Software has been eating the world for at least the last decade, if not more. Software’s role in enterprises has evolved from merely a functional tool to the primary means of engaging with customers, managing day-to-day operations and driving actionable business insights. In keeping pace with the changing role of software in enterprises, the process of software delivery has also materially evolved: from waterfall (1970s to early 2000s) to Agile 1.0 (late 2000s to mid 2010s) to DevOps (mid 2010s to present).

The key evolution has been that software delivery has gotten much faster as it rapidly incorporates innovations and customer feedback — shipping software used to take several months and now it takes just days. Software developers (‘Devs’) and IT operations (‘Ops’) used to be in their silos but now form part of the same collaborative team to ship software rapidly (hence the term, ‘DevOps’).

What is DevOps?

DevOps is the latest evolution of software delivery that integrates software development and operations — shortening, automating, and improving the software build and release workflow. DevOps is a combination of cultural and organizational shift towards collaboration, automation and measurement that enables a continuous, fast, and secure software release cycle. The chart below illustrates the key practises of DevOps:

Impact of DevOps

DevOps has become a potential source of competitive advantage. As demonstrated in the table below from The Phoenix Project, tangible benefits of DevOps include faster rollout of new products & features, faster customer feedback, more developer time for core tasks such as writing new code and higher code reliability.

Current State of DevOps

Given DevOps importance for enterprises in the current era of digital transformation, it is a top-of-mind priority for management and engineering leaders. As per Bain & Co, 90% of the companies regard DevOps as a top strategic priority. However, over 80% find it ‘somewhat challenging’ or ‘very challenging’ to scale DevOps within their organizations. Key reasons for the challenge in scaling DevOps mentioned were 1) lack of skills/talent, 2) complexity of managing tools/environments and 3) culture resisting change.

Key DevOps Investment Theses

Thesis #1: The DevOps market is highly fragmented today but best-of-breed providers will emerge over time. Tech-first enterprises will likely gravitate towards best-in-breed providers that integrate with each other while tech-enabled enterprises & SMBs will likely adopt platforms with end-to-end product suites

DevTools space is highly fragmented (see data from a UBS survey below) primarily due to a Cambrian explosion in DevTools targeting different segments of the software delivery process. There are certain platforms that offer end-to-end product suites (e.g., AWS, Azure/GitHub, Atlassian and GitLab) while certain DevTools solely focus on specific functionality (e.g., Harness for CI/CD, JFrog for artifact repository, Hashicorp’s Terraform for Infrastructure as Code). The debate is whether customers would gravitate towards end-to-end platforms or choose best-in-breed DevTools.

The benefits of using best-in-breed tools are a 1) rich & best-in-class feature set, 2) teams ability to choose their preferred tool set (“ask your developer” philosophy) and 3) avoiding vendor lock-ins.

However, the downsides include 1) a wild-west environment where there is lack of consistency in software delivery practises across teams within the same enterprise, 2) more expensive, not only because of the cost of the software, but also because of the cost of integration and 3) not all companies require 100% of the feature set of best-in-breed tools, sometimes 70% of the features set works just fine.

As such, I hypothesize that large technology-first enterprises will gravitate towards a limited set of best-in-breed tools as chosen by their ‘Centre of Excellence’ teams and spend resources in integrating these tools in their tech stack & delivery pipelines. Meanwhile technology-enabled enterprises and SMBs may gravitate towards the end-to-end product suite of DevTools offered by platforms such as Azure/GitHub, AWS and GitLab with the benefits being much tighter integration, lower cost and acceptable functionality.

Thesis #2: The DevOps market has numerous tailwinds and is sizable as a whole but the fragmentation of DevTools makes individual businesses not as large (e.g., very few >$10bn businesses). I expect largest businesses to be the best-in-breed solutions in large categories such as Plan (Atlassian), SCM + CI/CD (leading contender: GitLab) and observability (leading contender: DataDog)

As per IDC, the DevTools market is expected to grow at c. 15% p.a., with cloud-based tools growing at 21% p.a., from $12bn in 2020 to $25bn by 2025 primarily driven by 1) digital transformation being top strategic priority, 2) high cost of software developers & DevTools impact on improving developer productivity, 3) need to ship software fast, 4) increasing complexity of software applications driven by microservices architecture and multi-cloud & hybrid cloud deployments and 5) increasing need for online collaboration given trends related to remote work.

Despite a sizable market, the budgets for DevOps vendors have been on the smaller side versus other enterprise software giants such as Salesforce, WorkDay and SAP, where ACVs run in the millions of dollars. As per a survey conducted by Redpoint Ventures, >75% of enterprises have <$1m budget for 3rd party DevOps vendors and majority expect it to increase by 10% over the next 5 years. I believe that the low spend is primarily driven by fragmentation of the market as well as downward pricing pressure exerted by cheap bundled tools offered by the hyperscalers as well as cheap open source alternatives in certain categories.

I believe that DevTools that are able to gain dominant market share of certain large categories (such as Atlassian in Plan/Collaboration) and/or able to execute a platform strategy should be able to drive higher ACVs. Based on the same Redpoint survey, the large categories include observability (leader contenders: DatDog), CI/CD (leader contenders: GitLab, Harness), API dev & management (leader contenders: Postman).

Thesis #3: Typically, DevTools have low moats given constant technological innovation. However, certain categories of DevTools have moats primarily driven by heavy integrations with the enterprises’ tech stack & delivery pipelines resulting in customer stickiness

DevTools market is highly fragmented and the tools have short product cycles. Leading cause of short product cycles is the constant technological innovation which makes existing tools obsolete with the customers often churning to a newer, more innovative tools. Developers are also known to be more receptive to trying out the latest trending tools.

I believe that best-of-breed DevTools in more stabilized verticals of the DevOps toolchain with high integrations in the company tech stack/delivery pipelines and/or strong ecosystems have considerable moats.

As a case study, let’s examine the moats of Atlassian’s Jira product that has made Atlassian into a $68 billion market cap company (as of 5 Aug 2022):

• Planning & collaboration being a stabilized vertical with low technological innovation: This may change if the software delivery philosophy moves away from Agile to something new — does not seem imminent
• Jira being a best-of-breed product: Jira Software allows for heavy customizations, security, compliance and permissioning that are needed by enterprises with large dev teams
• Customer stickiness enabled by the Atlassian ecosystem & heavy integrations with the enterprise’s tech stack/delivery pipelines: Perhaps most importantly, Jira has a robust ecosystem enabled by Atlassian Marketplace and enterprise customers tend to use several of the bespoke features available on the marketplace. Jira also integrates with the rest of the customer’s tech stack such as their dedicated solutions for SCM, CI/CD pipelines, artifact repository, observability, etc. Ripping out Jira and replacing it with something new is not a hassle that most CTOs are willing to go through unless there are heavy $ savings (which is not the case as Jira is very reasonably priced at ~$5 per dev per month)

Thesis #4: ML will permeate DevOps resulting in higher developer productivity. This will allow DevTools that integrate & utilize ML effectively to achieve higher differentiation and emerge as best-in-breed

Machine learning is permeating several industries and software development is not left untouched. ML models incorporated in DevTools will help developers in software delivery by generating source code, optimizing test runs, patch vulnerabilities, rolling back code changes, etc. The artificial intelligence R&D team at Facebook / Meta believes that over time, programming will become “a semiautomated task in which humans express higher-level ideas and detailed implementation is done by the computers themselves.” As an example, GitHub Copilot is able to provide predictive single line and full code snippet recommendations, thereby removing the need to rewrite repetitive code and allowing reuse of undifferentiated functions across-teams. DevTools that integrate and utilize machine learning effectively will achieve higher differentiation versus competitors allowing them to potentially emerge as best-in-breed in their respective categories.

Global DevOps Market Map

As mentioned previously, the DevTools space is highly fragmented. The DevOps market map from the Sapphire Ventures, reproduced below, makes a great attempt at highlighting the key DevOps companies globally:

DevOps in Asia

Asia is well-positioned to be the home for several unicorn DevTool companies. Atlassian, the bellwether in the space, originated from Asia and continues to inspire the next generation of developer-focused product-led companies.

In the Part 2 of this series, I will cover the key drivers that make Asia an attractive home for DevTool companies and the Asia-focused market map. Stay tuned!