10 KiB
5 steps to building a cloud that meets your users' needs
This article was co-written with Ian Tewksbury.
However you define it, a cloud is simply another tool for your users to perform their part of your organization's value stream. It can be easy when talking about any new paradigm or technology (the cloud is arguably both) to get distracted by the shiny newness of it. Conversations can quickly devolve into feature wish lists set off by a series of never-ending questions, all of which you probably have already considered:
- Will it be public, private, or hybrid?
- Will it use virtual machines or containers, or both?
- Will it be self-service?
- Will it be fully automated from development to production, or will it have manual gates?
- How fast can we make it?
- What about tool X, Y, or Z?
The list goes on.
The usual approach to beginning IT modernization, or digital transformation, or whatever you call it is to start answering high-level questions in the higher-level echelons of management. The outcome of this approach is predictable: failure. After extensively researching and spending months, if not years, deploying the fanciest new technology, the new cloud is never used and falls into disrepair until it is eventually scrapped or forgotten in the dustier corners of the datacenter and budget.
That's because whatever was delivered was not the tool the users wanted or needed. Worse yet, it likely was a single tool when users really needed a collection of tools that could be swapped out over time as newer, shinier, upgraded tools come along that better meet their needs.
Focus on what matters
The problem is focus, which has traditionally been on the tools. But the tools are not what add to your organization's value stream; end users making use of tools are what do that. You need to shift your focus from building your cloud—for example, the technology and the tools, to your people, your users.
Beyond the fact that users using tools (not the tools themselves) are what drive value, there are other reasons to focus attention on the users. The tools are for the users to use to solve their problems and allow them to create value, so it follows that if those tools don't meet those users' needs, then those tools won't be used. If you deliver tools that your users don't like, they won't use them. This is natural human behavior.
The IT industry got away with providing a single solution to users for decades because there were only one or two options, and the users had no power to change that. That is no longer the case. We now live in the world of technological choice. It is no longer acceptable to users to not be given a choice; they have choices in their personal technological lives, and they expect it in the workplace, too. Today's users are educated and know there are better options than the ones you've been providing.
As a result, outside the most physically secure locations, there is no way to stop them from just doing what they want, which we call "shadow IT." If your organization has such strict security and compliance polices that shadow IT is impossible, many of your best people will grow frustrated and leave for other organizations that offer them choices.
For all of these reasons, you must design your expensive and time-consuming cloud project with your end user foremost in mind.
Five-step process to build a cloud for users' needs
Now that we know the why, let's talk about the how. How do you build a cloud for the end user? How do you start refocusing your attention from the technology to the people using that technology?
Through experience, we've learned that the best approach involves two things: getting constant feedback from your users, and building things iteratively.
Your cloud environment will continually evolve with your organization. The following five-step process will help you create a cloud that meets your users' needs.
1. Identify who your users will be.
Before you can start asking users questions, you first must identify who the users of your new cloud will be. They will likely include developers who build applications on the cloud; the operations team who will operate, maintain, and likely build the cloud; and the security team who protects your organization. For the first iteration, scope down your users to a smaller group so you're less overwhelmed by feedback. Ask each of your identified user groups to appoint two liaisons (a primary and a secondary) who will represent their team on this journey. This will also keep your first delivery small in both size and time.
2. Talk to your users face-to-face to get valuable input.
The best way to get users' feedback is through direct communication. Mass emails asking for input will self-select respondents—if you even get a response. Group discussions can be helpful, but people tend to be more candid when they have a private, attentive audience.
Schedule in-person, individual meetings with your first set of users to ask them questions like the following:
- What do you need in order to accomplish your tasks?
- What do you want in order to accomplish your tasks?
- What is your current, most annoying technological pain?
- What is your current, most annoying policy or procedural pain?
- What ideas do you have to address any of your needs, wants, or pains?
These questions are guidelines and not ideal for every organization. They should not be the only questions you ask, and they should lead to further discussion. Be sure to tell people that anything said or asked is taken as feedback, and all feedback is helpful, whether positive or negative. The outcome of these conversations will help set your development priorities.
Gathering this level of personalized feedback is another reason to keep your initial group of users small: It takes a lot of time to sit down with each user, but we have found it is absolutely worth the investment.
3. Design and deliver your first iteration of the solution.
Once you've collected feedback from your initial users, it is time to design and deliver a piece of functionality. We do not recommend trying to deliver the entire solution. The design and delivery phase should be short; this is to avoid making the huge mistake of spending a year building what you think is the correct solution, only to have your users reject it because it isn't beneficial to them. The specific tools you choose for building your cloud will depend on your organization and its specific needs. Just make sure that the solution you build is based on your users' feedback and that you deliver it in small chunks to solicit feedback from them as often as possible.
4. Ask users for feedback on the first iteration.
Great, now you've designed and delivered the first iteration of your fancy new cloud to your end users! You didn't spend a year doing it but instead tackled it in small pieces. Why is it important to do things in small chunks? It's because you're going back to your user groups and collecting feedback about your design and delivery. What do they like? What don't they like? Did you properly address their concerns? Is the technology great, but the process or policy side of the system still lacking?
Again, the questions you'll ask depend on your organization; the key here is to continue the discussions from the earlier phases. You're building this cloud for users after all, so make sure it's useful for them and a productive use of everyone's time.
5. Return to step 1.
This is an iterative process. Your first delivery should have been quick and small, and all future iterations should be, too. Don't expect to be able to follow this process once, twice, or even three times and be done. As you iterate, you will introduce more users and get better at the process. You will get more buy-in from users. You will be able to iterate faster and more reliably. And, finally, you will change your process to meet your users' needs.
Users are the most important part of this process, but the iteration is the second most important part because it allows you to keep going back to the users and getting more information. Throughout each phase, take note of what worked and what didn't. Be introspective and honest with yourself. Are we providing the most value possible for the time we spent? If not, try something different in the next phase. The great part about not spending too much time in each cycle is that, if something doesn't work this time, you can easily tweak it for next time, until you find an approach that works for your organization.
This is just the beginning
Through many customer engagements, feedback gathered from users, and experiences from peers in the field, we've found time and time again that the most important thing you can do when building a cloud is to talk to your users. It seems obvious, but it is surprising how many organizations will go off and build something for months or years, then find out it isn't even useful to end users.
Now you know why you should keep your focus on the end users and have a process for building a cloud with them at the center. The remaining piece is the part that we all enjoy, the part where you go out and do it.
This article is based on "Design your hybrid cloud for the end user—or fail," a talk the authors will be giving at Red Hat Summit 2018, which will be held May 8-10 in San Francisco.
Register by May 7 to save US$ 500 off of registration. Use discount code OPEN18 on the payment page to apply the discount.
via: https://opensource.com/article/18/4/5-steps-building-your-cloud-correctly
作者：Cameron Wyatt 译者：译者ID 校对：校对者ID 选题：lujun9972