The new year is when I like to take the time to reflect on the PaaS (Platform as a Service) market. I’ve long been a huge proponent of PaaS and have been waiting a few years now for PaaS adoption to explode.
What has happened in the last 18 months?
PaaS pure plays, especially Pivotal, have shown impressive growth in the last year and half. However, most PaaS pure play implementations are being done in a group or department within a larger company. This is a maturity issue. Most companies wisely choose to invest in PaaS to solve a specific problem, or to address a group of applications before rolling it out company wide. I believe in the next year or two we will see more wide-scale implementations across organizations, but not all organizations are looking for enterprise-wide PaaS adoption just yet.
Two developments are challenging the traditional PaaS pure players. First, the major public cloud vendors all have a good PaaS story. Both Google and Microsoft have a robust PaaS solutions that integrate with their IaaS solutions. AWS doesn’t have a packaged PaaS solution, but they do offer a number of high level PaaS services that are fully integrated with their IaaS services. AWS offers PaaS services in IoT, Big Data, machine learning, mobile, and many other areas.
What the pure PaaS players won’t admit is that not every workload is a candidate for PaaS. There are in fact many workloads still better suited to IaaS. The integration that the Big 3 public cloud providers deliver between their IaaS and PaaS layers reduces a ton of complexity compared to using a PaaS pure player and integrating that solution with other IaaS providers.
The other big development is containers as a service. Many companies are looking at containers to build their own PaaS-like platforms to meet their specific requirements. This area is new and emerging, but should gain a lot of steam in 2016 and beyond.
The PaaS stack players (e.g., Pivotal, CloudFoundry and Red Hat, Apprenda, OpenShift, etc.) will dismiss the DIY PaaS concept. They love to use the following image to ask why in the world would anyone roll their own PaaS when you get a solution from them out of the box.
The picture does present a compelling argument, but what if the company doesn’t want to lay bricks in a straight line? What if they’re required to build something really complex, with unique requirements that aren’t met out of the box by PaaS providers? That’s very often the case in large enterprises, which is why they’re very intrigued with using Docker to build their own containers as a service solution.
Many companies are experimenting now with Docker containers to build their own PaaS. Docker is aggressively addressing their gaps in areas like security, orchestration, network and storage layer abstraction and others. So, with every release of Docker, the DIY PaaS story becomes more compelling.
Another factor that is playing a part in the decision making is that a lot of companies are declaring that they are “all in” when it comes to the public cloud. Companies like Capital One and General Electric have made very strategic bets on the public cloud and are greatly reducing their data center footprint over the next several years. Many of these same companies will likely look at the PaaS offerings that their IaaS providers have developed, rather than go down the private PaaS road.
The reality here is that both pure play private PaaS solutions and public cloud providers are gaining lots of PaaS customers. But keep an eye on Docker. There’s a good chance they’ll claim a large number of PaaS workloads in the near future, as their portfolio grows and matures.
What is driving PaaS demand?
When enterprises first enter the cloud, they usually approach it from an infrastructure mindset. Putting SaaS aside, most enterprises pick an IaaS provider, either public or private, and then spend a good 2 to 3 years or more learning and maturing, as they implement their first few iterations of cloud solutions. Eventually, they get fairly good at IaaS and then start looking for ways to get even more agile. That’s when they start taking PaaS seriously as the way to reach the ultimate level of agility. What we’re seeing now is that many enterprises have been through those early years and have now become cloud competent.
These enterprises understand the value of PaaS and are becoming more interested in it than in previous years, as they seek to increase innovation and reduce time to market. This is why you’re seeing the Pivotals of the world doing so well. The demand is finally here- and it’s only going to increase.
Who is going to win?
That’s the million dollar question. For the next 2 to 3 years, there’s going to be a lot of winning by a lot of different companies. Pivotal is killing it in the private PaaS area. The Big 3 public cloud providers are killing it as well, although AWS is gaining way more PaaS workloads than Microsoft and Google right now.
Docker is set to explode in this area over the next couple of years. I predict that we’ll see more consolidation in the PaaS pure play area and more pivots. Last year we saw CloudBees pivot from a PaaS pure play to a Jenkins as a Service offering. Stackato was acquired by HP. Pivotal, Docker, Microsoft, IBM and others are gobbling up smaller companies to beef up their portfolios. There will be a lot more of this activity in 2016 and beyond.
PaaS adoption is on the rise for 2016. The choices are many and there is enough demand for multiple vendors to do extremely well this year. With the surge in public cloud adoption and the emergence of containers, it will be very interesting to discover who the winners will be 3 to 5 years down the road. Expect to see more consolidation and more PaaS vendors pivoting. It will be hard for the smaller players to compete with the money that Pivotal, Docker and the Big 3 IaaS providers have. We have already seen many IaaS vendors pivot and we will see the same with the PaaS vendors this year. It should be fun to watch.