2025 State of Internal Developer Portals

Our second annual report on the state of internal developer portals dives deeper into the challenges engineering teams face and why they look towards platform engineering best practices as a way to overcome these challenges.

Key takeaways

  • 75%

    Of developers lose between 6-15 hours weekly due to tool sprawl

  • 7.4

    Is the average number of tools used for everyday operational tasks

  • 78%

    Of engineering teams wait a day or more for SRE/DevOps assistance

  • 50%

    Of eng teams lack trust in the data quality of their central system of record

  • 94%

    Of developers experience dissatisfaction with self-service tools

  • 15%

    Of developers say standards are clearly defined across domains

Developer tool sprawl could be costing companies $1m in lost productivity each year 

Engineers rely on multiple tools for various use cases such as AppSec, code quality, and incident management. On average, development teams use 7.4 tools, forcing engineers to switch contexts all the time, creating information gaps and diminishing productivity. As a result of this tool sprawl, 75% of developers lose between 6-15 hours weekly. This inefficiency has a tangible cost: based on data from the US Bureau of Labor Statistics, for a team of 50 engineers, tool sprawl adds up to nearly $1 million in lost productivity every year.

Average: 6.7 hours
Time lost by developers due to the number of tools used (hours)

Engineering teams spend far too long waiting for SRE and DevOps assistance

One of the biggest frustrations for developers is the delay in getting help, while SREs and DevOps teams struggle with ticket overload. In 16% of cases, respondents admit a maximum wait time of a week or more, and just 22% say that on average they can get their issue resolved within one day. However, teams that have adopted an internal developer portal see this number rise to 30% thanks to better visibility and self-service capabilities, which reduce wait time and boost developer productivity. 

Average developer wait time for DevOps or SRE assistance

Developers aren’t satisfied with the self-service tools they’re using

Self-service tools streamline workflows, automate repetitive tasks, and empower developers to solve problems independently. They enable developers to complete tasks such as creating a cloud resource, scaffolding a new service or API, and creating a Kubernetes cluster, all without having to wait for assistance. When we asked developers how they handled operational tasks, 20% said they used an interface for self-service via a CI/CD tool, 19% said they used self-service via internal tools and 11% said using an internal developer portal. Other methods include provisioning and managing infrastructure through Infrastructure-as-Code tools, creating and managing cloud resources via cloud management tools, and using an API tool to discover, test and deploy APIs.

But we found that many of these tools are not up to scratch, with only 6% of respondents stating they were very satisfied with these tools. Many of these tools require in-depth knowledge of underlying infrastructure, come with an unintuitive interface and don’t allow for standards to be built-in. This negatively impacts developer engagement — with no way to ensure developers follow golden paths, these platforms force developers to resort back to TicketOps.

Satisfaction with self-service tools

Engineers need to be able to define standards and gain visibility into compliance

Only 15% of engineers believe they have clarity over the standards required across different domains. Meanwhile proportionally more engineering leaders express doubt on whether software complies with organizational standards than engineers. This highlights the need for domain owners to be able to better define, communicate, and enforce standards across their organization and for tools that can provide visibility into standards compliance at all levels.

Is it easy for developers to understand the standards required by different domain owners?

Engineers don't trust the data they rely on to build and ship with confidence

Only 3% of engineers feel the data quality of their metadata repository is completely trustworthy and 50% have doubts about its accuracy. Trust in metadata is crucial, as without it, developers are unlikely to rely on it and will turn to DevOps, SREs, or other colleagues for information. This reliance on institutional knowledge is something engineering teams are striving to move away from. The ideal would be to have a single source of truth for this metadata. 

This trust gap may stem from infrequent metadata updates: 53% of teams update their software asset metadata no more than once per week, potentially leading to outdated or inaccurate information that undermines reliability.

Trust in the data quality of central hub/resource metadata

…which is why internal developer portals are gaining popularity

There’s clearly a rise in popularity for using a portal to maintain metadata, with 53% of respondents stating that they use the open source portal Backstage, an internal developer portal like Port, or both. 

However, traditional methods like Confluence, Jira, CMDBs, or even spreadsheets are still relied upon — often because engineering leaders have invested years into tailoring them to fit their needs, built workflows around them, or simply feel resistant to the disruption of transitioning to modern alternatives.

Top central hubs/resources for maintaining metadata
No email required

Internal developer portals, the cornerstone of platform engineering, can help teams to overcome these challenges by:

  • Providing a consolidated software catalog for visibility and data reliability

  • Enabling engineers to use self-service actions such as creating new services, spinning up a new environment or performing Day-2 operations with golden paths and guardrails built-in

  • Enabling engineers and domain owners to define standards on any kind of resource, and using scorecards to check compliance

Starting with Port is simple, fast and free.

Let’s start