The Stratification Problem, Explained

Every tool you build makes assumptions about who can use it. Those assumptions create winners and losers. This is ALC's core insight, and understanding it changes how you think about software design.

What Is Stratification?

In the ALC framework, "stratification" refers to how tool design creates different tiers of users based on their ability to navigate the application layer. It's not about intelligence or even technical skill in the traditional sense — it's about a specific kind of fluency: the ability to communicate with and through software systems.

Think about a simple example: a productivity app that requires users to set up keyboard shortcuts, customize workflows, and integrate with other tools to be effective. For users with high ALC fluency, this is empowering. For users without it, the app is frustrating or unusable.

The tool didn't intend to exclude anyone. But by assuming a certain baseline of application layer fluency, it effectively did.

The Invisible Barrier

What makes stratification tricky is that it's often invisible to the people who don't experience it. If you have high ALC fluency, you might not realize that others struggle with things you find intuitive.

Consider these common assumptions built into modern software:

• Users understand that data syncs across devices (eventually, with caveats)
• Users can troubleshoot when something doesn't work as expected
• Users know to look for settings in the "hamburger menu" or gear icon
• Users understand the difference between "save" and "publish"
• Users can navigate nested menus and hierarchical structures

Each assumption seems reasonable. But stack them up, and you've created a tool that only certain people can use effectively.

Why This Matters Now

Stratification has always existed, but it's becoming more consequential. More of life happens through software. Employment, healthcare, government services, social connection — all increasingly mediated by digital tools.

If those tools stratify users by ALC fluency, we're creating a society where some people can navigate systems effectively and others can't. That's not a technical problem — it's a social one with technical roots.

The AI Amplification Effect

AI tools have introduced a new dimension to stratification. Users who know how to prompt effectively, integrate AI into their workflows, and verify AI outputs gain enormous productivity advantages. Users who don't are left further behind.

This isn't AI's fault — it's a design problem. AI tools could be built to meet users where they are. Many aren't. They assume users already understand the conventions of prompting, the limitations of the model, and the appropriate use cases.

Reducing Stratification

The goal isn't to dumb down tools. Users with high ALC fluency shouldn't lose capabilities. The goal is to provide multiple paths to the same destination.

Some strategies:

• Progressive disclosure: Start simple, reveal complexity as users demonstrate readiness
• Multiple entry points: Let users accomplish tasks through different interaction patterns
• Explicit defaults: Make assumptions visible and changeable
• Graceful degradation: When users don't use advanced features, the basic experience should still work
• Agent-friendly design: Let AI assistants help users who can't navigate directly

Diagnosing Stratification

If you want to understand your tool's stratification profile, ask:

• What does a user need to know before they can use this effectively?
• Where are the failure points for users without that knowledge?
• What happens when something goes wrong — can users recover?
• Do power features require the same fluency as basic features?
• Can users with different abilities accomplish the same core tasks?

These questions surface the assumptions baked into your design.

Learn More

The stratification problem is a central theme in the ALC framework, developed by Roger Hunt as part of his dissertation research. For deeper exploration, see the "Beyond Knowledge Graphs" essay series on The Sausage Mill.