Platform Engineering in 2025: Beyond Backstage and the IDP Paradox

Platform engineering has become the new DevOps—the organizing principle for how organizations scale their infrastructure practices. At the center of this movement are Internal Developer Portals (IDPs), promising to unify developer experience across fragmented toolchains. But after years of Backstage implementations, the industry is learning hard lessons about what portals can and cannot deliver.

The Portal Paradox

Backstage, Spotify's open-source developer portal, has captured remarkable market share—by some estimates, 80-90% of IDP implementations. It's a CNCF incubating project with 2 million developers across 3,400+ organizations. The adoption curve looks like success.

But there's a paradox. Many organizations that spent 12-18 months implementing Backstage report adoption rates stalling around 9%. The portal exists, but developers aren't using it. Thoughtworks and other consultancies are seeing this pattern repeatedly across enterprises.

Many teams mistake the portal for the entire platform, but Backstage is not your platform. Portals are the front end of the platform, a UI that allows developers to discover and access the platform's underlying capabilities.

What Actually Drives Developer Productivity

The research is clear on what slows developers down: on average, development teams use 7.4 tools, forcing constant context switching. 75% of developers lose 6-15 hours weekly to tool sprawl. For a team of 50 engineers, that's nearly $1 million in lost productivity annually.

A portal addresses discoverability—knowing what tools exist. But it doesn't solve the underlying fragmentation. If developers still need to switch between 7 tools, knowing about them from a central catalog doesn't reduce the cognitive load.

What Successful Platforms Provide

The Shift to Commercial Solutions

Gartner's 2025 Market Guide for Internal Developer Portals signals a shift: the market now favors IDPs that deliver immediate value and faster ROI. Commercial offerings like Cortex, Port, and Humanitec are gaining ground by providing opinionated solutions rather than frameworks to customize.

This isn't surprising. Backstage's strength—extreme flexibility—is also its weakness. Organizations must build plugins, define entity schemas, and create service catalogs from scratch. That's a significant investment before any developer sees value.

Principles for 2025

Platform Engineering Success Factors

• The platform exists to make developers faster, not to look cool
• Every feature must reduce cognitive load or it gets deleted
• If a human approves it in Slack, automate it
• Paved roads are mandatory; off-roading is opt-in only
• Success is measured by how little developers think about infrastructure

The industry is converging on a "vending machine" model for infrastructure. Developers select from curated options that are standardized and optimized for the enterprise. The platform team maintains these options; developers consume them without understanding the implementation details.

Gartner's Forecast

The numbers are striking: Gartner predicts that by 2028, 85% of organizations with platform engineering teams will have internal developer portals—up from 60% in 2025. Platform engineering has been named a top strategic technology trend for both 2023 and 2024.

But the same forecast warns that 80% of enterprises will have some platform engineering initiative by 2026. "Initiative" doesn't mean success. The organizations that will thrive are those treating the platform as a product, with dedicated teams, user research, and continuous improvement.

Our Perspective

Having worked with government agencies where developer experience is often an afterthought, I see platform engineering as genuinely transformative—when done right. The failure mode is treating it as an infrastructure project rather than a product initiative.

My recommendation for organizations starting this journey: don't begin with Backstage. Begin by identifying the three things that most slow your developers down. Solve those problems—even if the solution is a script and a Confluence page. Once you understand what developers actually need, then evaluate whether a portal adds value.