SEO & Search

Pointer Target Size on B2B Websites: Measure the Clickable Area and the Space Around It

A download icon looks small. The link around it may be large.

A resource card looks large. Only its title may respond to a click.

Measure the area that activates the function. Then inspect the targets around it. A screenshot of the icon leaves both questions unanswered.

This review isolates pointer target size under WCAG 2.2. It covers a fictional resource library with pagination, file downloads, and linked cards. The measurements and findings are invented. The worksheet is a proposed review method, not a W3C testing form.

Name the Target Before Measuring It

Start with the buyer's task. In this example, the buyer needs to open the next results page, download a specification, and read a resource.

Give each activation target an identifier. Record its function beside the identifier. A single card may contain several functions, so keep their records separate.

For the fictional download, identify three boundaries: the visible arrow, the clickable link, and the neighboring preview control. Mark each boundary on an annotated capture. Keep the original capture beside the annotation.

The bounding box encloses the active shape in an axis-aligned rectangle. Its center locates the spacing circle. Inactive corners stay inactive. W3C bounding-box discussion.

Use CSS pixels for the measurements. Do not count pixels in an enlarged screenshot. W3C evaluates the target in CSS units independently of the page's zoom factor. W3C size explanation.

During a real review, record browser version, viewport, page version, and zoom. Those fields let another reviewer reproduce the state. They do not change the threshold.

Ask the developer to confirm which element receives activation when the boundary is unclear. Treat a computed element rectangle as measurement evidence to investigate. Confirm that overlays, clipping, or event handling do not leave part of that rectangle inactive.

Leave uncertain geometry unresolved. A guessed width cannot support a completed finding.

Keep the Minimum and Enhanced Criteria Separate

WCAG 2.2 SC 2.5.8 is Level AA. Its minimum is 24 by 24 CSS pixels, subject to named exceptions. SC 2.5.5 separately sets 44 by 44 at Level AAA, with its own exceptions. Minimum criterion, enhanced criterion.

For the minimum size check, an axis-aligned 24-pixel square must fit inside the active area. A 24-pixel circular target cannot contain that square. An undersized target may qualify through spacing. W3C geometry explanation.

Write the selected criterion above the worksheet. Put any stronger internal design target in another field. Keep the acceptance question visible when someone requests larger controls.

A reviewer may recommend more generous dimensions after a minimum criterion check succeeds. Label that recommendation as a design decision. Do not rewrite the measured result to support the preference.

This article does not certify mobile usability or a whole site's accessibility. Continue separately with the keyboard navigation review and the zoom and reflow review.

Inspect Spacing in Every Direction

For an undersized target, center a 24-CSS-pixel-diameter circle on its bounding box. Check that circle against other targets and circles around other undersized targets. Intersections prevent using the spacing exception. W3C spacing explanation.

Use the bounding box center even when the active shape is irregular. Keep the active shape and the inspection circle visible in the evidence. They answer different questions.

For the proposed worksheet, record horizontal and vertical neighbors. Include controls in the row above and below. A generous gap on the right says nothing about a control directly underneath.

Prefer examples with clear separation during an initial review. A disputed measurement at a boundary deserves closer inspection. Preserve the unrounded coordinates before anyone decides the result.

The fictional examples below use flat, axis-aligned rectangles and exact coordinates. They exclude overlays and transformations. Each circle has a radius of twelve CSS pixels. Those assumptions make the arithmetic inspectable.

Do not transfer an example's conclusion to a page with different geometry. Use the page's measured coordinates.

Work Through Fictional Pagination

Suppose two adjacent pagination links each have an active rectangle twenty CSS pixels wide and twenty high. Their centers sit at coordinates zero and twenty-two on the same horizontal line.

The first rectangle extends from minus ten to ten. The second extends from twelve to thirty-two. Their visible gap is two CSS pixels.

The inspection circles extend twelve pixels from their centers. Their center distance is twenty-two, less than their combined radii of twenty-four. They overlap. The fictional pair cannot use the spacing exception.

Now propose increasing the gap to six CSS pixels while retaining the twenty-pixel targets. The second center moves to twenty-six. The circles have two pixels between their nearest edges.

Assume the example contains no other targets near either circle. The revised pair qualifies through spacing. Record the qualifying route through the criterion instead of calling the targets twenty-four pixels wide.

A different proposed revision enlarges each active rectangle to twenty-eight by twenty-eight. Space the revised rectangles so neither overlaps its neighbor. Keep the displayed page number unchanged. Verify that the enlarged rectangles activate their intended pages throughout their area.

Choose a revision with the component owner. This article specifies the evidence to collect, not a universal pagination design. Retest all visible page-number links, including the current-page state where its behavior differs.

Check the Download Beside the Resource Card

In a second fictional component, a sixteen-pixel arrow sits inside a thirty-two-by-thirty-two rectangular download link. Assume the whole rectangle activates the download.

The icon dimension does not establish a failure. The active rectangle contains the required square. Record the icon size as context and the active dimensions as the criterion evidence.

Now consider a separate preview icon beside a linked resource card. Its active rectangle is sixteen by sixteen. Place its center at zero. Its right edge is eight pixels from that center.

The large card's active area begins ten pixels to the right of the preview center. The preview circle extends to twelve. It therefore overlaps the card by two pixels horizontally. Assume the card crosses the horizontal line through the preview center.

The card's own generous dimensions do not resolve the preview target's spacing problem. Compare the undersized circle with the card's active area. A second circle around the large card is unnecessary for this comparison.

Propose moving the card's left edge to fourteen while keeping the preview fixed. The circle then ends two pixels before the card. Under the example's stated assumptions, the spacing conflict is removed.

Check other nearby targets before closing that row. A bookmark control above the preview could introduce another conflict. Keep every relevant neighbor in the measurement record.

Document Exceptions Before Using Them

The minimum criterion names Spacing, Equivalent, Inline, User Agent Control, and Essential exceptions. Equivalent requires another qualifying control on the same page. Inline concerns sentence links or line-height constraints from non-target text. User Agent Control requires unmodified browser-determined size. Essential concerns necessary or legally required presentation. W3C exceptions.

For this proposed process, attach evidence to every exception decision. A label in a spreadsheet is insufficient.

Suppose a tiny download icon and a large text link both appear on the fictional resource page. Compare their functions. Does each retrieve the same specification in the same format? Record the destination and behavior before proposing Equivalent.

If the large link opens an overview page, leave the equivalence claim unproven. A reviewer should not assume the buyer can recover the original function somewhere downstream.

For a sentence link, capture the surrounding sentence and its layout. Contrast that evidence with a standalone pagination control. Calling a number text does not describe how its target fits the exception.

For a browser control, ask which styles the author changed. Save the answer with the inspected state. Do not apply User Agent Control merely because the component resembles a native input.

Treat Essential as a claim requiring an explanation. Brand preference and a crowded mockup do not establish necessity. If someone cites a legal requirement, request its exact source and applicability before recording it.

Copy a Pointer Target Worksheet

Use one row per target and inspected state. Keep supporting captures outside the table when they need more space.

Component Activation area CSS dimensions Neighbors Exception evidence Finding Retest state
Pagination link Confirmed rectangle 20 by 20 Next center 22 away Spacing circles overlap Spacing unavailable Open
Revised pagination Confirmed rectangle 20 by 20 Next center 26 away Circles separated; no other neighbors Qualifies through spacing Proposed, untested
Specification download Full link rectangle 32 by 32 Record adjacent controls Size check used Size condition met in example Synthetic only
Preview beside card Small preview rectangle 16 by 16 Card begins 10 from center Circle reaches 12 Spacing unavailable Open
Other component [Confirmed active shape] [Measured units] [All nearby targets] [Evidence or unresolved] [Bounded result] [Build and observation]

Add the criterion, environment, evidence identifiers, owner, and review date above the table. The template's fictional outcomes demonstrate how to write a row. Replace every input during a real review.

Keep proposed and observed results separate. A changed design file is not a retest. Leave the row open until someone examines the implementation that will ship.

If several pages reuse the component, list the inspected instances. Record any variant excluded from the review. Shared markup gives a reason to investigate related pages, not evidence that every instance behaves identically.

Hand Over a Finding Someone Can Reproduce

Write the finding around the active area and the conflicting neighbor.

For the fictional preview, a useful record states: the sixteen-by-sixteen preview target's spacing circle intersects the adjacent card in the inspected layout. Include the coordinates and the marked capture.

Attach the proposed revision separately. Name who will choose between enlarging the preview target and changing its position. Preserve any unresolved interaction with the card's link behavior.

During retest, repeat the original state first. Then inspect the revised component at the other recorded layouts. Record fresh dimensions instead of copying values from the design ticket.

If a shared style changes, inspect the nearby download control again. A revision can change neighboring geometry. Keep the final conclusion attached to the measurements used to reach it.

For forms, continue with the accessible lead-form review after this bounded pointer check. A sufficiently large submit target still needs a usable surrounding form.

Close the worksheet when each row has evidence or an explicit unresolved condition. Keep the clickable boundary visible.

Sources

Topic map: Accessible B2B websites

Articles in this cluster, then the 3 clusters that share the most topics with it.

This cluster: Accessible B2B websites

Neighbor: Technical SEO, crawling, and migrations

Neighbor: Customer and usability research

Neighbor: Topical authority and keyword strategy