Hi, I’m Forrest Brazeal at Trek10, and this is ‘Think FaaS’, where we learn about the world of serverless computing in less time than it takes to run a Lambda function. So put five minutes on the clock - it’s time to ‘Think FaaS’.
Last week was ServerlessConf San Francisco, and my co-host Jared Short and I did a talk there called “How To Win Coworkers and Influence Organizations For Serverless.” That talk will be available as a video on A Cloud Guru’s platform soon. I hope you’ll watch it because it has prop comedy and cartoon characters and all sorts of awesome stuff. But for those who may have missed the conference itself, I wanted to give you a sneak preview today.
Basically, as Jared and I have worked with a variety of different organizations as consultants here at Trek10, we’ve identified some types of folks who have predictable difficulty getting on board with serverless. It’s not necessarily their fault, either. They may have very real business pressures that are getting in the way. They may have legitimate technical concerns, because as we all know serverless is not a silver bullet for every use case, and tradeoffs have to be made.
And what we’ve come up with is sort of a three step approach to help you talk to these people and get them on the road to serverless adoption. We use the words Empathize, Educate, and Empower. The idea is you want to listen to them first, hear their concerns, rather than just running roughshod over what objections they may have. You want to help educate them where they may have misperceptions due to the FUD that swirls around these technologies, perhaps from what I like to call ‘unscrupulous vendors and sinister purveyors of legacy IT’. And finally you want to actively move them to a point where they’re seeing tangible benefits from serverless in their role.
We have five personas we’ve identified within this framework. The first is the legacy leader. This is someone who has teams reporting to them, maybe has invested money in some older technology like a datacenter. So to empathize with them, you need to realize that committing a lot of time and human resources to every new thing that comes along may be a nonstarter. With that said, you can educate them by pointing to the very real time to market and TCO impacts that serverless can have. You can help them find upcoming projects that can get done in record time using serverless. Enabling quick wins that make the whole team, and by extension the boss, look good, is not a bad way to empower a legacy leader toward serverless adoption.
Second, we talk about the antagonistic architect. All software architects are a bit antagonistic from time to time, because they’re protecting the health and well-being of the stack. But they’re going to have concerns about things like vendor lock-in, about performance at scale, basically — does serverless live up to the hype? And you can educate this person on these subjects by providing some of the details in previous episodes of this podcast, but recognize that they’re probably still going to have concerns, and that’s fine. You can empower them by getting them started on a project that’s been a pain point with traditional architectures, like a system that’s handling spiky workloads. And as they start to see the power of serverless, they’ll run with it on their own.
Third, we have the server-hugging sysadmin. Maybe this person learned to code on a PDP-10, maybe they’re relatively new to the field, but they can’t imagine running their shop without access to OS-level metrics, performance settings, and so on. Really, this persona gets hit the hardest in the serverless world, because being an admin is all about control. You administrate things, it’s right in the job title, and serverless is telling you to let go of that. You can educate and empower this person by showing them that serverless certainly still has an operational component. Software is still getting delivered to users, and somebody has to make sure those users are happy. But the responsibility shifts from maintaining infrastructure to instrumenting code and config.
That brings us to our fourth persona, the disrupted developer. This hits home for me personally, because let’s face it, the serverless development experience is not always awesome. The tools aren’t where we want them, the old-school frameworks seem to have gone out the window, and don’t even get me started on local testing. You can educate the developer by pointing them to tools we’ve mentioned here before like Cloud9, and steering toward client-side frameworks like React that still play a big role in the serverless world. But kind of as the flip side of what we told the sysadmin, developers actually get to have a little more control with serverless than they used to. Granted, the infrastructure is abstracted. But the pieces you need to assemble to deliver an application are now high-level enough that you can hold them all in your repository and, to some extent, in your head. And that’s a pretty empowering thing.
The fifth persona, somewhat tongue in cheek, we called the HackerNews hater, because you usually encounter this person behind the comforting anonymity of a comments section. This is the person saying “Serverless still has servers”, and all the other things that add more noise than light to the discussion. In our talk, Jared and I encouraged the audience not to focus on engaging with this type of debate. Far better to use the limited time and energy you have to win hearts and minds within your own organization.
Because ultimately, this is about delivering value. And you’ll create value by empathizing with, then educating and empowering, the people who are key contributors to your organization’s ongoing serverless success. In the meantime, you can keep up with all things serverless by following Trek10 on Twitter @Trek10inc, I’m there as well @forrestbrazeal, and we’ll see you on the next episode of Think FaaS.