Understanding the root causes of API drift

Bill Doerrfeld | November 20, 2024

Many APIs stray from their definitions. My latest feature on Nordic APIs explores the root causes behind API drift.


A recent report found 75% of APIs stray from their OpenAPI specifications. That surprised me and inspired me to ask: why is drift so common?


There are a ton of answers to why drift is endemic. Intentional documentation and spec-first cultures are rare. APIs keep multiplying yet API governance is still in its infancy. Others attribute it to a lack of comprehensive testing and tools to produce quality descriptions. Or, a lack of discipline in general.


Today on Nordic APIs, I dive deeper into API drift to understand some of the root causes and to see how drift impacts developer experience and the business at large. We also put together some solid tips on ways engineering leaders can mitigate drift.


All in all, it seems like solving the drift problem will require tools to detect drift, conformance testing, as well as spreading more knowledge-sharing and clarifying responsibilities. 


Thank you to APIContext for igniting some interesting discussion and to everyone who participated in this story!


Read: Understanding The Root Causes of API Drift
Large action models LAMs story Bill Doerrfeld The New Stack
By Bill Doerrfeld June 10, 2025
AI researchers are calling the next class of models large action models (LAMs). For The New Stack, I explored what LAMs are, what examples are emerging in the market, and what experts think.
7 proven AI prompting strategies for coding to try today
By Bill Doerrfeld June 9, 2025
My article for LeadDev explores specific prompting techniques proven to sharpen your AI-assisted software development.
Nordic APIs ranked #1 API blog on the web
By Bill Doerrfeld June 7, 2025
Nordic APIs, the API-specific blog I edit, was recently ranked the top API blog online by FeedSpot. After ten years managing this presence, I reflect a bit on the journey thus far.
Tips to improve your AI vibe coding
By Bill Doerrfeld June 3, 2025
Developers are realizing that being productive with AI coding assistants takes a lot more than just asking nicely. There's real craft to it.
AI coding is the easy part. Now it's time to focus on production.
By Bill Doerrfeld May 30, 2025
AI coding is the easy part. Now it's time to focus on DevOps to get it into production. In a recent interview for LeadDev's DirectorPlus, Honeycomb's CTO, Charity Majors, shares expert tips on how to accomplish this.
New tools help LLM devs improve training data decisions
By Bill Doerrfeld May 29, 2025
My latest for The New Stack explores new research from Ai2, DataDecide, which helps LLM developers make better decisions around training data.
MCP security vulnerabilities
By Bill Doerrfeld May 21, 2025
My APISEC|CON talk covers the hype around agentic AI and MCP, and delves into inherent flaws in MCP architectures and suggests mitigations.
Knowing when to use AI coding assistants Doerrfeld InfoWorld
By Bill Doerrfeld May 6, 2025
AI coding assistants are a productivity dream in some cases — and a debugging nightmare in others. So, where’s the line?
How semantic caching reduces LLM API calls
By Bill Doerrfeld May 5, 2025
Semantic caching is like typical caching, but for AI. It could eliminate a lot of redundant API calls to LLMs, reducing costs and improving performance.
Using agentic AI for business workflows
By Bill Doerrfeld April 30, 2025
For CIO.com, leading executives shared with me how they're actively utilizing agentic AI to enhance core business workflows.
More Posts