Project overview

This thesis project explored how accessible UX writing can improve the usability of IKEA’s internal product-information tool. Over three weeks, I audited microcopy, labels, and semantics against WCAG 2.2 AA, focusing on inclusive language and assistive-technology support. The result was clearer, more consistent content and actionable guidelines for accessible writing within IKEA’s design system.

 

Notice: This project is covered by a non-disclosure agreement (NDA). Certain details — including internal tool names, system identifiers, and brand-specific visuals — cannot be shared publicly. The confidentiality requirements also limited the possibility of conducting user research and usability testing outside IKEA internal environment.

Project goal

Project main goal was to improve clarity and accessibility in the internal tool’s interface to help IKEA co-workers work more efficiently and independently. Reduce friction and exclusion caused by unclear copy and missing semantics, ensuring the tool is fully usable with keyboard and screen readers and easier to understand for non-native speakers and users with cognitive or visual impairments.

My role

UX designer focused on UX writing and accessibility

Tools

Figma, VoiceOver, Microsoft Teams, IKEA design system, IKEA internal guidelines, WCAG 2.2 AA, UX best practices

 

Time frame

February 2025

Research

Methods

  • Theoretical research – reviewing WCAG 2.2 AA, POUR, inclusive and universal design principles, IKEA internal accessibility guidelines, and best UX practices.
  • Content audit – analysing microcopy, labels, placeholders, alt text, and abbreviations to identify accessibility barriers.
  • Assistive-technology testing – using the VoiceOver screen reader to detect navigation, labeling, and responsiveness issues.
  • User and expert feedback – gathering insights through Accessibility Fika sessions with IKEA employees (including users with disabilities) and internal experts in UX, writing, and development.

Identified issues

  • Abbreviations and acronyms in filters and buttons caused confusion.

  • Frequent ALL-CAPS and capitalised microcopy reduced readability.

  • Missing or unclear labels and inconsistent placeholders made navigation unpredictable.

  • Keyboard navigation and screen-reader support were incomplete or inconsistent.

  • Interface relied on colour alone to communicate states.

  • Language and retail-unit selectors were not announced to assistive technologies.

  • Image alt texts were long, system-generated strings that provided little meaning.

  • Use of typographic signs (like “&”) in labels caused incorrect or missing pronunciation by screen readers.

Impact on co-workers and users with disabilities

Blind or low-vision users

Struggle to navigate due to unlabeled controls, missing roles, and poor contrast

Colour-vision deficiencies

Can’t perceive status information shown only by colour.

Cognitive differences

Face higher effort and more errors from abbreviations and inconsistent copy.

Keyboard-only users

Can’t move smoothly through the page because some elements don’t get focus or trap the cursor.

Screen-reader users

Struggle to understand page structure because some buttons and labels aren’t described properly. Symbols like “&” or “/” are read aloud in full, which makes names and phrases sound confusing.

New co-workers

Often face confusion and slower onboarding when the tool uses unclear terms, abbreviations, making it harder to learn and navigate independently.

Non-native speakers

Struggle to understand unclear or inconsistent wording, especially in elements like the language selector or filters. Ambiguous terms and abbreviations make it harder to find the right option and complete tasks confidently.

All co-workers

Experience reduced efficiency and independence due to unclear, inconsistent copy and use of abbreviations, especially in time-pressured environments like stores where quick task completion is essential.

Content design workflow

During the work process I collaborated closely with accessibility specialists, designers, writers and developers to ensure every change in wording or component behaviour improved both clarity and technical accessibility. Together we reviewed proposed updates, verified them against WCAG standards, and discussed implementation details in the design system.

Before and after

Key problems:

  • No label.
  • Placeholder text is too long and lacks clarity.
  • Capitalization.
  • No keyboard navigation and screen readers support.

What has changed:

  • A visible label added for clear context.
  • Placeholder text shortened.
  • Capitalization replaced by sentenced case.
  • "X” (clear) button in placeholder added to delete typed text if needed.

Key problems:

  • Unclear and vague microcopy.
  • Capitalization.
  • Lack of keyboard navigation and screen readers support.

What has changed:

  • Clear and descriptive labels used.
  • Capitalization replaced by sentence case.

Key problems:

  • Language selection component (dropdown) is inconsistent.

  • Unclear labeling - vague division for English language into "Official English" and "English".
  • All languages listed in English by default.
  • No ARIA support for dynamic updates - when a retail unit is selected, the language list updates, but screen readers do not announce the change.
  • Lack of proper keyboard navigation.

What has changed:

  • Language selection component (dropdown) became consistent due to using component from design system.
  • Clear labels used for division for English - standard and localized English variants: English (US), English (UK), English (Canada), etc.

  • Language names used in their native form, e.g., "Français" instead of "French",

    “Svenska” instead of “Swedish”, etc.

  • If there is English in the list, it comes first by default.

Key problems:

  • Unclear and vague microcopy.
  • Capitalization.
  • A lot of abbreviations without explanation.
  • Lack of keyboard navigation and screen readers support.

What has changed:

  • Radio buttons were replaced by checkboxes providing a user a multi-optional choice. As user can choose more than one alternative there is no need to use unclear option "Yes/Not Yet" anymore.
  • Capitalization replaced by sentence case.
  • Tooltip on hover added for abbreviations.

Creating guidelines

The final step of the project was transforming all research findings and accessibility issues into a set of clear, actionable UX writing guidelines. These guidelines translate accessibility principles into practical rules for everyday work, helping designers, writers, and developers create consistent, inclusive interfaces across IKEA internal tools. Here is an excerpt from the guides.

General writing guidelines

Labels and microcopy

Placeholder text

Alternative text

Outcomes

Recommendations for future work

  • Test with more diverse co-workers and roles.
  • Include additional assistive technologies (e.g. NVDA, JAWS).
  • Refine keyboard navigation and focus order.
  • Improve screen-reader feedback for all interactive elements.
  • Include accessibility in all development stages.
  • Build a shared accessibility library for future projects.

Reflection

This project strengthened my understanding of how language directly impacts accessibility and daily workflows. Working within IKEA internal environment showed the importance of collaboration between roles and how even small text or structure changes can remove barriers. With more time and broader testing, the results could evolve into a scalable model for accessible writing across all IKEA digital tools.