The End of an Era: Puppet Calls It a Day
Today marks the true end of the era of configuration management platforms. Perforce acquired Puppet after the configuration management pioneers had to abandon their IPO plans. After the $220M acquisition of Chef by Progress Software in September of 2020 (see my analysis), Perforce snapping up Puppet was little surprising. Like its former archrival Chef, Puppet had accumulated a massive amount of funding ($189M) with a painful last round of $40M in debt financing by BlackRock in 2020. In a market without excitement around configuration management and with the pole positions in related segments like infrastructure as code (IaC), Kubernetes automation, DevOps automation and GitOps already occupied by incumbents like Terraform, Red Hat Ansible, VMware Salt, Jenkins, Atlassian, Microsoft GitHub, and GitLab, there was nowhere left to go for Puppet. This is why I’m sure the investors asked (more or less impatiently) management to take the deal and call it a day.
HashiCorp Went IPO, while Puppet Had to Sell
On December 13, 2021, HashiCorp celebrated its successful IPO (see my analysis), raising $1,25B in cash and ending the day at a valuation of $15B. These are impressive numbers considering HashiCorp’s annual revenue of $212M at an expected annual run rate of $350M. How did HashiCorp achieve this massive valuation while Puppet had to scrap its own IPO a few months earlier that same year? Are these not both infrastructure automation vendors with similar financial ratios?
Three Key Differences between HashiCorp and Puppet
- HashiCorp managed to move 97 percent of its annual software revenue to subscription business.
- HashiCorp’s Terraform has developed into today’s de-facto standard for infrastructure as code (IaC).
- HashiCorp Consul and Vault have shown fast growth in today’s high-growth disciplines of container networking and container security.
Puppet Missed the Kubernetes Train
Today, Puppet runs on Kubernetes. However, it took the company too long to adjust to this Kubernetes-dominated reality. We have to remember that Puppet had been evangelizing a declarative approach toward IT infrastructure automation from the company’s very beginnings in 2005. When Kubernetes emerged in 2014/2015, the company was very close to its envisioned IPO and doubled down on its own declarative story instead of embracing Kubernetes. Thinking back to many conversations with Puppet in 2015/2016 I remember the company describing Kubernetes as “one of many” deployment models playing at the infrastructure level only. When it became clear that Kubernetes would become the dominating application platform with its own vast ecosystem of tools for deployment and management of complete application stacks on different clouds, the air became much thinner for Chef and Puppet. Organizations were looking for integrated management platforms that would help them tame Kubernetes complexity from a developer’s and an operator’s perspective alike. At the same time HashiCorp successfully positioned Terraform as the infrastructure as code (IaC) platform that would lift Kubernetes to hybrid and multi-cloud environments. The combination of Docker containers, Kubernetes scheduling, Terraform IaC, Jenkins for pipeline automation, and GitHub for code management and ultimately GitOps became the new standard. All of a sudden, both Chef and Puppet were isolated and scrambling to integrate their tools into a consistent platform. Due to the vast user base of their legacy products, this was not easy. Talking about exciting topics such as “policy as code” or “application automation” is difficult when tens of thousands of your open source users merely interested in enhanced infrastructure automation and orchestration features while CIOs are already starting to jump on Docker containers, Kubernetes scheduling, and serverless public cloud services. Like Docker (my analysis from 2019), Puppet had missed the Kubernetes train and would never be able to catch up.
Perforce Picks Up 40,000 Open Source Users and a Mature Automation Platform
Minneapolis-based Perforce has approximately 1,200 employees and a large portfolio of developer and DevOps productivity tools. The Puppet automation platform seems like a good fit to tie together all of these offerings into a unified DevOps play. Puppet could become Perforce’s underlying automation platform for deploying, operating, and scaling the hybrid cloud infrastructure required for agile software development. Combining automation and DevOps tooling could help Puppet accelerate the conversion of its 40,000 open source users into commercial enterprise customers.
Final Thoughts
While this analysis was easy to write with the benefit of hindsight, I do not know if all of this was preventable. Sure, Puppet’s hesitation to embrace Kubernetes and the company’s at times confusing marketing strategy did not help, but in 2015 Puppet had too much history to just rebuilt its portfolio into a Terraform-like container-centric IaC play. Remember that in 2015 it was still unclear if other big players in the market would throw their weight behind Kubernetes. Kubernetes was still miles away from true production readiness and many customers were still nursing black eyes from their OpenStack adventures.
All that said, please regard this article as my perspective on what happened. It is not at all a judgement of Puppet’s decisions and strategies. Things could have turned out differently and instead of seeing $15B IPOs of Terraform and GitLab a few months back, we could have seen similar IPOs of Chef and Puppet.
Addendum: The Public Reaction — April 15, 2022
Many people I have known for a long time were genuinely surprised about my interest in this acquisition. “I wouldn’t have known about the acquisition, had you not talked about it,” was the tenor of many of them. Nobody was surprised that Puppet was acquired and it generally seems like people just have moved on. But what was surprising was the quantity of opinions I received over the week, including today (April 15, 2022). Many people, including myself, remember the days when Puppet and Chef were still at the top of the world and when all the cool kids had Puppet or Chef stickers on their server-grade laptops. Therefore, I wanted to capture the conversation on LinkedIn about this topic, to get a complete overview of people’s thoughts.
The topic map shows 4 main themes, all of which I answered in the above article:
- Puppet (Orange): People still closely relate Puppet to Chef and to a lesser degree to Jenkins, Ansible, Terraform, and GitHub. One of the key questions I received this week was about the comparison between Terraform, Ansible, and Puppet.
- The Acquisition itself (Green): LinkedIn users where wondering about the strategic fit and the risk of this acquisition, considering that a 1,100 people company just acquired a 500 people one. How will this change both players’ position in the market-place.
- Perforce (Red): Questions about the fit between Puppet and the Perforce Perfecto test automation platform came up a few times. People where speculating about the potential repositioning of Perfecto in the market place and also about the cross pollination between Perfecto and Puppet.
- Developer Impact (Purple): The fourth and final topic on this map is the impact of the acquisition on the daily life of software developers. This goes in line with people in the Puppet Slack asking about how the acquisition would affect the Puppet open source repository. These people were immediately reassured of Perforce’s commitment to support the open source community.
