Skip to content
Insights/Article

Faint Text and Tiny Buttons: How to Audit Website Readability and Touch Targets

Key takeaway

Audit readability by separating WCAG 2.2 AA floors from UX targets: enforce 4.5:1 text contrast (3:1 for large text and UI borders), 24x24 CSS px targets or clearance, visible focus, and 320px reflow—while aiming for 44–48px tap areas and 16px+ mobile inputs.

Blue-and-white website audit diagram showing 4.5:1 contrast gauges, 24px vs. 44px touch target hit boxes, a visible keyboard focus ring, and 320px narrow-viewport wrapping

Minimalist corporate website mockups almost always shine on a calibrated studio display: pale gray captions, sleek 16px icons, and clean inputs with focus outlines stripped away. Once real buyers open that site on a phone in daylight or on a standard office laptop, those choices backfire quickly. Light captions fade out, tapping a form field triggers sudden iOS page zoom, tiny close buttons cause mis-taps, keyboard Tab focus disappears, and long English words overflow horizontally on a 320px viewport.

You cannot verify readability and interaction accessibility by eyeballing a Figma artboard. The reliable method is splitting acceptance into two tiers: mandatory W3C WCAG 2.2 Level AA baselines (can users actually see and operate every control?) and engineering and UX targets (does the interface feel natural for thumbs, mobile browsers, and bilingual text?).

Distinguish conformance criteria from design starting points below. Items labeled with a WCAG level have their stated exceptions; input font size, line length, and line height recommendations still need checks with the actual font and target device. They are not all mandatory accessibility thresholds.

Separating Two Tiers: WCAG 2.2 AA Baselines vs. Engineering & UX Targets

Four reading and interaction checks
  1. Text and background contrast
  2. Target size and spacing
  3. Visible keyboard focus
  4. Reflow and zoom

See the article for criteria, levels, exceptions and design recommendations.

1. Text and UI Contrast Ratios (4.5:1 vs. 3:1)

Under W3C WCAG 2.2 SC 1.4.3 Contrast (Minimum) (Level AA), normal text and images of text require at least 4.5:1 contrast against their background, while large text (18pt / ~24px regular, or 14pt / ~18.67px bold; equivalent scale for CJK) requires at least 3:1. Watch three strict rules: ratios cannot be rounded (4.499:1 fails 4.5:1); placeholder text and text shown on hover or focus must also comply; and only disabled controls, pure decoration, and brand logotypes are exempt—brand guidelines never exempt regular UI text, and intentionally muting logos until hover is not exempt either.

Meanwhile, W3C WCAG 2.2 SC 1.4.11 Non-text Contrast (Level AA) requires at least 3:1 against adjacent colors for active input borders (when no other cue exists), checkbox ticks, custom focus rings, and meaningful standalone icons or chart lines. Because anti-aliasing thins light strokes, aiming for 7:1 or higher on body copy keeps pages legible under outdoor glare.

2. Target Size: 24×24 CSS px Floor vs. 44–48px Touch Targets

Designers and engineers often debate 24px versus 44px because each comes from a different level:

  • AA floor is 24 × 24 CSS pixels: W3C WCAG 2.2 SC 2.5.8 Target Size (Minimum) (Level AA) covers mice, styluses, and touchscreens, requiring an axis-aligned 24×24 CSS px square inside each target. Five exceptions apply: Spacing (undersized targets pass if an imaginary 24 CSS px circle centered on their bounding box does not intersect another target or undersized circle—so 20×20px buttons 4px apart pass, while touching 20×20px buttons fail), Equivalent controls, Inline sentence links, unmodified User Agent Controls, and Essential dense pins. Relying on user pinch-zoom never counts as passing.
  • Mobile ergonomics target 44–48px: For thumbs, W3C's Level AAA SC 2.5.5 Target Size (Enhanced) sets 44 × 44 CSS px; Apple's Human Interface Guidelines on Accessibility recommend 44 × 44 pt on iOS/iPadOS; and Google web.dev's accessible tap targets guide recommends around 48 device-independent pixels with about 8px spacing. Keep icons visually compact at 20–24px, and expand their clickable hit box via CSS padding to ≥ 44×44 CSS px.

3. Visible Keyboard Focus (SC 2.4.7)

Under W3C WCAG 2.2 SC 2.4.7 Focus Visible (Level AA), every keyboard-operable control needs a persistent visible focus indicator; stripping outline without a visible replacement fails directly (Failure F78), and custom rings must meet SC 1.4.11's 3:1 adjacent contrast. Use :focus-visible so mouse clicks stay clean while Tab navigation highlights a clear ring—ideally two-toned when pages mix white cards and dark navy sections.

4. Narrow-Screen Reflow, 16px Inputs, and Word Wrapping

W3C WCAG 2.2 SC 1.4.10 Reflow (Level AA) requires vertical pages to reflow at a 320 CSS px width (1280px zoomed 400%) without losing functionality or forcing horizontal scrolling across body text; wide data tables belong inside their own scrollable wrapper. Never block zoom using user-scalable=no or maximum-scale=1, which MDN's <meta name="viewport"> reference warns harms low-vision users (and modern iOS Safari ignores by default).

On top of that baseline, enforce three engineering rules: first, per Google web.dev's Learn Forms guide, keep mobile <input>, <textarea>, and <select> font sizes at ≥ 16px (1rem) so iOS Safari does not auto-zoom on focus; second, per Google web.dev's responsive typography guide, keep Latin single-column measure around 45–75 characters (66ch) with unitless line-height (1.5–1.75); third, follow MDN's overflow-wrap guidance by using overflow-wrap: break-word (or anywhere) for long URLs or model codes instead of word-break: break-all, which splits ordinary English words mid-syllable.

Four Hands-On Acceptance Checks You Can Run Right Now

  1. Check 1: Light-Gray Text, Placeholders, and Inverted Blue/Navy Cards Inspect footer captions, form placeholder text, ice-blue tinted cards, and dark navy banners in Chrome DevTools or Lighthouse. A hypothetical sRGB calculation illustrates why visual guessing fails: cool gray #94A3B8 on #FFFFFF gives only 2.56:1. Darker #64748B passes white at 4.76:1, yet drops to 4.47:1 on a #F4F8FF ice-blue card—failing W3C's unrounded 4.5:1 floor. Similarly, brand blue #1557F3 hits 5.69:1 on white but drops to 2.81:1 on #12213D navy, where you need crisp white (16.02:1) or bright cyan (#12C8D2 hits 7.8:1 on #12213D, though only 2.06:1 on white and decorative only).
  2. Check 2: Narrow 320px / 390px Wrapping and 200% Zoom Overflow Preview both Chinese and English pages at 320px and 390px widths plus desktop 200% zoom: confirm zero page-wide horizontal scrollbars; check that wider English buttons and SKUs wrap cleanly without overflowing or splitting ordinary words mid-word; and verify wide tables scroll inside their own wrapper without widening surrounding prose.
  3. Check 3: End-to-End Keyboard Walk (Tab, Shift+Tab, Enter, Esc) Set the mouse aside on core service pages and your consultation page. Tab forward and backward from header to footer: focus rings must be spotted within half a second across light and dark cards, never hidden behind sticky bars, and pressing Esc on drawers or popovers must dismiss them and return focus to the trigger.
  4. Check 4: Real-Phone Tap Zones and Form Input Focus Open your contact page and consultation form on a smartphone—especially iOS Safari. Tap every field to confirm no auto-zoom jumps occur; verify in DevTools Computed boxes that language toggles, × close icons, pagination pills, and checkboxes clear 24×24px (aiming for 44–48px on key touch actions); and tap checkbox <label> text to confirm it toggles the control.

Website Readability & Accessibility Acceptance Matrix

Audit Dimension

Mandatory WCAG 2.2 AA Floor

Engineering & UX Target

How to Test & Pass Signal

Body & Placeholder Contrast

Regular text, text images, placeholders, hover/focus text ≥ 4.5:1 (unrounded)

Aim for ≥ 7:1 on body text; avoid hairline weights on gray or tinted cards

Measure foreground vs. card background in DevTools; legible in daylight

Large Text Contrast

Large text (≥ 24px regular or ≥ 18.67px bold; CJK equivalent) ≥ 3:1

Keep ≥ 4.5:1 headroom on tinted cards or image gradients

Inspect lowest-contrast spot behind hero and section headings

UI Borders & Icons

Active input borders, checkboxes, states, informative icons ≥ 3:1

Pair subtle fills with clear solid borders rather than ultra-faint rules

Verify borders, dropdown arrows, and close icons hit ≥ 3:1 adjacent contrast

Target Size & Spacing

≥ 24×24 CSS px, or non-overlapping 24px spacing circle (5 exceptions)

Use padding to expand mobile buttons & icons to 44–48 CSS px, ≥ 8px apart

Inspect Computed dimensions in DevTools; test one-handed thumb taps

Keyboard Focus

Keyboard controls show persistent focus; custom rings ≥ 3:1 contrast

Use :focus-visible with a two-tone ring across white and navy cards

Walk Tab/Shift+Tab/Esc without a mouse; focus never hidden

Mobile Input Size & Zoom

Scales cleanly to 200%; never block zoom with user-scalable=no

Set <input>, <textarea>, <select> font size to ≥ 16px (1rem)

Zero auto-zoom jump in iOS Safari; pinch-zoom remains unrestricted

320px Reflow & Tables

Zero 2D horizontal prose scrolling at 320 CSS px; no lost content

Isolate wide tables in scrollable containers; wrap button groups cleanly

Check 320px, 390px, and 200% zoom for scrollWidth overflow

Measure, Leading & Wrapping

Enlarged text reflows without overlap or clipped lines

Keep Latin prose 45–75 chars (66ch), line-height 1.5–1.75, overflow-wrap: break-word

Verify narrow English views wrap long strings cleanly without mid-word splits

Two Audit Boundaries to Keep Clear

Avoid two common audit traps. First, do not waste time arguing over valid WCAG exceptions while ignoring real blockers: inline sentence links are exempt from 24px target height so paragraph leading stays intact, and disabled buttons or brand logotypes are exempt from 4.5:1 contrast. Direct fixes toward pale footer copy, faint input borders, small close icons, and missing focus rings. Second, bake these tokens into shared UI components early rather than patching pages one by one before launch, pairing cleanly with mobile navigation structure and real-device WeChat/browser testing.

If you are planning a website redesign or preparing for pre-launch QA, put the four checks above straight into your acceptance script. You can also explore our Corporate Website & Standalone Site Services to see how we build responsive accessible components, or reach out on our Contact page with your current site URL.

Sources

Let's talk about your project

Websites / Business systems / AI applications

Chat on WhatsApp