Here, you see a shift in intentions to use containers, a pretty large one: less people are planning to use them. To me, containers are mostly useful for custom written software, not business application workloads.
So, several years ago, containers seemed like a cheaper VMware strategy where you just generically throw your apps in and reduce costs.
But, that doesn’t really work. Apps have to be container friendly, plus, you know, you have to manage your new container orchestration thing – figure out kubernetes. Even that has only been a thing (an option a generic IT team would know about and find viable enough to consider) for about the past year.
(I mean, maybe, if you soften the idea that kubernetes is a platform for building platforms and just think of it as a platform for running apps, that is, a platform. I don’t know what the fuck is going on in that definitional-wrangling space.)
These companies, I’d theorize, then, had the wrong assumptions, investigated container usage, and realized it wasn’t what they were expecting.
Containers are for running custom written software. If you don’t want to do that, they’re probably not useful to you.
As an important side-note, let’s assume use here means penetration, which is to say, respondents use at least one, or as few as, container. That means overall usage could be a tiny percentage of their total workloads – or big. We have no way of know how many containers they use. Not a big deal if you’re interested, as here, in y/y trends, that is, growth. That’s what investors want to see.
Equally important and enlightening here, as always, is to look at the demographics:
I don’t know what the the n is, the total number of respondents. There’s a good chunk that of what I’d fall “enterprise”: 10,000+ employees, and lots from finance. Let’s say around half? Spending wise, education usually doesn’t spend much (or write that much custom software?), and “Technology” typically less. Tech companies usually don’t buy shit and DIY it choosing to spend on their own people instead of vendors – they are, after all, technology companies, they think.
Also, it’s worth weighting this all by how few insurance and retail companies are in the respondent base: they have tons of custom written software, the stuff you’d put in containers.
So, you’ve really got two very different surveys and conclusions going on here, split by two different markets: those who write and run their own software (mostly large organizations) and those who don’t (mostly smaller organizations).
You see the general conclusion in the footnote: 10,000+ people companies (who have a good chance of writing and running their own software) already use containers and plan on using more.
Anyhow, half of respondents are small and mid-sized companies, plus those tech companies that don’t spend. So, spending wise, selling containers is probably mostly a large enterprise play cause that’s where they get used and paid for. They rest of the companies, likely, want SaaS and security.
Check out the rest of the report. It covers much, much more that the container neck of the woods.