
Building Internal Developer Platforms That Developers Actually Want to Use: Lessons from High-Performing Platform Teams
Why Most Internal Platforms Fail
Companies like Spotify, Netflix, and Amazon operate at a scale where engineering velocity is paramount. To accelerate delivery, many organizations build Internal Developer Platforms (IDPs)—self-service layers that streamline infrastructure management, CI/CD, security, and deployment pipelines.
But here’s the problem: Most IDPs fail not because of bad technology—but because developers don’t want to use them.
According to a 2024 report from Humanitec, over 60% of internal platforms suffer from poor developer adoption. Gartner echoes this concern, highlighting that “developer-facing tools often collapse under unclear ownership, lack of user empathy, and no measurable success criteria” (Gartner, 2024).
So how do you build a platform your developers will actually want to use?
The Answer: Treat Your Internal Platform Like a Product
This post explores lessons from top-performing platform teams at companies like Shopify, Netflix, Spotify, and Wealthsimple. We’ll cover:
Why treating your IDP as a product—not just a project—is critical.
How to create tight feedback loops with developers.
The platform metrics that matter (beyond uptime).
Real-world case studies and actionable implementation guides.
We’ll also link to related CrashBytes posts, such as Optimizing CI/CD Pipelines for Global Teams and Developer Experience Metrics: What Engineering Leaders Need to Track.
The Shift: From Infrastructure Automation to Developer Product
What is an Internal Developer Platform?
An Internal Developer Platform (IDP) provides developers with self-service access to tools, workflows, and environments. At its best, it removes cognitive load from engineering teams and accelerates delivery.
Typical IDP features include:
Self-Service Infrastructure: Provision Kubernetes clusters, databases, or queues without ticketing ops.
Golden Paths: Predefined deployment pipelines, security policies, and standardized templates.
CI/CD Integrations: Seamless build, test, and deploy workflows.
Environment Management: On-demand staging environments with parity to production.
But tooling alone doesn’t create value.
As Camille Fournier, author of The Manager’s Path, points out:
“Internal platforms aren’t valuable unless developers actually want to use them. Adoption is the outcome, not the starting point.”
Lesson 1: Treat Your Internal Platform as a Product
Platforms Are Products, Not Projects
High-performing platform teams act like product teams.
At Spotify, the Backstage platform team follows a product mindset. They prioritize user research, roadmap iteration, and regular feedback cycles—not just technical delivery.
This requires:
Mindset Shift | Project Approach | Product Approach |
Ownership | One-time delivery | Ongoing iteration |
Success Metric | Feature complete | Developer satisfaction, NPS |
Process | Waterfall | Agile, iterative |
Feedback | Optional | Mandatory |
Create a Dedicated Platform Product Owner
Companies like Netflix and Wealthsimple assign Platform Product Owners (PPOs) who:
Conduct user interviews.
Prioritize features based on developer pain points.
Define KPIs tied to business outcomes.
Example: Netflix
Netflix’s engineering blog describes their IDP as “a product with its own backlog, dedicated PM, and UX research.” They run quarterly developer satisfaction surveys to adjust priorities (Netflix Tech Blog).
Lesson 2: Build Tight Developer Feedback Loops
Why Developer Feedback Is Non-Negotiable
According to McKinsey, “teams that continuously gather and implement developer feedback deliver platform features 30% faster” (McKinsey, 2023).
Feedback prevents:
Overengineering solutions nobody needs.
Building tools that don’t fit real workflows.
Losing trust with developers.
Methods for Collecting Feedback
1. Embedded Platform Advocates
Shopify embeds platform engineers within feature teams. This creates natural, trust-based channels for feedback.
2. Open Office Hours
At Wealthsimple, the platform team hosts weekly “Ask Me Anything” (AMA) sessions via Slack and Zoom.
3. Developer Journey Mapping
Map out common developer workflows. Identify friction points and gather qualitative input.
Template: Developer Journey Map
Stage | Developer Action | Pain Point | Improvement Opportunity |
Deploy New Service | Configure Kubernetes YAML | YAML complexity | Provide CLI abstraction |
Request Database | File IT ticket | Slow turnaround | Automate DB provisioning |
Lesson 3: Define Success Metrics That Actually Matter
Avoid Vanity Metrics
Traditional platform teams often measure:
System uptime
Number of deployments
Tickets closed
These are operations metrics, not product success indicators.
Developer-Centric Platform Metrics
DORA Metrics Extension for Platforms
Metric | What It Measures | Why It Matters |
Time to First Commit (TTFC) | Onboarding friction | Measures how fast new developers can push code |
Mean Time to Recovery (MTTR) | Failure recovery speed | Tracks how quickly developers can fix broken deploys |
Developer NPS (Net Promoter Score) | Satisfaction | Indicates adoption health |
Cognitive Load Index | Workflow complexity | Measures friction in everyday tasks |
Lesson 4: Build for Adoption, Not Control
Make Golden Paths Optional (but Desirable)
At Spotify, golden paths are recommended, not enforced. This builds trust while encouraging consistency.
Example: Spotify Backstage
Backstage offers templates for microservice creation but allows teams to customize. This balance between standardization and autonomy is critical for adoption.
Lesson 5: Show Developers the “What’s In It For Me”
The Adoption Equation
Developers adopt platforms when:
Perceived Value > Perceived Cost
How to Increase Perceived Value:
Save developers time (faster deployments, fewer meetings).
Reduce context switching (single pane of glass for dev workflows).
Improve developer happiness (less toil).
How to Reduce Perceived Cost:
Lower cognitive load (no YAML unless necessary).
Provide great docs and examples.
Offer excellent support (Slack channels, docs, office hours).
Case Study: Wealthsimple’s Platform-as-Product Model
Company: Wealthsimple Platform: Kubernetes-based IDP Approach: Treat platform as a product
What They Did:
Appointed a Platform Product Manager.
Held weekly developer feedback sessions.
Built a self-service portal with templates for common tasks.
Measured time-to-onboard and developer satisfaction quarterly.
The Result:
Reduced service deployment setup from 3 weeks to 2 days.
Achieved a +52 Developer NPS, up from +5 before the changes.
Increased platform adoption from 20% to 90% across teams.
Source: Wealthsimple Engineering Blog
Real-World Example: Amazon’s Internal Platform Culture
At Amazon, internal platform teams follow a Working Backwards approach:
Write a press release for the platform before building it.
Focus on customer obsession—in this case, internal developers.
Use metrics-driven prioritization, tying platform features to customer impact.
This approach led to the creation of services like:
Apollo: Amazon’s deployment platform, used by over 100,000 deployments daily.
Brazil: Amazon’s internal service-oriented architecture management system.
Read: How Amazon Builds Platforms
Implementation Guide: Building Your Developer-Centric Platform
Step 1: Set Up a Platform Product Team
Roles Needed:
Platform Engineering Lead
Platform Product Manager
Developer Advocate or UX Researcher
Step 2: Conduct Developer Research
Use these questions to guide interviews:
What’s the hardest part of your daily workflow?
Where do you feel blocked?
What tools do you avoid—and why?
How do you prefer to get help? (Slack, docs, office hours?)
Step 3: Prioritize Features by Impact
Use Impact vs Effort mapping to rank initiatives.
Feature | Impact | Effort | Priority |
On-demand staging environments | High | Medium | High |
Custom Kubernetes operator | Low | High | Low |
Golden path templates | High | Low | High |
Step 4: Measure Success Continuously
Set up quarterly reviews using:
Developer NPS surveys
Time-to-onboard measurements
Cognitive load assessments
Common Pitfalls (and How to Avoid Them)
Pitfall | Solution |
Overcomplicating the platform | Start small, iterate |
Ignoring developer feedback | Build continuous feedback loops |
Lack of clear documentation | Treat docs as part of the product |
Treating platform as a side project | Assign dedicated product ownership |
The Future of Internal Developer Platforms
Emerging Trends:
Platform Orchestration Tools: Tools like Humanitec are automating IDP assembly.
Developer Experience (DevEx) Teams: More orgs are formalizing DevEx ownership.
Platform Engineering Guilds: Cross-company collaborations like the Platform Engineering Community.
Beyond the Tooling
Ultimately, building an IDP is about people, not just tech.
As Charity Majors, CTO of Honeycomb, says:
“Tools don’t create high-performing teams. People do. But tools can remove the roadblocks that frustrate people.”
Conclusion: Build for Trust, Iterate for Value
If you want your Internal Developer Platform to succeed:
Treat it like a product.
Listen to your developers—continuously.
Measure what matters, not just uptime.
Prioritize adoption over control.
By following these lessons from high-performing teams, you’ll create a platform that developers actually love to use—not one they’re forced to use.
Crosslinks & Backlinks
Optimizing CI/CD Pipelines for Global Teams
Developer Experience Metrics: What Engineering Leaders Need to Track