Short answer: Platform engineering is now the norm. The 2025 DORA research found that 90% of organizations use an internal developer platform and 76% have dedicated platform teams. More importantly for CTOs, DORA found that when platform quality is high, AI adoption has a strong positive effect on organizational performance; when platform quality is low, the effect is negligible.
Key platform engineering data at a glance
- 90% of organizations use an internal developer platform (DORA 2025).
- 76% have dedicated platform teams (DORA 2025).
- Developer independence from enabling teams brings a 5% productivity gain at team and individual level (DORA 2024).
- 90% of DORA respondents use AI at work, about 2 hours a day (DORA 2025).
- AI adoption pays off strongly only where platform quality is high (DORA 2025).
[Table 1]
What is platform engineering?
Platform engineering is the practice of building an internal developer platform: shared tools, environments, pipelines and standards that product teams use through self-service. The goal is to let engineers ship safely without filing tickets or waiting on other teams.
Why does the platform decide AI returns?
DORA describes AI as an amplifier. AI-assisted engineers produce more code, and that code moves through the same testing, security review and deployment steps as before. Without a strong platform, the extra output piles up in those downstream steps. With a strong platform, automated gates absorb the volume and the gains reach customers. See how CTOs measure the ROI of AI coding agents.
Which platform capabilities matter most?
[Table 2]
Build, buy or assemble?
- Build when the platform is a competitive advantage or compliance needs are unusual, as in regulated fintech.
- Buy when speed matters more than customization and the team is small.
- Assemble, the most common choice, combining open-source and commercial components behind a thin internal layer.
How should CTOs organize the platform team?
- Treat the platform as a product. Give it a product manager, a roadmap and internal customers.
- Measure adoption and satisfaction. A platform nobody uses is overhead.
- Fund it before scaling AI. Invest in testing, observability and golden paths ahead of wider agent rollout.
- Size it to the organization. Many B2B SaaS companies start a platform team between 30 and 60 engineers. See engineering org design.
Frequently asked questions
What percentage of companies use platform engineering?
90% of organizations in DORA's 2025 research use an internal developer platform, and 76% have dedicated platform teams.
Does platform engineering improve AI results?
Yes. DORA found AI adoption has a strong positive effect on performance when platform quality is high, and a negligible effect when it is low.
When should a company create a platform team?
When several product teams duplicate infrastructure work or wait on each other to ship, often between 30 and 60 engineers.
Compare platform decisions with other CTOs: apply to join CTO Club.
[Table 1]
Metric | Figure | Source
Organizations using an internal developer platform | 90% | DORA 2025
Organizations with dedicated platform teams | 76% | DORA 2025
Productivity gain when developers can work independently of enabling teams | 5% at team and individual level | DORA 2024
Effect of AI adoption when platform quality is high | Strong and positive on organizational performance | DORA 2025
Effect of AI adoption when platform quality is low | Negligible | DORA 2025
Respondents using AI at work | 90%, about 2 hours a day | DORA 2025
[Table 2]
Platform capability | Why it matters for AI-assisted delivery | Question for the CTO
Self-service environments | Agents and engineers can test changes without waiting | How long does a new environment take?
Automated testing and security checks | Catches defects from higher code volume | What share of changes pass automated gates?
Golden paths and templates | Keeps agent output consistent with standards | How many services follow the paved road?
Observability | Finds instability before customers do | Can every team trace a failed change?
Cost visibility | AI and cloud spend stay accountable | Can each team see its own spend?
