Red Hat's Hidden Treasure: The People Worth Billions to IBM
At $34 billion, IBM has pulled the biggest software M&A ever, paying Red Hat shareholders a $14 billion premium in cash, in a bid to dominate the hybrid-cloud market. Avalia, a firm specialised in providing software due diligence for mergers and acquisitions, has analysed one of Red Hat’s fastest growing products and, in the process, gained insights about Red Hat’s strategic value, the risks in the deal, and the actions IBM should take to ensure it succeeds.
While discussions about the deal have focused on Red Hat products that use open-source technologies that have dominated the market, such as Linux (RHEL), JBoss, and Kubernetes (OpenShift), we were interested in Ansible, their IT automation software, which is fighting for dominance against several other open-source options, most with a longer history.
We start by discussing the business relevance of Ansible to IBM, including its role in hybrid-clouds and why it is a critical element in high-value deals, IBM’s target segment.
We then analyse software development data from sources such as GitHub - used to manage Ansible’s code, development activities, and developer community - to gain insights into how their people, processes, and the product impact the business value they deliver to their clients.
In our assessments, we usually interview the people creating, deploying, and supporting the software in question, but we have not interviewed anyone from Red Hat in this case, and recommend that the reader keep that in mind when viewing this report.
However, the data speaks for itself, providing a technological perspective into what is behind IBM’s billionaire bet.
IBM targets high-value deals with clients that have on average 1’000 different applications. These clients typically have 5 to 16 different clouds, but only 20% of their applications are running in the cloud.
Expanding the benefits of cloud computing to the remaining 80% is a greater challenge with a variety of constraints. Clients want to avoid vendor lock-ins, and want to ensure their applications and data can flow securely across their in-house and cloud infrastructures, managing these hybrid-clouds as if they were one single platform.
With the acquisition of Red Hat, IBM gains power in fundamental cloud technologies. Red Hat is a leading open source contributor with influence in key hybrid-cloud components such as Linux operating systems, OpenStack cloud platforms, Kubernetes cloud orchestration, and Ansible IT management automation.
We have chosen Ansible for our software due diligence because cloud automation is still an emerging technology, and if Ansible becomes the dominating cloud IT automation software, IBM would have a significant competitive advantage in shaping its roadmap, and winning their target high value projects.
This graph ranks the all-time top 15 people contributing to Ansible based on the number of files they have modified (for more information about them check Annex 1: Ansible’s hall of fame). These contributors have authored as much of Ansible’s code as all the other 4’600 developers combined, even though code is not the only kind of contribution.
Data comes from GitHub, where Ansible’s project files and source code is kept. The green bars run from the first to the last contribution dates and are ordered according to the number of files modified, with colour highlighting the difference in values.
Red Hat Ansible development leaders are almost all still there (source: GitHub):
Straight away we see that 13 of the top contributors are still active in the project, their bars stretching all the way to the end. Five of them joined before 2014, when Ansible was starting, preserving essential tacit knowledge and technical leadership that make it easier to onboard contributors and grow the community.
A search on LinkedIn showed that Red Hat employs 12 of the top 15 contributors, retaining all but the co-founder of Ansible after its acquisition in 2015. These numbers indicate Red Hat has a positive corporate culture for technical talents.
Most development leaders of competing projects are not contributing anymore (source: GitHub):
Open source competitors Puppet, Chef, Salt, and Terraform show a different picture, with many of their top developers no longer contributing. While in Ansible we see a growing core team of top contributors, in the others we see several drop-outs.
A large and stable core team indicates strong leadership in the open source community, helping it to grow and become the dominant technology. This data shows Ansible has an advantage in its leadership ranks, bringing us to the next question: how does their army of developers compare against the others?
Onboarding and retaining developers is a critical process in a successful open source project. Ansible’s has an active “contributor experience” project with 54 people involved in improving their community’s experience. From having bots automatically saying thank you to contributors, to providing better templates, and asking first time contributors about their experiences, the care and effort invested in the community provides outstanding results.
Ansibles’ open-source community is diverse and growing (source: github):
In this graph, each circle represents a person’s contribution (i.e. commit) in a quarter. Circle sizes indicate how many contributions each person has made in the period, with their share of the total appearing in %, and the top 5 contributors highlighted with colours.
Between the 3Q/2015, right before Ansible was acquired by Red Hat, and 3Q 2018, the number of active contributors grew from 330 to 490 people, with the top 5’s share moving from 64% to 20% of the contributions, and while the top 15 developers changed as many lines of code as all the other 475 combined, LinkedIn data shows that while 9 of them work for Red Hat, the other 6 work for other companies, including hybrid-cloud vendors Cisco, Google, and Microsoft.
Competing communities are small and concentrated in the hands of a few (source: github):
Looking at the numbers from Puppet, Chef, Salt, and Terraform we see different results. These projects reliance on their top contributors is much higher. Over 50% of the contributions are made by 5 people in Puppet and Salt, 2 people in Chef, and only 1 person in Terraform.
Their whole communities of active developers put together amounts to half of Ansible’s in the same period. Our forecast, based on data from the past 3 years, estimates Ansible’s lead to widen, consolidating its position as the cloud automation largest open-source community.
We forecast Ansible’s community to remain larger than all the others combined.
Ansible’s open source community have created a high quality software. We have applied an automated tool to analyse all of Ansible’s 908 thousand lines of source code in search for potential bugs, vulnerabilities, and poor coding practices (code smells), as well as to estimate the effort required to fix them, expressed in number of days of “technical debt”. An expert interpretation of this data is essential to understand the business impact of these metrics. Dr. Olivier Liechti, Avalia’s CTO, had the following comments about Ansible’s code analysis:
“The number of bugs and vulnerabilities is low in relation to the project’s size. What the tool has flagged as bugs and vulnerabilities are choices of coding style that differ from a specific standard, but bear no risk. However, an area that deserves the attention of Ansible’s technical team is code complexity. Complex code increases support time and costs. Maintaining the code, as well as introducing new features, also becomes more expensive as complexity grows.”
See more data here.
A deeper look shows that the technical debt associated with code complexity is of 155 days, or 80% of the total. As a reference, Puppet and Chef, while having smaller code bases of 260K and 185K lines of code respectively, only have around 3 days of technical debt related to code complexity each. This is an area where Ansible can and should improve.
Google searches for Ansible are higher than all competitors combined (source: google trends):
Despite its code complexity Ansible stands out for being easy to use. According to surveys in Gartner Peer Insights, Ansible’s main differentiator is its ease of deployment. Google data shows its growth in search popularity while others decline, with the exception of Terraform.
However, Terraform’s active developer community shrunk significantly in 2017, and is over 10x smaller than Ansible’s, raising questions about its ability to keep up with its product development. Ultimately, Ansible holds a very privileged position in the market, and is likely to hold on to it.
Avalia Systems software due diligence summary and recommended actions:
IBM’s CEO Ginni Rometty aims to add over $8 billion in sales over 5 years with Red Hat’s deal. She says IBM will become the number 1 provider in hybrid-clouds, thanks to Red Hat’s open-source culture, developer community, and technologies. Our software due diligence of Ansible, Red Hat’s IT automation product, showed that IBM might be getting what they paid for.
Taking Ansible as a reflection of Red Hat’s open-source culture, community, and technologies, IBM is making an excellent acquisition. Ansible’s core leadership belongs to Red Hat, and their care has translated into a growing and diversified community of contributors, that in turn have built a great quality software. With this acquisition, comes great power and great responsibility with Red Hat’s developer community, and many questions still need to be answered.
Will a shock of cultures drive Red Hat’s technical talents away? Will IBM try to cut costs and lessen the resources invested in supporting open-source communities? Will they change the transparent and inclusive governance we saw in Ansible, or could their big business image, encourage some to fork out and compete with IBM using Red Hat’s own technology? Will IBM invest in making the code less complex, despite the profit opportunities that could come from having a dominating technology that requires the consultancy of experts to be properly used?
The biggest risks to the deal’s success might come from IBM themselves. No wonder Ms. Rometty will keep Red Hat as an independent company, despite the billions she used to buy it.