Lessons learned from a year in the SHI Cloud: The business model and transformative ease-of-use

Lessons learned from a year in the SHI Cloud: The business model and transformative ease-of-use

If you’ve been watching this space over the past few weeks, you know that my past two blog posts have been part of a series of posts dedicated to reflecting on the first year of the SHI Cloud.

When we launched our cloud offering last summer, we knew that interactions with our customers would validate the true differentiators in our service and help us identify areas for improvement. So far, I’ve covered how our unique networking design has helped us build a successful cloud model for our customers, how we learned that customers prefer simplicity in their cloud service, and how keeping track of every single detail of the cloud for the past year has paid off.

Today, I’m nearing the end of the series with the last two lessons that the past year has taught us: the business model conversation and the importance of transformative ease-of-use.

Lesson #4: It’s a business model conversation
Over the past year, we’ve worked with large enterprises and small businesses (admittedly, small businesses with significant technological needs), research universities, state and local governments, and a lot of customers in between. We’ve learned that the SHI Cloud model scales all the way up, because every size company can always leverage truly excellent infrastructure as a service (IaaS).

One of the most successful adaptations we have made over the past year is in how we interact with our customers. We are more and more focused on the “use case” and less on promoting the details of our service. After all, our service is simply an enabler of customer projects and their use cases. The conversation we have with customers now is a business and operational model conversation, not a product conversation.

For example, one of our first customers was a company with a few hundred employees that develops laboratory information management systems (LIMS) — the software that keeps a lab organized, compliant, and productive. When the customer contacted us, it needed a couple of things. The first had to do with the training sessions it hosts on how to use its software. For every class it needed to instantly spin up 20, 30, or more virtual machines running its software as if it was a full production environment. Each student would have a complete environment to use as his or her own during the class. When the class was over, the customer’s IT staff would wipe out the environment and then spin it all up again the next week.

So when we met with this customer for the first time, everyone saw a great opportunity to provide an elastic, utility model for their virtual training environment. But the customer is a growing, forward-thinking company that also wanted to build a software-as-a-service business to host its application for smaller laboratories in a multi-tenant environment. This presented a very different use case, and our conversation was suddenly less about virtual machines and more about helping the customer’s business grow by supporting a new business model.

The more conversations we have with customers, the more we come to see that each one about the cloud is a strategic conversation in support of new business models the customer is pursuing. SHI’s ability to advise the customer, as we have in so many other areas of IT, is what sets us apart as a service provider in this new technology paradigm. The SHI Cloud is an enabler of new models for our customers. That alone makes coming to work every day an exciting proposition.

Lesson #5: Ease-of-use is transformative
Our conversation with the above mentioned customer led us to look at how we could make it easier to requisition and clone new virtual machines. The idea wasn’t necessarily to create something specific for that customer, but to look at its use case to see if there were things it was doing that might appeal to other customers. Simple enhancements to the portal interface and our orchestration layer made things more organized and efficient, allowing the customer to prepare for classes quicker and free up staff to do other, more valuable tasks than infrastructure management. So we dove in and started to make things more flexible. We really thought about their particular use cases and tried to envision the types of activities other customers might be doing that would need that same flexibility.

As an example of how this sort of thing builds on itself, our improvements to the requisition and cloning functions certainly helped the initial customer. We were then able to build on those with even more improvements for a high-performance computing project at a major university to accelerate its ability to take advantage of rapid cloning to support burst computing requirements. Then we engaged with a gaming company that does very large-scale testing of their massive, multi-player role-playing games. They spin up hundreds of virtual machines to test their games before they release them to the public. So they built off of the work that we did with the university, just as the university built off our work with the laboratory management firm.

As we talked among our customers, we started to see people grouping around certain models of computing: traditional, persistent infrastructure to support production applications; time-boxed burst computing to support both predictable and unpredictable capacity spikes; and high-performance or transaction-intensive computing to support large scale processing requirements (such as research and load testing). There are more models that we will inevitably confront as use of the cloud evolves, but these three became apparent to us quickly. And interestingly, these models are not industry-specific or market-segment-specific. Different organizations — and it could be multiple organizations within the same company — engage in different models of computing to support wildly different business use cases. By paying close attention to what makes our customers successful and how they approach computing, we begin to see how computing is evolving and we have focused our design process on understanding and accommodating these core cloud consumption models.

We continue to spend a lot of time on the user experience, striving to create a smooth, no frills experience. It’s about building good business process into the SHI Cloud and beginning to create more flexibility, because every IT department is going to manage things a little bit differently and you want to be able to support what the customer needs to do. There’s more and more standardization around ITIL, so using ITIL as the focal point for our operations team and for our UI design has the potential to allow us to accommodate the widest variety of customers in terms of their management process.

That wraps up the five lessons SHI has learned since launching our cloud offering one year ago. But we’re not quite done yet. In my next and final post in this series, I’ll outline my predictions for the SHI Cloud’s future.