Have a look at this week's Cold Start newsletter from ServerlessOps!
The Week In Serverless Hot Takes
The past week I've seen serverless Twitter aflutter with a blog piece and a tweet. Both of these are I think distractions.
Serverless Without Containers
To start, how my serverless platform works is largely irrelevant to me. That’s the point. What matters to me is what the platform provides. For every fault of Lambda@Edge, I'm sure there's a ticket somewhere at Amazon to address it. Discussions about containers versus JavaScript V8 Isolate and how one company architects their serverless platform is fun for some people to nerd out on but not relevant to future serverless growth.
Infrastructure Provisioning Is Hard
Second, don’t confuse individual developer problems with organization problems. How many organizations are making devs configure their own infrastructure? How many even let them? Individual engineers largely don't pay money to solve their problems, their employer does. If it's not a problem to their employer then no one is going to give you money.
That leads me to...
What Really Moves Serverless Forward
Serverless still has to connect with business decision makers to further grow to relevancy. It doesn’t connect with arguments about whose's platform has a better architecture or the developer experience of one of its employees on their own time.
Trying to go out and "sell serverless" early on, and failing, was super important. This is why I, and as a company we, talk about serverless so differently from many others out there. Once you leave the serverless bubble and talk to regular people you realize we have a lot further to go.
Serverless needs to be positioned as addressing problems. This is the reason ServerlessOps leads with "DevOps transformation" and not "AWS serverless adoption". One is a problem. The other is part of a solution to a problem.
People buy solutions to solve problems. We need to remind ourselves of this regularly. We are here as companies to solve people's problems. Though, that may be hard to see when we as a community are still primarily in the phase of building tools to solve the problems our solutions create.
We also need to remember that most of the gains we tout for serverless are still theoretical at best. "Serverless will make you more efficient and therefore a greater success." If you adopt serverless you don’t automatically obtain its benefits or reap its rewards. Tools enable you. That is all. Success or failure is up to you. We need to still prove that our technical solution (serverless) actually solves problems in the real world and demonstrate that value back to people.
Serverless Tool Roadmaps
The folks at The New Stack released data and infographics on serverless tooling adoption from their recent survey. With a sample size of 382 there's only so much you can possibly conclude from the data. But there is one striking thing in the monitoring space.
The data shows 47% of respondents use CloudWatch while the SaaS products out there are all in low single digits. What that indicates to me is for each of these companies the biggest competitor is the status quo and not each other.
Reasons for these numbers? Here's some thoughts.
- CloudWatch is really good enough for people (no market)
- Serverless applications have not arrived to scale where CloudWatch is not good enough (market not yet arrived)
- Products are not yet compelling enough to adopt
- Audience lacks of awareness of products out there
- A reason I've missed
I don't know the answer but they're interesting questions to ask and think about.
Closing
I'm Tom from ServerlessOps and we provide a range of services around DevOps transformation, AWS migration, serverless training, and startup cloud operations. See our services to learn more.
Also let us know what you think of this week's Cold Start using the 👍/👎 links at the end of this email!
Cheers,
Tom at ServerlessOps
