A menu has very little room to explain itself. Each label needs to give somebody enough confidence to choose a route before they have seen the destination. I start with the questions visitors bring, then check whether the menu uses words they would recognise.
An organisation chart rarely makes a helpful first draft. Internal teams, product families and department names may mean something to staff while leaving new customers guessing. The navigation should help the person outside the business understand where to go.
Start with a handful of real tasks
Write down five things someone should be able to do: find a suitable service, compare products, check delivery, read useful advice or ask a question. Use tasks that belong to the actual site. Then point to the first menu item you expect that person to choose for each one.
If several tasks have no obvious route, adding a decorative icon will not solve the problem. If every task points to one enormous category, that category may need a clearer landing page or more useful subcategories. I want a route that can be explained in ordinary language.
Labels such as “Solutions” are not automatically wrong, but they ask the visitor to supply missing context. “Website support” tells a small business more about the destination. Where specialist terminology is necessary, the destination page can explain it without forcing the menu to carry a paragraph.
Make the parent item’s behaviour deliberate
A label with a submenu presents a small design decision: does pressing it open the submenu, visit an overview page, or both through separate controls? I decide this explicitly and keep the behaviour consistent. People should not have to discover a different rule for each menu item.
A useful overview page can serve somebody who knows the broad area but not the exact service. Removing that route because a dropdown exists can make the site harder to explore. Conversely, a heading that has no destination should not pretend to be a working link.
I check the menu with a mouse, keyboard and touch. A submenu that relies on keeping the pointer over a narrow invisible gap can frustrate someone even when its visual styling looks excellent. Opening, moving through and closing the menu are all part of the same task.
Review the cramped version early
The longest menu label is a better test than the shortest. I try it alongside the logo, search control and any prominent contact button, then reduce the available width gradually. The change to a compact menu should happen when the content needs space.
On a phone, I check whether the open menu makes its own location obvious, whether the close control is easy to find and whether a person can reach the last item. If a menu contains nested sections, its back controls should describe the movement clearly rather than resembling browser navigation by accident.
Try a five-minute language test
Give somebody the task list and the menu labels without showing the rest of the design. Ask where they would start and what they expect to find there. This is a small qualitative check, not a statistically reliable conversion study, but hesitation can reveal ambiguous wording quickly.
Finally, open each destination and compare it with that expectation. A clear label leading to an unexpected page still breaks the promise. Keep a short note explaining the purpose of each top-level item so future additions do not gradually turn the menu into a list of internal requests.