You ship a feature. The tests pass. Adoption flatlines. Weeks later, a customer describes the exact problem the feature solves and asks when you are planning to build it. You already did. They could not find it. Every unit test, integration test, and end-to-end test confirmed the code worked. The failure is that nothing in your test suite asked whether a real user could discover the feature through your actual navigation.
Navigation coverage is the percentage of an application’s features that are reachable through actual user navigation paths. Whilst code coverage measures how much of your source code your tests exercise, navigation coverage measures how much of your product your users can actually reach by clicking through the interface.
What conventional testing misses
Conventional tests answer whether code behaves correctly given a specific sequence of steps. They do not ask whether a user would ever encounter those steps in the first place.
Consider a settings page buried three levels deep behind a menu that only appears on certain screen sizes. An end-to-end test navigates directly to /settings/notifications/advanced and confirms the toggles work. The test passes. But if no user can find that page by clicking through the interface, the feature might as well not exist. The test validated the implementation, not the discoverability.
This gap exists because testing frameworks operate on URLs and selectors, not on navigation paths. They test what the application does when you arrive at a page. They do not test whether you can arrive at that page through the routes your interface provides. Features that work perfectly but cannot be found generate no value for users and no return on the engineering investment that built them.
The problem compounds across complex applications. Every new feature adds potential navigation paths. Every redesign reshuffles them. Without a systematic way to measure which features remain reachable after each change, undiscoverable functionality accumulates silently. By the time anyone notices, the gap between what the product does and what users experience has become significant.
How Glia Quest works
An AI agent navigates your site the way a real user would, clicking links, opening menus, and following navigation paths, then reports what it found and what it could not reach. No SDK to install. No code access required. No test scripts to write. You provide a URL.
The agent builds a Navigation Map of every reachable feature and the specific path it took to get there. This is not a sitemap comparison or a link crawler. The agent interacts with the interface: JavaScript-rendered navigation, dropdown menus, modal triggers, conditional UI elements that only appear after specific interactions. It discovers your site the way your users do, through the interface itself.
The Navigation Coverage Score tells you what percentage of your product is actually discoverable. A score of 100 per cent means every feature the agent expected to find was reachable through at least one navigation path. A score below that means features exist in your codebase that a user cannot reach by clicking through your site.
For every unreachable feature, the report includes an investigation prompt: a structured, copy-pasteable brief you can hand to your development team or paste directly into a coding assistant. It describes what the agent expected to find, where it looked, and what navigation path it attempted. This turns a vague “something is missing” into a specific, actionable fix.
Who this is for
Anyone responsible for a web product who suspects their navigation has gaps but has no systematic way to measure them. The common thread is a simple question: can your users find what you have built?
Engineering teams reach for Glia Quest after a redesign or refactor, when the navigation structure has shifted and nobody is quite certain what moved. Product managers use it to verify that the user journey they designed matches the one the interface actually delivers. Founders use it because they want an honest, external assessment of their product’s discoverability, one that does not rely on a team member clicking through the site and saying “looks fine.”
The tool requires no technical integration. It works from the outside, the same way your users do, by visiting your site and trying to find things. If your users need a URL and a browser, so does Glia Quest.
What is next
Glia Quest is live at app.glia.quest. Free preview audits provide a Navigation Coverage Score and a Navigation Map for sites of up to 50 pages. The preview takes approximately 60 seconds and requires nothing more than a URL.
For teams managing larger applications with complex navigation structures, persona-aware navigation paths, or multi-locale sites, full audit capabilities are available through the platform.