OpsLevel vs Backstage: What's the Best Internal Developer Portal?

Since it was open sourced in 2020 Backstage.io—a platform for building internal developer portals (IDPs) initially created for internal use at Spotify—has generated significant buzz in the engineering community. While Backstage wasn’t the first IDP, it did much to raise awareness for the entire space, bolstering business for commercial entities that offered out of the box alternatives. One such entity is OpsLevel. OpsLevel formed in 2018 to tackle a growing need for service catalogs, but has since evolved to offer more traditional IDP capabilities like scoring and scaffolding. In this blog we’ll cover:

• How OpsLevel compares to Backstage
• The pros and cons of each according to your goals
• Why you might consider a third alternative

While the developer portal and developer platform space is starting to get crowded, and confusing—with some established players like DataDog or Atlassian offering catalogs or pared down versions of an IDP as a feature rater than a standalone product—it wouldn’t be unusual for a team to compare OpsLevel and Backstage directly. The two have many similarities, but also differ:

At a glance, Backstage and OpsLevel:

• Both provide ways to centralize service information
• Both offer a mechanism to score services against certain standards of development
• But OpsLevel requires fewer resources to deploy, with more out of the box functionality
• While Backstage offers a greater level of flexibility with a robust plugin framework and near limitless optionality for building the portal that best fits your team’s needs

OpsLevel: The Developer Platform for High Performing Teams

OpsLevel is a prescriptive internal developer portal that enables engineering teams to catalog, measure, and scaffold software according to best practice. They boast the ability to build your developer portal in minutes rather than months, which cannot be said about Backstage given the need to architect every component from scratch.

OpsLevel is a good budget-conscious option for teams that are primarily struggling with service visibility, only need to lean on a small set of required web-hooks, and are comfortable with a single rubric for different types of scorecards.

It’s not best for teams that are rounding the corner to greater software and standards complexity, including those with certain data portability and hosting requirements.

Features & Benefits of OpsLevel

Easy to setup

• Catalog all your microservices, systems, teams, and tools
• Set standards of development and maintenance
• Automate checks for alignment to standards
• Execute actions from the platform

Checks some boxes for service catalogs

• Detect services across your stack and add to your catalog
• Add service descriptions with AI
• Show dependencies between services

Enables standards tracking

• Add rules to different tiers for security and reliability use cases
• Use a single rubric for all use cases (can’t customize rubrics for different use cases)
• Create a one-off campaign to drive change (can’t be re-used)

Developer self-service and output optimization

• Centralize view of tools supported by web-hooks
• Create templates that can be used when deploying software

What Are The Drawbacks of OpsLevel?

There are a few things to consider when evaluating OpsLevel, some of which are tied to your org size, maturity, and complexity.

Poor ratings for user experience and adoption

• User experience has been called “clunky”
• Clunky UI has led to historically poor developer adoption, casting doubt on case study claims of 100% adoption from the entire engineering team in under 30 days.

Limited supported data types

One the reasons why OpsLevel can tout rapid setup has to do with a rigid data model which restricts the types of content you can import, and the relationships those entities can hold

• OpsLevel only supports service catalogs rather than other software types like ML models, repositories, or Kubernetes clusters. Other entities must be encoded as “services" or “teams.”
• OpsLevel also doesn’t support full integrations, only web-hooks for tools like GitLab and Jenkins which require you to do a bit more work to get your entire ecosystem of tools connected to push rather than pull data
• Users also cannot pull in custom data from internal or other third-party tools

Limited automation

• According to their blog, OpsLevel believes auto-syncing service ownership from identity providers is poor practice, thus requiring you to manually add groups and update when people join, leave, or move teams.
• Manage team membership manually through UI, API, YAML, or Terraform only.
• While OpsLevel talks about continuous checks, Scorecard evaluations actually aren’t automated which means checks must be manually created and run

Inflexible scorecarding

• While users can create different Scorecard types, they utilize a “global rubric” which means only only rule format is supported‍
• Scorecard rules don’t support regex, which prevents teams from creating non-case-sensitive checks for the presence of certain directories

Limited Developer Experience capabilities

• Developer portals should enable customization of user experience to suit developer workflows. OpsLevel does not allow you to recongfigure the layout or visible modules of homepages or entity pages to suit user needs
• No prioritization in the homepage. OpsLevel’s homepage for each developer contains all the content and context they need, but it’s not ordered in a way that enables them to take action quickly.‍
• Alerting isn’t targeted, which can be noisy for developers

Inflexible reporting, and no dev metrics

• According to their documentation, OpsLevel enables users to report by services or teams, but not infrastructure, resources, product lines, project groupings, etc.
• OpsLevel doesn’t provide any reporting relating to productivity or performance by team or individual, which means software health might only be half the story.
• Slack integrations are “pull” only, requiring users to ask questions of the integration to get necessary information

OpsLevel Pros & Cons

OpsLevel has some unique features, though lacks a lot of table stakes functionality of teams that need a certain level of flexibility, automation, and customization to abstract away overhead for managers and devs alike.

Pros

• Easy to get going quickly
• Comprehensive documentation
• Clear visibility of services with auto-suggested descriptions where lacking

Cons

• Limited data and catalog types (services and teams)
• Webhooks instead of supported integrations
• Inflexible scorecard rubrics
• Requires manual ownership updates
• No customization
• No targeted or prioritized alerting for devs
• Small customer base primarily in mid-market

Spotify Backstage: An Open Platform for Building Developer Portals

Backstage is arguably one (if not the most) talked-about internal developer portals. Originating within the engineering team at Spotify and open-sourced in 2020, it’s become a magnet for forward-thinking engineering teams that want to enable developer self-service via a build-your-own portal experience. Built on top of a flexible plugin architecture, Backstage offers more extensive customization capabilities compared to other IDPs like OpsLevel.

As an open-source project backstage is of course free of any commercial obligations, though they do offer plugin packages with support options at a set cost, typically $100/per dev/per month. Backstage can be a good investment for organizations that have the resources to dedicate wholly to setup, maintenance, UI building, plugin creation, and integration development. We’ve seen anywhere from 3-15 FTEs staffed on Backstage projects across the mid-market and enterprise organizations we speak to.

However, the “open” framework that can be so enticing to developers carries hefty time, skill, and cost considerations. Now three years into its history, organizations who’ve tried Backstage have begun to realize the previously obscured cost of building and maintaining a fully custom internal developer portal. Many organizations either spend months trying to stand it up, or at the other end of the spectrum, have great success in doing so and realize too late the amount of overhead required to maintain all of the functionality they’ve built.

Features & Benefits of Backstage

Opensource

• Backstage is an opensource project which means it’s an open framework for development, and can be continuously improved by the community that supports it—not just those at Spotify.
• Opensource also means no licensing costs

Large community

• Connect, share, and get help from users all over the world

Great documentation

• Backstage may require enormous effort to set up and maintain but you’ll have no shortage of documentation on how to begin the process

Portable

• No fear of vendor-lock-in
• Supports techdocs or docs-as-code

Developer Experience and platform engineering focus

• Tools to build developer self-service and pipeline monitoring solutions

What Are the Drawbacks of Backstage?

Low developer adoption

• While Spotify’s internal adoption of Backstage is high, Spotify’s own VP of Engineering has publicly shared that adoption at external organizations averages about 10%.

• No task prioritization for devs means no way to weed through the noise they may receive when alerting isn’t targeted or elevated by business impact.

Exceptional effort to stand up and maintain

• Effort required is the top complaint from Backstage-curious development teams, and is referenced many times across Reddit forums.
• It’s not an extension of existing efforts, it’s a completely new architecture. Build all software catalogs, integrations, and plugins from scratch.
• There are no out of the box integrations, and while a plugin marketplace exists, you may have other use cases you’ll want to build for. You may also be unlikely to find developers that would enjoy taking a full-time role creating integrations or plugins.

Requires skills that may be in short supply

• Backstage support may require you to hire for skills that don’t already exist on your team, or may be in short supply. This includes individuals familiar with: SAML/OpenID authentication, Authorization, REACT, Typescript as well as specialists for every integration and plugin you’d want to create.
• You’ll need to dedicate frontend engineers to build any workflows, dashboards, or reporting you may want
• Once you’ve moved past initial implementation just maintaining Backstage requires extraordinary effort—with teams alotting anywhere from 3 to 15 FTEs depending on the size of the engineering team they need to support.

Limited scoring capability

• Backstage offers a paid plugin package that includes scorecards, at 2-3x the cost of full commercial SaaS IDP like OpsLevel.
• Their scorecarding functionality is quite limited, with a UI-based rules editor (no Git Ops or YAML support), and manual check running via API
• Scorecards also only three support three data types for rule creation: Source Control Management, GitHub, or metadata from integrations you build into your catalogs

Backstage Pros & Cons

Pros

• Opensource model for unlimited flexibility and portability
• Large supportive community
• Supports creation of any software catalog
• Supports software development best practice

Cons

• Compared to out-of-the-box offerings like OpsLevel, Backstage requires significant effort to build and maintain
• No out of the box integrations
• Costly Backstage plugin marketplace at 2-3x the cost of full IDP solutions
• Very rudimentary scoring as a paid plugin
• Requires Frontend resources with React and Typescript skills
• No support packages offered except through third-party agencies

Cortex: The Internal Developer Portal that cuts noise for developers with paved paths to production.

While the two are often compared, there’s an ocean between the opinionated nature of OpsLevel and a fully open platform like Backstage. On one of the spectrum setup is quick, but capabilities are limited. On the other, there is no limit to what you can build, but doing so may completely tank your engineering team’s productivity, with very little developer adoption to show for it.

That’s why organizations of all sizes are moving to Cortex. Like OpsLevel, Cortex began as a service catalog, but quickly accelerated development across the stack once realizing the deep connection between developer productivity, and software ownership, health, and velocity. Cortex was the first Internal Developer Portal to offer Scorecards—a now baseline feature for all IDPs with the exception of Backstage—but was also most recently to add engineering metrics as a flywheel for software health, enabling engineering leaders to track productivity metrics within the context that matters most for their interpretation and response.

Out-of-the-box capabilities with plenty of flexibility

• Cortex comes out of the box with all the features and functionality that your team would consider rote overhead.
• We offer 50+ fully supported integrations across the SDLC (with a full engineering team just dedicated to building more). Just add your API key and you’re good to go
• We provide a flexible framework for Scorecards, Initiatives (projects with deadlines), Scaffolding (building new software with software templates), and executive Reporting
• Our reporting can be filtered by more than just service or team—track progress by domain or product line to provide updates the rest of the C-suite can understand
• Get up and running in under 10 minutes with Cortex QuickStart. Just connect your Cloud provider, Git provider, and Identity provider with account number or API token.
• Extend your experience with the only plugin framework offered by a SaaS IDP. Visualize data from internal systems or as-yet unsupported tools you don’t want to leave behind, and slice views by role to reduce noise
• Bring in custom data using our API to create rules in scorecards or reports

Support for any catalog type

• Unlike OpsLevel, Cortex enables users to define and catalog any software type—from services and infrastructure, to systems, workflows, environments, and anything else your organization has already thought up in your existing information architecture.
• Don’t jam distinct entities into “service” or “resource” if that’s not how your organization typically thinks about them!
• If you’ve already begun building in Backstage, Cortex also offers a migration helper to ensure you can preserve what you’ve already built while transitioning over.

Automation for improved experience

• Auto-update ownership information from your identity provider, plus support for manual changes via UI, API, YAML, or Terraform (the latter of which is the only way you can update this information in OpsLevel).
• Continuous monitoring of alignment to Scorecard rules without the need to initiate manual assessments as with Backstage

True support for developer experience and adoption

• Prioritization in the Developer Homepage ensures no critical task goes unacknowledged, while smart-suggestions provide devs with ways to make an impact even if they have no more than 5 minutes to spare.
• Targeted alerting from long-term projects or short-term initiatives ensures devs only hear about the action items that are relevant to the software they own, when they need to know about it.
• Push and pull Slack integration serves up useful information for users, while also enabling them to ask questions of the platform.
• Dedicated customer success and customer engineering teams that provide guided education for platform use and adoption.

Customize your experience

• Developers are happier when context-switching is at a minimum, and new tools mirror experiences they’re already familiar with. Cortex enables users to customize the UI to better align with a developer’s jobs-to-be-done. Surface tabs and tools that are most relevant to your team.
• Any-rubric scorecards enable users to create scorecards for any set of standards—binary, fractional, or value-based using information from any supported integration or custom data. Add 3 tiers, 10 tiers, or no tiers, and mix and match rule types by use case.

First IDP to support engineering metrics

• Cortex recently launched Eng Intelligence which provides insight from jira, GitHub, GitLab, PagerDuty, and deployment data to provide a rich picture of developer productivity where it matters most—in context with critical information about software health, team composition, and on-going projects.
• Eng Intelligence automatically aggregates and contextualizes data enabling teams to improve resourcing, tailor improvement plans, and drive greater productivity without sacrificing developer happiness
• Connection with Scorecards enable teams to finally connect the dots between software health, and the negative or positive effects that has on team productivity. When testing practices are poor, for example, productivity on net-new projects will wane.

Largest number of customers

• Founded in 2019 Cortex is well-entrenched within the Internal Developer Portal conversation, with the largest number of customers around the world spanning the mid-market and enterprise segments.
• Customers like Outreach, Workday, and SolarWinds have found success in use cases ranging from production readiness alignment, to AWS migration support, and new product launch.

If you’re evaluating Backstage or OpsLevel we hope this guide has been helpful (and we hope you’ll consider a third alternative in Cortex). If you’re ready to dig in a little more, check out our self-guided tour, or connect with us to book a fully custom demo today!