Seven things, including Self-Service Automation, that IT teams must do to retain relevance with end users.
I was talking with an industry analyst (think Gartner, Forrester, IDC) the other day around a broad range of trends impacting IT. Somehow we got onto a discussion around the issue of IT teams losing relevance with their line of business customers. I can’t recall the conversation exactly but it went something like this. “David, I talk with different IT teams every week and many of them ask me the same question: “What can we as IT do to be more relevant to our end users [line of business]?”.
Jim (not his real name) told me that the first question he asks these teams in response is “Are you offering your customers self-service?”. This analyst told me that the answer he hears back most often is “no, we haven’t gotten to that yet”. Jim then goes on to advise these teams to A) leverage an automated approach to service delivery to speed up resource delivery (if they are not already doing so); and B) be sure to also implement self-service that makes it drop dead easy for end users to get the services they want.
If you think about it, not implementing self-service is denying the reality that line of business partners have choices beyond enterprise IT. It also fails to recognize that increasingly our expectations of how things should work, at work, are shaped by our personal and consumer experiences. Self-service and the near instant gratification that comes from it just makes more sense today than submitting tickets and waiting weeks for resources to be available for your next critical project.
My Top “X” List For IT
This exchange got me thinking about the big-ticket items that most IT teams must tackle to be more relevant to their end users. If the # 1 thing that IT teams must do to retain or regain relevance is embrace self-service; what does a top ten list look like? Sorry to disappoint but I don’t have a top ten list. There are however some things that I feel do stand apart from the rest of the pack when it comes to looking at how IT operates. So, in that spirit here is my list of the top seven things IT must do to remain relevant.
1. Implement Self Service for Resource Requests
2. Market IT Services to your End Users
3. Enable Infrastructure as Code
4. Become an IT Developer
5. Begin to Think about Multi-Cloud Networking
6. Go Beyond Infrastructure and Deliver Complete Stacks
7. Help App Dev Teams Move Containers to Production
There are undoubtedly other things that IT teams can do that would increase their relevance to line-of-business (LOB) partners. Having said that, I do think this is a pretty good list to start with. There’s too much here to cover in a single blog so I’ll elaborate on each of these in this blog and several others that will follow. Hopefully, along the way I will provide you enough insight on each to give you a good idea of what it is that IT must do along with some additional thoughts on how to get it done.
Starting with Self Service
According to Wikipedia and depending on how you look at it, Amazon Web Services has been around since 2002 or 2006. Early adopters flocked to it because of two reasons in my opinion. The first reason was an ability to get infrastructure fast. The second reason was the ability to get these resources without having to file one or more tickets with the help desk.
Today, implementing the ability to get end users resources fast is simply a matter of automation. Many organizations have adopted automation to dramatically speed up the provisioning of infrastructure resources. Application level resources is a different matter but we’ll cover that elsewhere.
I have first-hand experience talking with many IT teams who used to take 4 or more weeks to provision resources in the past but now routinely do it in under in under thirty minutes. Of course, with Amazon you can get those resources in just a few minutes, so taking 30 minutes or so is still longer than what it would take using AWS. But let’s be honest – how many developers find out about a project and then need to be coding it 5 minutes later? Thirty minutes is plenty fast for most needs.
While many organizations have, or are in the process of adopting automation to speed up service delivery, not nearly as many have implemented self-service as part of that process. Many still rely on existing request fulfilment processes that existed before automation was implemented. The most common example of this is organizations using Service Now for requesting resources, which in turn generates a ticket to the platform automation team which then initiates an automated process to fulfill the request.
Leveraging an existing ticketing process isn’t necessarily a bad approach and there are some good reasons for doing it. The main reason that I am aware of is that this approach means that any existing process for determining who has access to specific resources doesn’t need to be re-codified into the automation that supports self-service.
That’s not a bad reason to keep the existing process, but remember that if you are an internal IT team, your competing with the public cloud and on the public cloud – self-service means self-service. No tickets and no help desk. So, going the extra mile to enable true self service where entitlements and other forms of governance are matched between users and resources might be worth it for your IT team given the world we live and compete in.
Now a few caveats around the idea of self-service. Different end users have different needs. Many end users are perfectly happy selecting resources from a pre-populated catalog. VMware vRealize Automation is a great example of an automation platform that supports this model of self-service.
In this model, blueprints can be created to represent everything from a single machine to a complex, multi-tier application, with storage, networking, security and even monitoring agents all represented in the blueprint. These blueprints then become catalog items that once selected by end user are instantiated in near real time.
Other users might prefer a self-service model that is closer to what they would experience on Amazon. This model is declarative in nature and resources are requested either through a CLI or through an API (using scripts or through another tool) in the form of small building blocks that represent infrastructure elements such as compute, storage, or network. For IT teams looking for such a model to satisfy their end users, VMware Integrated OpenStack (VIO) might be the best choice for a service delivery automation platform.
A hybrid model might be the best choice for others. In this model vRealize Automation is used to offer VM level resources from a catalog but it is also used to reserve resources for a VIO based developer cloud that an App Dev team would like to implement. In this model vRealize Automation would also be used to provision the components necessary to instantiate a VIO based Developer Cloud for that same App Dev team.
Just for completeness, I should point out that vRealize Automation can also support the idea of blueprints as code, where blueprints are created or modified using YAML. These blueprints can then be imported into vRealize Automation and offered to end users through the catalog. These same blueprints can of course be exported as YAML as well.
The Right Self-Service Model for Your End Users
Hopefully you can see that solutions to the self-service problem exist along a continuum. Figuring out what type of self-service model to implement is very much a function of understanding your users. There are different approaches and you won’t be sure which approach makes the most sense unless you are actively engaged in understanding the needs of your users.
Having a deep understanding of what your end users need is also the prerequisite for our next “must do item” which is effectively marketing what you do offer to your end users. More to come on that in the next installment of this series.