As IT organizations attempt wide-scale cloud adoption, the importance of common best practices across applications and products is growing, sparking an exciting new conversation about platform teams and related disciplines like platform engineering.\n\nThe problem statement driving the investment in platform teams is clear: developing, operating, and optimizing a modern application is becoming too complex for many product delivery teams to solve independently. In response to this friction, leading organizations are taking a new approach, allowing workstreams following a given application pattern\u2014perhaps a Java microservice or Kubeflow data pipeline\u2014to use a repeatable, secure set of starters, UX, and automation\u2014a \u201cgolden path.\u201d\n\nThe game-changing insight of golden paths is an application-centric and workstream-focused approach. From IDE to production, golden paths align development teams with an organization\u2019s cross-functional best practices. The correct way to work becomes the easy way. As a result, many platform teams deliver a more than 50% improvement in developer onboarding speed, eliminating friction and uncertainty for all the cloud applications they enable.\n\nRepeatability is fundamental to realizing value from the cloud at scale. Before adopting a golden path culture, autonomous application teams might share version control systems or continuous integration tools but deploy and update similar applications in needlessly variable ways. Golden paths propose a new level of standardization, implementing internal developer platforms (IDPs) and shared workflows, which guide an application from its first commit to Day 2 operations.\n\nEffectively diffusing cloud best practices requires both a technological and cultural evolution. Many cloud application teams are entirely disconnected; other groups may have no repeatable access to their hard-won cloud and application expertise. The shared patterns and workflows on internal developer platforms become a nexus of repeatable standards, continuously improved by updates from across the organization.\n\nSpotify experienced a lack of repeatability and responded with a significant update to its cloud strategy. They shifted their culture to center around a collaborative set of best practices and workflows as a baseline for every cloud application\u2013golden paths enabled by their platform team. The results proved so decisive they wondered how they operated without them, recalling: \u201cThere was a time when engineers at Spotify couldn\u2019t imagine life with golden paths; now we can\u2019t imagine life without them.\u201d\n\nBuilding a cross-Functional Golden Path\n\nPlatform teams building golden paths codify and automate the application patterns across architectural, operational, and security domains. Comprised of experts in cloud architecture, DevOps, security, and automation, platform teams work closely with application development teams to enable an end-to-end experience.\n\nMany application teams are building similar cloud applications but need a cultural prompt and shared platform capabilities to collaborate on architectural best practices. A platform team's first task is to enable simplified creation and deployment of an organization\u2019s most common application types\u2013preferably with a simplified application deployment manifest.\n\nCommon and simplified deployment manifests unlock a powerful new way of thinking of cloud applications as fleets vs. one-off projects. Applications with a tolerance for horizontal scaling employ an easy-to-automate and consistent approach to secure application networking and autoscaling. New security affordances and controls may also be cumbersome to implement without a shared and developer-friendly platform but become the easy-to-consume defaults with one.\n\nMore secure by default often becomes a top focus for golden paths. Platform teams should use common automation for container builds, removing developer toil and allowing ongoing Day 2 updates. Security posture and vulnerabilities should be holistically tracked and exposed with an intuitive UX by the platform team. More secure infrastructure configurations can be automated by the platform approach such as enabling mTLS for application traffic by default.\n\nIf this new approach sounds right for your cloud strategy, here are some aspirational outcomes and metrics proven to help guide and motivate a platform team investment:\n\n1: Ease developer onboarding: When a new engineer joins the team, it's a living test of your cloud productivity experience. Spotify recorded a rough halving of their total time to value for new hires with golden paths. What\u2019s easier to learn the first time is also often far easier to repeat reliably\u2013making this a great first metric to focus on for a platform team.\n\n2: Reduce inefficient manual tickets: While the promise of cloud-based delivery is on-demand automation through declarative APIs, the reality is often many layers of approval tickets and forms. Golden paths bake in policies and best practices by default. Writing code, not tickets, is a rallying cry for successful golden paths in enterprises. Reducing the number of tickets to initiate and deploy an app by more than 50 percent is a common starting goal.\n\n3: Automate Day 2 operations: Day 2 operations on applications and the platform infrastructure become more automated and frequent. Successful platform teams often rebuild their entire estate, including security updates, weekly. Golden paths also leverage more secure-by-default automation whenever possible, relieving the burden on developers and improving security posture. \n\nThis unique combination of outcomes across developer velocity, day 2 operations, and security is powering an exciting industry shift in cloud consumption. As Gartner analyst Mark O'Neill recently observed: \u201cLooking at our inquiry trends on platform engineering\u2026 What a rocket ship of a topic.\u201d\n\nExpect golden paths and platform engineering to remain one of the most important trends as our industry grows from early cloud platform explorations to repeatable execution at scale.