Over the last decade, almost every function in the enterprise has seen an uplift in productivity thanks to transformations from data-driven, AI enabled analytics tools. These tools allowed teams to start measuring what once was hidden, allowing them to act based on data rather than on gut.

This has been the case in sales, where new technologies like “conversation intelligence” help teams win more deals; product, where teams can now understand user journeys in real time and make better decisions; marketing, where tools help employees boost conversion and measure ROI; and customer success, where teams can better delight and retain users. This overall shift transformed productivity and decision making so much that in all likelihood, organizations are never going back.

But there’s an outlier here: software engineering. While this business function has become more data driven in recent years, it still heavily relies on intuition and manual processes. Despite software engineering being one of the fastest-growing professions—development platform GitHub is adding more than 10 million developers a year, with 56 million total developers now on that platform–software delivery is still more of an art than a science.

DevOps Research and Assessment (DORA), a group that helps organizations boost their DevOps and organizational performance through data, pioneered critical metrics from the point of committing code to releasing code to production. These metrics include deployment frequency, lead time for changes, mean time to repair and change-failure rates. This is great. Yet it doesn’t account for the fact that so much of the development process happens before code is committed.

On top of that, the complexity of developing software is only increasing with the availability of more tools, more stages, higher disaggregation, more modules, more repositories, more changes and more developers. The need for better analysis, and for software developers to step up productivity amid all this complexity, has never been greater.

Engineering productivity is measured across code and process reliability, repeatability, and velocity. Software developers live in five or six different systems daily, which include their git system, project-management solution, CI/CD platform and monitoring solution. This has increased the complexity in delivering quality code while impacting the core tenets of engineering productivity. The processes around software development have also changed from monolithic to microservices, on-premise to cloud, and waterfall to agile. So, identifying software bottlenecks requires insights from across the various stages of the development process.

But until recently, engineering leaders were trying to shed light on team performance and project status using self-built spreadsheets and watercooler chats. Covid-19 and the transition to a work-from-home environment has made this situation even more tenuous, given the added difficulties of asynchronous work and managing remote teams.

And it’s not just individual developers who want more visibility across their releases; VPs of R&D and directors of engineering/team leaders feel the pain as well. In order to effectively run daily standups, managers need to manually gather data from multiple data sources that exist in different organizational silos. This consumes expensive engineering firepower and makes it very difficult to stay in front of delivery failures and find initiatives at-risk.

This is why we are excited to partner with Ori Keren and Dan Lines, the founders of LinearB*. The company is pioneering technology in a new category addressing exactly these issues for software development. Software Delivery Intelligence, as they have termed it, acts to increase the visibility and effectiveness of engineering teams and their processes. Ori and Dan experienced the problem first-hand as senior engineering leaders at CloudLock (acquired by Cisco), where they struggled with rapid growth across their technical organization and often found themselves making engineering decisions based on intuition versus data.

Git-based systems (e.g., GitHub) and project-management software (e.g., Jira or Clubhouse*) serve as the underlying systems of record for development teams, but they provide little context around where processes get stuck and where there are opportunities for continuous improvement. Specifically, LinearB provides a unified dashboard for developers to view, analyze and improve their code-to-commit processes. The solution integrates with core engineering systems and correlates git, project and release data to provide engineering teams with real-time project insights and team metrics that are commonly linked to production issues.

The technology is consistent with our firm’s focus over the last several years on investing in developer tools and other productivity technologies that make the software-development process more efficient. This thesis resulted in investments in Cypress.io* and LogRocket* for testing automation; Bridgecrew* for security automation; JFrog* for binary repository management; and Harness* for continuous delivery. Concurrently we’ve focused on data-driven optimization for other business functions, such as through our investments in Gong.io* provides for sales operations; Pendo* and Amplitude* for product management; and Gainsight* for customer success.

We are excited to partner with LinearB as it makes software engineering more data-driven and moves into its next chapter of growth.