Yesterday I attended a presentation from NTT Communications, a managed hosting provider, on the plans of 200 CFOs and CIOs from larger UK organizations (500+ employees) with respect to cloud computing. Since NTT would presumably like more companies to stick more stuff on its hosted servers, I presume it was hoping for a strong endorsement of the idea. Unfortunately for NTT, that was not the case. Fewer than 20% of those surveyed think they are using cloud computing now, a bit more than 20% think they will adopt some of it in the next two years, but – and here’s the real killer – cloud computing is way down the list of investment priorities, at around 5%. I’m not clear 5% of what exactly; but the report says it is the lowest priority.
What are companies spending money on instead? Servers and storage, network infrastructure, security, company web sites, backup and disaster recovery, unified communications, desktops and laptops, software, almost anything else in other words.
What’s wrong with the cloud? The three top issues, for those surveyed, are security, immaturity, and reliability.
These are valid concerns, though each one is open to debate; but the entire survey was undermined by the fact that most of those surveyed admitted to not knowing what cloud computing is. The reason is not ignorance, but the many and various ways the term is used. The common strand is that it is something to do with the internet, but even that is undermined if we describe virtual on-premise servers as a “private cloud”.
What are the varieties of cloud? Almost infinite, but here are a few:
- Multi-tenanted applications such as Salesforce CRM, Google Docs, NetSuite. This is the model that has the biggest inherent economic advantage.
- Hosted application platforms including Google App Engine, Microsoft Azure, Force.com. These are hosted application servers, where you write the code, taking advantage of integrated hosted services for storage, identity, transactions and so on.
- Utility services such as Amazon S3. It’s a great example: S3 offers nothing but storage, though you can use it in conjunction with other Amazon web services.
- On-demand infrastructure such as Amazon EC2. You get virtual servers to do what you like with. NTT’s services are mainly in this broad category. It’s cloud but you are mostly not getting the benefits of multi-tenancy.
- Anything on the internet. Running a web application? Hey, you’re in the cloud.
If we are going to have a sane discussion about these things, we need to know what we are talking about. Maybe rather than asking companies whether or not they are doing that cool cloud stuff, it would be better to enquire how they see their use of the internet evolving.
Another big question is the extent to which companies are willing to buy in their IT infrastructure as a third-party service. Although it makes obvious financial sense in most cases, it is a big ask given how business-critical it is, hence the concerns about security, immaturity, reliability.
Smaller companies with ad-hoc IT systems are likely to be more amenable to the idea, but this group was not covered by NTTs survey.
Conclusions? The main one is “watch this space”. In the end I reckon sheer economics will drive cloud computing adoption – in all the areas described above – but the one thing NTT’s survey proves is that larger organisations are in no hurry to make that jump.
The thing is, in the past any company which buys in their IT services from a third-party has usually results in really poor IT services.
You can’t beat in-house, familiar with the specific implementation, IT support.
Granted, that is changing. As with cloud computing its less and less about specific implementations and more about keeping the basic infrastructure, which all implementations are designed to run on, working. It will still to some expect be down to individual companies to deal with their specific implementation issues but its a lot less about the hardware you are running it on today.
I for one though do not relish the day when I trust all my e-mail on a third-party. I long since stopped using webmail for that reason as I would much rather keep my e-mail storage under my own control, where I can easily back it up if I want or transfer it to another PC easily.
Again that may be changing, I do not know if such as GMail let you backup your inbox/filters/etc as I do not use it, but the way things are evolving I can see it being perfectly plausable as they want to pull everyone over to cloud computing and knowing where your backups are is one of the biggest hurdles of trusting someone elses servers. Security of that information of course being the biggest. I mean, are Google reading my e-mail?
Of course for e-mail its relatively mute. While a lot of servers talk SSL these days it by no means guarantees that you e-mail has reached you without anyone reading it on its way. But then, neither does traditional post or the telephone which people tend to feel is more secure but in reality is not.
For me personally, I like the security of having a traditional web host for my sites and e-mail. I can easily move all my data to a different host if they go out of business or get too expensive/unreliable. I am unsure how this works with cloud computing though.
However open standards based it may get, I cannot visualise how it works and it makes me concerned that relying on a specific provider locks me into that provider. Running your own private cloud (which is basically just your own server farm, what most companies have had for years but a new name for it) you know where your data is which is very reassuring.
(declaration of interest: I work for NTT Europe Online, the company that commissioned the original report…)
The debate continues as to what cloud computing has to offer large businesses at this stage of its evolution. The Cloud or Fog? report was aimed at finding out what CIOs and CFOs are prioritising in terms of IT investment in the next few years and the fact that 67% are not looking to invest in cloud computing in the next two years is no real surprise (to us at least). As you say, there remains a lot of confusion and uncertainty around the term itself, and rather than undermining the results, I think this points to what we all know already – that too many products and services are being marketed as cloud computing and unfortunately it’s become a technical sell rather than a business and operational discussion, which is where the true value lies.
Our report did reveal that 60% of CIOs and CFOs at UK enterprises are more inclined to invest in software and services delivered online as a result of the recession and that the vast majority are looking for more flexibility in their software licensing agreements and want to avoid long term IT contracts. One could argue this can be delivered via a cloud model, but as enterprises remain uncertain of the security, reliability and maturity of cloud computing it appears alternative options are required.
Alex Atkin’s comments on private cloud are very relevant here and I would say are representative of many people: private clouds delivered by managed hosting organisations offers flexible use and dynamic provisioning of computing resource. The provision of an SLA can (despite Alex’s concerns) provide peace of mind for reliability and security of business critical IT services. The type of solutions provided by the top managed hosting providers are not comparable to shared web or public cloud services such as EC2 though. The demand for the capex vs. opex benefits (or even more effective use of opex) offered through online delivery of software and services remains a consideration and as the hosted private cloud model ticks all the boxes for the concerns outlined by CIOs and CFOs it can be an attractive option for mid-size and enterprise businesses.