People. Not edge cases.

Adaptations & assistive technology

• Screen readers
• Refreshable braille
• Voice control & dictation
• AI visual assistance

Friction patterns we hear about

• Unlabelled buttons, icon-only controls, and decorative imagery without alt text
• Custom widgets that bypass the platform accessibility tree (drag-and-drop, canvas, custom dropdowns)
• Form errors announced visually but not programmatically, so submission appears to fail silently
• Heading structure that is purely cosmetic, breaking the screen reader landmark map
• Inaccessible PDF receipts, statements, and authentication challenges (CAPTCHA)

Adaptations & assistive technology

• Magnification
• Dark mode
• High contrast
• Large text
• Inverted colours
• Screen readers (paired)
• Voice control & dictation
• AI visual assistance

Friction patterns we hear about

• Text that loses meaning when scaled to 200%; content reflows but key labels disappear off-canvas
• Insufficient colour contrast on body copy, links, focus rings, and form-field borders
• No dark mode for participants who get headaches or abandon a flow because of glare and inverted colours
• UI that depends on subtle hover states or thin 1 px dividers that fade under magnification
• Modals and tooltips that anchor at fixed coordinates so the relevant content never lands inside the magnified viewport
• Information conveyed by colour alone (red error states, green success indicators)

Adaptations & assistive technology

For many Deaf participants, sign language is a visual language and their first language. Written English is a second language, so dense paragraphs and complex terminology create more friction than audio alone. Iconography and step-by-step guides reduce the reading load and make complex flows scannable visually.

• Hearing & audio aids
• Live captioning
• Interpretation
• Visual & haptic alerts
• Plain-language tools
• Iconography & step-by-step guides

Friction patterns we hear about

• Complex terminology, idioms, and dense paragraphs of written English. Sign language is a visual language and the first language for many Deaf participants, so heavy text adds significant load
• Long-form instructions without iconography or visual step-by-step guides to anchor each stage of a flow
• Auto-play videos with audio narration but no captions or transcript
• Customer-service flows that route to a phone call with no chat, RTT, or sign-language alternative
• Notifications that fire only as a system sound, with no visual alert, vibration, or in-app banner
• Auto-generated captions that lag, mis-spell critical words (medical, legal, financial), or are switched off by default

Adaptations & assistive technology

Over-65 participants are less likely to use a named screen reader, but very likely to lean on system-level adjustments: larger system font, increased contrast, browser zoom, the operating-system magnifier, and voice control. Many have hearing aids paired to their phone for calls, even if they don’t self-identify as having a hearing impairment. Confidence matters as much as capability for this cohort: poor service design is exacerbated when patterns are unpredictable or when critical visual cues are missing, and small frictions compound into abandonment faster than they would for younger participants.

• Visual display adaptations
• Magnification
• Hearing aids (paired)
• Voice control
• Accessibility services

Friction patterns we hear about

• Unpredictable or inconsistent navigation patterns that change between sections of the same product
• Absence of critical visual cues (where am I in this flow, what just happened, what is the next step)
• Multi-factor authentication that times out before a code can be retrieved from a different device
• Small targets and dense layouts that fight against age-related motor and visual changes
• Jargon and acronyms that assume prior digital literacy
• Modal patterns that hijack the back button or trap focus unexpectedly
• Aggressive session timeouts that destroy work in progress with no recovery path

Adaptations & assistive technology

Limited English participants rely on translation, AI interpretation, iconography, step-by-step guides, and simple content.

• Translation
• AI interpretation
• Iconography
• Step-by-step guides
• Plain-language / simple content

Friction patterns we hear about

• Complex language: idioms, financial jargon, and government-speak that survive direct translation but lose meaning
• Walls of text that overwhelm reading-as-a-second-language comprehension and translation tools alike
• Images that relay information without a text equivalent or alt text, so translation tools have nothing to interpret
• In-app translation that switches the UI but leaves form labels, error messages, and PDFs in English
• Address, name, and ID fields that assume Western formatting (no second-family-name field, fixed postcode regex)
• Date and number formats that flip between locales without the participant noticing (DD/MM vs MM/DD, comma vs period)
• Customer service chat that hands off to a human who only speaks English

Adaptations & assistive technology

• Visual display adaptations
• Reading & cognition
• Accessibility services
• Plain-language tools
• AI focus & summarisation

Friction patterns we hear about

• Auto-play video, looping animations, and parallax scroll that overwhelm sensory bandwidth
• High-saturation accent colours used at scale, especially red error states or pulsing CTAs
• Aggressive deadlines (countdown timers, "you’ll lose your spot in 30 seconds")
• Implied-meaning copy that demands social inference: surprising flows, ambiguous CTAs, "obvious" next steps
• Inconsistent navigation patterns that change between sections of the same product