Practical guides

Choose and test tabbed web views

Tabs switch between panels within one set. They can suit alternative views of the same object. If each part needs its own address or people must compare content simultaneously, navigation links or visible sections may work better.

Go to the method

Laptop, phone and open notebook on a white table
Illustration

Check that panels concern one object

Name the shared object: a product, case file or result. Labels then describe different views of it. A collection of unrelated services is more likely navigation than a tabbed interface.

Decide how to share a view, return after reloading and print the information. Compact presentation should not prevent finding content from an external reference. When comparing offers, check whether hiding other panels demands memorising too many details.

Match activation to loading

The W3C tabs pattern distinguishes moving between tabs and activating them. Automatic activation suits content that appears without noticeable delay. Manual activation avoids starting a request at every move. Selection and keyboard focus should remain distinguishable.

Specify behaviour after a failed request. A panel title, waiting message and retry option help explain the state. Do not show old data under a new heading without making clear that it does not yet belong to the requested view.

Preserve input and reading context

Changing panels should not silently discard entered data. Define what remains provisional in the form and what has actually been saved. If a change must be abandoned, an explicit decision is better than an unexplained loss.

On mobile, labels should remain reachable without confusing scrolling with switching views. Try a long translation, zoom and five tabs. If the arrangement becomes awkward, sections with a contents list may be simpler to maintain.

Example: read and edit a case file

Educational example: Summary, History and Documents concern the same case. Reading History does not alter it. A form in Summary retains input when Documents opens; a separate command saves changes. Testing distinguishes the selected view, provisional data and confirmed saving.

Content updated on October 7, 2026

Acceptance matrix to adapt to your project

These proposed checks use synthetic cases. Decide the required behaviour with the team, record the result and assign unresolved gaps before release.

Scroll the table horizontally to read every column. With a keyboard, focus the table area and use the arrow keys.

Test cases, expected outcomes and useful evidence
CaseExpected outcomeEvidence to retain
Moving between tabsFocus and selected panel are identifiableKey sequence and active panel
Delay or errorState matches the requested viewSlowed request result
Input before switchingValues follow the stated ruleValues and saving state
Shared viewThe link or journey restores contextReference opened in a new window

Frequently asked questions

Is a row of links a set of tabs?

Similar appearance is insufficient. If each item opens another page, retain link and navigation semantics.

Should a panel activate on focus?

Only when that does not slow movement. Remote content may suit manual activation.

Should all essential information be tabbed?

Compare visible sections. People need to find and combine the facts their decision depends on.