Luca BertonThe Internal Developer Portal Decision for Early-Stage B2B SaaS
When you're building B2B SaaS, your first developer portal choice happens when you don't have a dedicated platform team. Early-stage means limited engineering bandwidth — you need something that accelerates feature delivery, not slows it down with months of infrastructure setup. This guide compares Cortex and Port (the two managed SaaS IDPs) against the open-source Backstage option, with a specific lens on early-stage B2B SaaS constraints.
For Early-Stage B2B SaaS (Under 50 Engineers)
Short answer: Port is the better choice for most early-stage B2B SaaS teams. It launches in days, requires zero dedicated engineers, and its flexible data model adapts as your platform evolves. Cortex is stronger if you want aggressive service maturity standards from day one. Backstage is not practical unless you already have 2+ platform engineers to spare.
- Port wins on speed: Days to first dashboard. No infrastructure.
- Cortex wins on standards: Built-in scorecards force engineering excellence early. Better for regulated industries.
- Backstage loses on resources: Needs 1-2 FTE minimum — a luxury early-stage SaaS can't afford.
Backstage vs Cortex: Full Context
For context: Backstage is the open-source, build-it-yourself option — unlimited customisation and no vendor lock-in, but it needs 1-2 dedicated engineers to run. Cortex is the managed alternative, built around the strongest scorecards and auto-discovery for driving service maturity, live in days to weeks without a dedicated platform team. Where Port fits in between is covered in the full three-way comparison below.
Kubernetes Recipes
A practical guide for container orchestration and deployment by Grzegorz Stencel & Luca Berton (Apress).
Watch on Skillshare →Head-to-Head Comparison
| Feature | Backstage | Port | Cortex |
|---|---|---|---|
| Model | Open-source (CNCF) | SaaS | SaaS |
| Service catalogue | Core feature, YAML-defined | Flexible data model, UI-configured | Auto-discovery + manual |
| Scorecards | Via plugins (Soundcheck, etc.) | Built-in scorecards | Strong built-in scorecards |
| Self-service actions | Software templates | Self-service actions (powerful) | Scaffolder |
| Plugin ecosystem | 200+ community plugins | Growing integrations | Pre-built integrations |
| Customisation | Unlimited (it's code) | Configuration-driven | Configuration-driven |
| Setup effort | High (1-2 engineers ongoing) | Low (days to weeks) | Low (days to weeks) |
| Cost | Free + engineering time | Per-user SaaS pricing | Per-user SaaS pricing |
Backstage: The Open-Source Powerhouse
- Pros: Unlimited customisation, no vendor lock-in, massive community (CNCF Incubating), used by Spotify, Netflix, American Airlines, HP
- Cons: Requires dedicated engineering team (1-2 FTEs minimum), plugin quality varies, steep learning curve for React/TypeScript frontend customisation
- Best for: Large enterprises (500+ developers) with platform engineering teams that want full control over the developer experience
- Compliance angle: Self-hosted means full data control — important for data residency requirements under GDPR and NIS2
Automating Azure DevTest Labs
Automate lab management and integrate with CI/CD pipelines.
Start on Pluralsight →Port: The Flexible SaaS
- Pros: Flexible data model (model anything as an entity), powerful self-service actions, fast time-to-value, good API
- Cons: SaaS dependency, customisation limited to configuration (no code-level changes), pricing scales with users
- Best for: Mid-size enterprises (50-500 developers) that want developer portal benefits without dedicating an engineering team
- Compliance angle: SOC 2 certified, but data lives in Port's infrastructure — evaluate against data residency requirements
Cortex: The Scorecard Champion
- Pros: Strongest scorecard and maturity tracking, auto-discovery of services, good for driving standards adoption across teams
- Cons: Less customisable than Backstage, SaaS dependency, focused primarily on service maturity rather than full developer portal
- Best for: Enterprises focused on service maturity, reliability standards, and engineering excellence metrics
- Compliance angle: Scorecards can track compliance metrics (security scanning, documentation coverage, SLO compliance)
AI Platform Assessment
Get a 2-3 week infrastructure audit with a concrete roadmap. No big-consultancy overhead.
Book Your Free Assessment →Total Cost of Ownership
3-Year TCO for 200 Developers
- Backstage: $0 licence + 1.5 FTE × $150K/yr = ~$675K (heavily front-loaded in year 1)
- Port: ~$30-50/user/month × 200 × 36 months = ~$216K-$360K (predictable, linear)
- Cortex: ~$30-50/user/month × 200 × 36 months = ~$216K-$360K (predictable, linear)
The hidden cost: Backstage's "free" open-source model often costs more than SaaS when you factor in engineering time. But at 500+ developers, the per-user SaaS cost exceeds Backstage's engineering investment.
Decision Framework
Choose Backstage if: 500+ developers, dedicated platform team, need full customisation, strict data residency requirements.
Choose Port if: 50-500 developers, want fast time-to-value, flexible data modelling, strong self-service actions.
Choose Cortex if: Primary goal is driving engineering standards and service maturity, want scorecards out-of-the-box.