Daily Archives: January 25, 2016

How Do You Value Cloud Services?

The clouds are here to stay. Whether they are dark nimbostratus storm clouds filled with hail or fluffy white cumulus clouds that dot the clear blue skies, they’re here. (That’s why the doctor recently co-authored a series over on Spend Matters Plus with the prophet on Supply Chains in the cloud.) Regardless of the doctor‘s opinion on whether your supply chain should be in the cloud, the clouds are sweeping supply chains up and the situation has to be addressed. (Thus, one has to do one’s best to insure that one’s supply chain is in the way of the right cloud.)

And while you should be well aware by now of how to cost a cloud-based platform, and compare it to a hosted ASP solution and an on-premise solution (as the referenced series and a number of posts here on SI have addressed this issue in detail in the past and even provided you with spreadsheet templates), you might not be aware of how to value a cloud-based solution.

When it comes to the cloud, valuation is a very difficult concept. There’s the hardware infrastructure and the reliability that comes from multiple locations that can store your data and run your applications. There’s the cloud-OS layer that handles real-time on-site and off-site data replication and back-up, automatic start-up of new processes and machines when a process or machine fails or becomes unavailable, automatic allocation of more processors and memory and storage when usage spikes, and so on. There’s the application layer that not only enables your processes but that is accessible anywhere with a data signal on any device your people happen to be carrying, that supports real-time data sharing and collaboration with your supply chain partners, and that supports innovative new capabilities not possible in on-premise apps.

There is a lot of value in each of these layers. Access to more hardware than you need, or can even afford, is valuable. Real-time off-site backup and failover is valuable too – compared to having to manually bring up an off-site location. And a better application with more capability and innovation is valuable too, but just how valuable?

In the traditional hardware world, the cost of filling a data centre is the cost of hardware plus the cost of a network engineer setting it up. Hardware is the cost of production plus a fair margin – there are enough essentially equivalent providers that costs are kept in check.

In the traditional software world, the cost of software is generally computed as the overhead cost of the company that produces it plus a margin that will produce an acceptable margin that the company can get away with based upon the perceived value differential between it and its competition that it can sell.

But the cloud is not set in the traditional world. In fact, the real-tine off-site backup and failover in a virtual OS layer didn’t even exist before the cloud. How much more valuable is having access to as many machines as is needed to power your application at full capacity at all times? While this power is known, failure — be it machine failure, power failure, or communication line failure — cannot be predicted and sometimes the entire application infrastructure must be ported in real time to a different part of the cloud.

And how much more valuable is having software that is maintained and regularly updated by the provider as compared to having software that must be manually updated and kept up by in-house development staff? Especially when that software might be capable of offering more real-time collaboration, real-time product tracking, market intelligence, and analytics than an on-premise platform. This is a much harder question to answer.

But one that should be asked. Just because a cloud solution is the cheapest alternative, that doesn’t mean that you are getting the full value you could be from your money. There are multiple providers, and they won’t all charge the same. Plus, if the technology is relatively simple, if its implemented as a true multi-tenant cloud based platform, and it doesn’t need to be updated very often to meet your needs, then the platform likely doesn’t cost the provider very much and may not have the value the provider claims if another provider offers essentially the same platform for three quarters of the cost.

There are no good answers here, but the questions should be asked and good answers should be expected before you commit to a solution, even if you are a non-profit that was donated a certain amount of cloud services — because you might not be getting what you think and may get hit with a big bill at the end of the year if your acceptance entails an agreement to pay for any usage above the donated amount of services.

Since there are no standards, providers are more or less free to “Value” services anyway they want, make extravagant claims as to support costs, and value a service at 5X its cost, or more. So be careful.