Additional services include security, for example built in DDOS protection, as well as many others. Quality and (cyber) security are important and more readily available in the cloud. On top, life cycle management and back-ups (in case things go South) is something you don’t need to worry about much, aging physical servers, daily insertion of back-up tapes etcetera can be history. And you pay what you use, so if you are able to shut down applications at night or in the week end, this saves cost in a Cloud environment versus traditional hosting. Of course this requires a different approach (and people capability) versus traditional Server Operations.

Yet another area is marketing initiated web sites. Often times, marketing agencies offer an end to end service for campaigns and persistent web presence. All good, but in my experience security aspects are oftentimes overlooked and maintenance is close to non-existent, leading to friendly and sometimes not so friendly hacks. Creating a central Cloud based web site service from within IT allows for better security and controls.

Worry about the Vendor eco-system

So do we move everything to the cloud? No. When crisis hits, you typically rely on your vendor eco-system to help resolve these. While hosting applications in the cloud is very attractive, you also need to take into account how important you are for your cloud partner. Will they really go the extra mile for you when your global critical systems are down? For that reason we decided to keep our global critical systems with traditional hosting partners for now. That might change in future but only after we gained sufficient experience with the vendor eco-system in the cloud. On top, some applications require local servers (for latency reasons for example). That’s a real challenge we are still grappling with. My definition of cloud nine for this area is to have zero local servers, or if they are needed, that they are managed by an IT company (not by the Dairy Company we are).

And worry about the impact on your network

In our network environment we typically have primary leased MPLS lines and secondary Internet break-outs. Moving to the cloud forced us to relook at our network strategy, rebalancing primary and secondary line capacities (for quality and cost) and starting our Software Defined Wide Area Network (SD-WAN) project to be able to better balance all types of traffic over the bandwidth we have contracted and pay for. As an example: when we initially moved on-premise Outlook e-mail and SharePoint to the cloud and added capabilities like OneDrive we needed to really upgrade our secondary lines. Offsetting this by decreasing primary line bandwidth was hard work but also saved quite a bit of cost in the end. Our SD-WAN strategy will ultimately prepare us for the future: an internet and cloud centric IT environment with user experience as the core success measure versus traditional up-time / bandwidth discussions.