Signs your website navigation is losing visitors, and how to test it

Your analytics show a steady stream of visitors, but enquiries do not match. The service pages get fewer views than expected, and the contact page even fewer. It is easy to blame the content or the prices, yet a surprising amount of lost business comes from people simply not finding things. Before changing anything, it pays to test website navigation properly: first by reading the warning signs already in your data, then by watching real people try to use your menu.

Warning signs in your analytics

None of these signals proves a navigation problem on its own, but several together are a strong hint. GA4 is used as the example below, although most analytics tools show similar data.

People search for pages that are already in the menu

If your site has search and GA4’s enhanced measurement is tracking it, look at the search terms. Repeated searches for “contact”, “price” or the exact name of a service listed in your menu mean visitors did not spot or understand the menu item.

High exits from pages that should lead somewhere

In the Pages and screens report, compare views with exits. A homepage or category page with a high proportion of exits is a crossroads that is failing to route people onwards.

Ping-pong paths

GA4’s Path exploration can show sequences like Home → Services → Home → About → Home. Bouncing back and forth between hub pages suggests visitors are hunting rather than following a clear route.

Key pages that almost nobody reaches

If your pricing or booking page receives a tiny fraction of homepage visits, check how many clicks and how much guessing it takes to reach it.

Very short engagement on landing pages from search

Visitors from Google who leave after a few seconds may have landed on the right page but found no obvious next step.

Warning signs outside analytics

• Customers calling to ask things clearly answered on the site.
• Staff sending the same page link on WhatsApp repeatedly because customers “couldn’t find it”.
• Enquiry forms submitted with the wrong service selected.
• Feedback such as “the website is a bit confusing” with no specifics.

These are free, honest signals. Write them down as they happen, because they point directly at the tasks worth testing.

Test 1: The five-second test

This checks whether your homepage and menu communicate clearly at first glance.

1. Show someone unfamiliar with your business the homepage for five seconds.
2. Hide it.
3. Ask: What does this business do? Who is it for? Where would you click to see prices or get in touch?

If people cannot name your main service or cannot recall any menu items, your navigation and top section are not doing their job. Repeat with five people and look for patterns rather than individual opinions.

Test 2: Tree testing your menu structure

Tree testing isolates the menu structure from design. Participants see only a text hierarchy of your menu and pages, with no visuals, and are asked where they would go to complete a task.

Running it cheaply

1. Write your navigation out as an indented list, including dropdown items and their child pages.
2. Prepare five to eight realistic tasks, such as “Find out whether they build online stores” or “Find the cost of annual hosting.”
3. Show the list on paper or in a shared document and ask participants to point to where they would click first, then next.
4. Record whether they reached the correct page, how directly, and where they went wrong.

Dedicated online tree-testing tools exist and speed this up with larger groups, but a paper version with a handful of people will still reveal the worst problems. A task that most people get wrong usually means a confusing label or a page filed in an unexpected section.

Test 3: Hallway usability testing

Hallway testing means asking whoever is nearby, colleagues from other departments, friends or family who match your customers reasonably well, to use the real website while you watch.

How to run a session

• Use a phone for at least half the sessions, since many of your visitors will.
• Give one task at a time, written on a card, phrased without using your menu labels. Say “find out if they can deliver to Pune” rather than “find the Shipping page”.
• Ask the participant to think aloud.
• Do not help, point or explain. Silence is uncomfortable but necessary.
• Note hesitations, wrong clicks, use of search, and moments they give up.

Usability practitioners commonly suggest that around five participants reveal most major problems in a round of testing. Run a small round, fix what you find, then test again.

Turning findings into fixes

Group what you observed by cause, since similar symptoms often have one root:

• Wrong clicks on a specific item: the label is ambiguous. Rename it using the words participants used.
• Nobody opens a dropdown: important pages are hidden too deep. Promote them or add links on the parent page.
• People use search for menu items: the menu is overlooked, perhaps too faint or hidden behind an icon on desktop.
• Participants go to the footer: the page they need is missing from the main navigation.
• Mobile-only failures: the collapsed menu, tap targets or dropdown arrows need attention.

Change one or two things at a time, then compare the same analytics signals over the following weeks, so you know which change made the difference.

Build navigation testing into routine

A short round of hallway tests every few months, or whenever you add a section, catches problems while they are small. Keep your task list and reuse it so results are comparable over time. If testing reveals that the structure itself needs rethinking, you can ask our experts to review the findings with you and plan the changes.

Frequently asked questions

How many people do I need to test website navigation?

Around five participants per round is a common guideline for spotting major issues. Several small rounds with fixes in between are more useful than one large test.

What is the difference between tree testing and card sorting?

Card sorting asks people how they would group content, helping you create a structure. Tree testing checks whether people can find things in a structure you already have.

Can I test navigation without paying for tools?

Yes. Paper tree tests, five-second tests with printed screenshots and hallway sessions on your own phone cost nothing but time.

Which GA4 reports help find navigation problems?

Pages and screens, Landing page reports, Path exploration and site search terms from enhanced measurement are the most useful starting points.