environment with Software as a Service. Another 7% are already using SaaS to complement CRM today. Thirty percent are using it for campaign management, lead management or the call center -- they’re spreading functionalities across different solutions.
In the US, replacement is on the stronger side. When we asked, ‘Are you planning to replace your CRM systems or have you already replaced most of your CRM?’ -- 9 and 13% in the US answered yes. In the UK, it is 7 and 8%. So replacement in the US is stronger while complementation in the UK is stronger.
What applications/services are businesses most commonly using cloud services for?
Reid: There are two categories of applications. One is the front office -- the external-facing activity of an enterprise -- your customers, suppliers; and on the other side you have the internal applications -- back-office apps like financial apps, balance sheets, back-office manufacturing and production. If you compare those, you see that the front- office apps are still leading with 7% more adoption than back-office applications. Until recently, people thought cloud computing was not really applicable to back-office apps because of security and privacy concerns, but our survey [shows] this concern is being overcome by the customer.
What are some cloud computing adoption barriers?
Reid: When we talk to clients, the No.1 concern with cloud computing is security and privacy, but if you only talk to those that are already using some cloud computing services, it turns out that that security is not the No.1 concern anymore. The No.1 concern is integration between the cloud and on-premise solutions. If you ask people that are heavily using cloud computing resources on the SaaS and IaaS side if they have seen a security incident, you hardly find one because the cloud computing providers are so eager and so professional about security -- and on a much higher level than corporations because they know they will go out of business.
I always say the more data you have in the cloud the more secure you are, because then you don’t have data on your laptop, which you can lose. There are some exceptions, of course, but the cloud offers economy of scale. For most corporate data centers, the level of professionalism they can get from a cloud computing provider or professional outsourcing provider is better than what they can get on their own.
What technical challenges should network managers know about?
Reid: Network managers should look at the different natures of their workloads. A workload is something you need a piece of hardware or data storage for. There are three types of workloads. You can have production with few data volumes, production with heavy volumes, and test and development workloads. Production workloads that include customer data you might put in a cloud where you can control the location. There are workloads that require a lot of data that is on-premise -- say a terabyte or petrabyte of data. For those kinds of workloads it might be possible to get computer power elsewhere, but you won’t get the data there because it’s too big.
Network managers should look at the elasticity of their workloads and how they are fluctuating. They should see how they can swap those out of the internal environment and do cloud bursting. When the performance requirement increases, you swap out the workload to a trusted local provider that can integrate with your on-premise environment. If you have an internal self-service portal and you need a new development machine, you go to the portal, and if the internal environment is at high load, IT automatically spins up the machine externally. Decisions based on strategic policies can happen very, very quickly and take effect immediately.
About the author: Crystal Bedell is an award-winning writer and editor specializing in technology. She has written articles, tips and guides to help IT professionals evaluate technology, secure and modernize their IT infrastructure, solve business problems and prepare for IT certifications. As Principal of Bedell Communications, Crystal also analyzes technology trends and offers editing and content development services. She can be reached at firstname.lastname@example.org.