não acredite em ferramentas milagrosas: Do not believe in miracle tools

Select your language

Select your language

There are two ways to approach digital accessibility. Which is yours?

Accessibility has always been "left for later" because it has never been seen as a way to improve product quality, but as "something companies are required to include".

Using tools to fill gaps creates a false sense of compliance, while the actual experience remains compromised.

Scroll down to understand the differences between the paths chosen...

The path you choose directly affects the quality of your product!

Follow the paths and reflect...

Two hands in black and white offer a red pill and a blue pill, referencing the film The Matrix. In the animated version, the hands turn over and the pills fall as you scroll.

Two choices. Different consequences.

What happens after the choice?

Follow five stages of a digital product. On one side, accessibility as part of the work. On the other, the expectation that an overlay will solve everything.

  1. The initial decision

    The promise of a shortcut

    “Just install it.”

    The team installs a tool and considers accessibility resolved, without investigating the product's barriers.

    Accessibility in practice

    It starts with understanding.

    The team identifies barriers, sets priorities and takes responsibility for addressing them.

  2. Preparing the teams

    The promise of a shortcut

    “The tool takes care of that.”

    Design, content and development teams continue working without training. Decisions that create barriers keep being repeated.

    Accessibility in practice

    Knowledge is part of the solution.

    Teams receive training to recognize barriers and incorporate accessibility into everyday decisions.

  3. Building the product

    The promise of a shortcut

    “We will adapt it later.”

    The product reaches implementation without accessibility requirements. The expectation is that a layer added to the website will make up for what was left out.

    Accessibility in practice

    Accessibility from the start.

    Requirements, behaviors and acceptance criteria are documented from the design stage and guide development.

  4. Checking the experience

    The promise of a shortcut

    “The button appeared. It is ready.”

    The presence of the feature is treated as proof of accessibility, without checking whether people can complete tasks.

    Accessibility in practice

    The experience needs to be checked.

    QA participates from the start. Automated and manual tests, assistive technologies and participation by people with disabilities help identify barriers.

  5. Ongoing work

    The promise of a shortcut

    “We have already solved that.”

    New pages and features are published without an accessibility review. Trust in the tool replaces ongoing product monitoring.

    Accessibility in practice

    Every change matters too.

    The team follows up on reports, checks for regressions and updates components and documentation to sustain improvements.

A long road...

Accessibility is far (very far) from being about installing a tool.

It is important to understand that building better products does not involve ready-made formulas,
but contexts of use...

Understand the tool

It is a tool for ADJUSTMENTS, not an accessibility solution.

Some accessibility tools add a settings panel to the page. They are known as overlays: a layer of features over the website you are using.

  1. The website remains the foundation. Content, structure and behaviors need to be accessible.
  2. The tool adds adjustments. The panel may include options such as text size and colors. Features vary between products.
  3. A button opens the panel. This is how these tools typically appear: an icon on one side of the page.

Most of the added features already exist in operating systems and browsers.

Simplified representation of two layers: the website and the tool's button over the page. The icon in this illustration is not a functional button.

Features, alternatives and limitations

What do these tools
actually add?

Compare the features offered, the options already available on devices and the barriers that still require work on the website itself.

Compare feature by feature

Feature inventory

Many tools present having a hundred features as a selling point. But how many of those features actually affect people's experience or need to be there?

The following table shows some features offered by overlay tools and compares them with features available in operating systems or browsers. Keep in mind that redundancy alone does not necessarily mean a better or worse experience; this is simply a comparison table.

Features, alternatives and limitations
Feature offered Alternative or responsibility Limits of the comparison
Increase and decrease text size High redundancy Browser zoom; text size in Windows, macOS, iOS and Android Page zoom and text size have different effects. Support for adjustments varies between websites and apps.
Screen magnifier High redundancy Windows Magnifier, macOS and iOS Zoom, Android magnification System magnification can cover the entire screen; a magnifier inside a website has limited reach.
Larger cursor High redundancy Pointer settings in Windows and macOS System settings apply to other apps. Website-customized pointers may behave differently.
Change cursor color High redundancy Pointer customization in the operating system Available options depend on the system and its version.
High contrast Significant redundancy Windows contrast themes; contrast options in macOS and iOS Contrast themes, increased contrast and recoloring a page have different scopes.
Invert colors High redundancy Color inversion in macOS, iOS and Android The treatment of images, videos and interface elements may vary. Inverting colors does not guarantee adequate contrast.
Grayscale High redundancy Operating system color filters Removing colors does not resolve information that relies solely on color to be understood.
Color blindness filters Significant redundancy Color filters in Windows, macOS, iOS and Android Filters and their results vary. Content still needs identification methods beyond color.
Dark mode Significant redundancy Appearance preferences in operating systems and browsers A dark system or browser appearance does not automatically transform every website. The result depends on the page's support.
Read text aloud High redundancy System speech features, read-aloud features and reading modes Listening to text is not equivalent to navigating and operating an interface with a screen reader. Available voices and languages vary.
Screen reader Specialized free alternative VoiceOver, NVDA, Jaws, TalkBack, Narrator, among others. Screen readers provide navigation and interaction as well as speech. The website needs compatible semantics and behaviors.
Sign language avatar Specialized free alternative Avatars may be considered useful in specific scenarios.
Example: VLibras (for Brazilian Sign Language) (opens in a new tab)
Sign language avatars can help some members of the Deaf community, and their features can be used independently of other tools with some usefulness in certain scenarios. However, prefer solutions that prioritize human interpreters.
Voice commands Significant redundancy Windows Voice Access, Apple Voice Control and Android Voice Access Availability depends on the system, version and language. Commands limited to a website are not equivalent to controlling the device.
On-screen keyboard High redundancy On-screen keyboards are available in operating system settings The keyboard enables data entry but does not fix components that cannot be reached or operated properly.
Reduce animations Partial redundancy Operating system reduced-motion preferences The preference does not guarantee that all movement on a page stops. The website needs to respect it and provide suitable controls.
Simplified reading mode Significant redundancy Reading modes in browsers and Android Not all pages are compatible. A simplified presentation does not replace operating forms and other components.
Reading mask Partial redundancy Line focus features in browser reading modes The mask visually reduces the content around a reading band. Availability of the alternative depends on reading mode; the feature does not fix content structure and must allow adjustment and disabling.
Reading guide Partial redundancy Line tracking features in browser reading modes A ruler or visual line helps mark the reading position. It is not necessarily equivalent to line focus in a reading mode or to keyboard focus; it must not obscure content or prevent interactions.
More readable font Partial redundancy Font customization in the browser and some reading modes Font application varies between pages and reading modes. No single font meets everyone's needs.
Eye or facial tracking Partial redundancy Windows eye control and facial-expression controls in compatible systems Eye control and facial control are distinct features. They may require specific hardware and depend on the device and its version. Note: It is important to know whether screen recordings are being stored, due to requirements related to LGPD, GDPR, CCPA or equivalent laws.

The features in this table are not exhaustive and are intended only as an initial comparison.

Where can you find documentation and examples of similar features?

Feature offered Synonyms and similar feature names Documentation and free extensions
Increase and decrease text size Font size; text size; page zoom
Screen magnifier Magnifier; magnification; screen zoom
Larger cursor Enlarged pointer; pointer size; large cursor
Change cursor color Pointer color; custom pointer
High contrast Contrast themes; increase contrast; high contrast
Invert colors Color inversion; inverted colors; invert colors
Grayscale Grayscale; monochrome filter
Color blindness filters Color filters; color correction
Dark mode Dark theme; dark appearance; dark mode
Read text aloud Read aloud; speech synthesis; text-to-speech (TTS)
Screen reader Screen reader; screen reading; Narrator
Sign language avatar Libras avatar; automatic sign language translation; VLibras
Voice commands Voice access; voice control; voice commands
On-screen keyboard On-screen keyboard; accessibility keyboard
Reduce animations Reduce motion; animation effects; reduced motion
Simplified reading mode Reading mode; reader view; immersive reader; reader mode
Reading mask Screen mask; reading window; line focus; reading mask
Reading guide Reading ruler; line guide; reading guide
More readable font Customize fonts; font family; font style
Eye or facial tracking Eye control; head pointer; eye tracking; head tracking

These names help you search for similar features, but do not imply equivalence in every context. Availability depends on the system, version and content; some links are in English.

Behavioral analysis

Reflective questions to guide your decision

The name says it: FUNCTIONALITY. Does this function actually help?

Feature Key question
Keyboard navigation Why should it be necessary to enable a feature to operate something that should work naturally with a keyboard?
Screen reader compatibility Why was the website not developed to work correctly with the screen readers people already use?
Automatic image descriptions How can you ensure that an automatically generated description accurately represents the purpose of the image?
Highlighting links and headings If elements cannot be identified, why not fix their presentation and semantics in the original interface?
Disability profiles Why assume that people with the same disability need the same adjustments?
Text spacing adjustment If someone changes the spacing, does the layout remain functional without loss of content?
Enlarging text and content When content is enlarged, can the person still read, access controls and complete their task without overlap or loss of information?
High contrast and color inversion Does the adjustment preserve the identification of fields, error messages, focus and control states, or improve one part of the page while making another harder to use?
Color filters and grayscale If information relies solely on color, does the filter resolve that barrier, or are text, symbols and other ways to identify it still missing?
Read text aloud Does reading aloud help people understand the content, and can it be paused and stopped without competing with the screen reader or other sounds the person already uses?
Animation reduction When movement stops, do carousels, menus and other components remain operable? Why not respect the system preference from the start?
Simplified reading mode Does simplification preserve essential information, reading order and necessary links, or remove parts the person needs?
Voice commands Can the person activate controls using the names shown on screen, without having to learn commands unique to this tool?
Changing fonts Does changing the font respect the person's preference and preserve the readability of all characters, without clipping text or disrupting components?

Consider the effort

The number of features is NOT a measure of accessibility

It is not uncommon to hear claims that the number of options offered is a positive factor for accessibility. But quantity will never be a measure of quality, especially when we consider various studies linking the number of options to increased cognitive load, such as this 2023 study (opens in a new tab), for example.

A possible sequence of attention, interpretation and decision-making when using a preferences panel:

Select a step to explore an example. Select it again to collapse it.

Stages of cognitive effort before returning to the task

  1. Find the featureInterrupt the task and locate the settings.

    Example situation

    You are filling out a form and need larger text. Before continuing, you look for the button that opens the preferences panel.

  2. Understand the optionsInterpret the names and imagine the effect of each alternative.

    Example situation

    The panel offers “Contrast”, “Saturation” and “Highlight”. You read the labels and try to work out which option will make the text easier to read.

  3. Choose an adjustmentCompare possibilities and decide which might meet the need.

    Example situation

    You compare increasing the font size, zooming in on the page and changing the spacing to decide which option to try first.

  4. Test the resultEvaluate the effect and, if necessary, reconsider the choice.

    Example situation

    You increase the text size, return to the form and check whether it is easier to read. If it does not help, you return to the panel to try another setting.

  5. Return to the taskRefocus and remember where you left off.

    Example situation

    After closing the panel, you look for the field you were working on and recall the information you intended to enter.

Learn from good ideas

Look at major companies' websites.
They are accessible and do not display "adjustments"

Developing products and services also involves recognizing and using good practices. Just as some good practices from major companies are adopted, also adopt their approach of developing products consistently by training their teams.

Important actions and positions involving governments and organizations in different countries

European Commission

The European Disability Forum, users with disabilities and many accessibility experts have expressed concerns about the use of overlays.

There are many more references to explore. Scroll a little further for a list.

For deeper study

Additional references

The following links include research, companies sued despite using tools, personal accounts, detailed analyses and information that can inform arguments and decisions about different types of tools.

  1. Overlay Fact Sheet (opens in a new tab) — how overlays work, their capabilities and limitations.
  2. Overlay False Claims (opens in a new tab) — critical documentation of commercial promises.
  3. Should I Use An Accessibility Overlay? (opens in a new tab) — arguments about unresolved barriers, assistive technologies and the risks of relying on overlays.
  4. The A11Y Project: Should I use an accessibility overlay? (opens in a new tab) — differences between temporary fixes and permanent plugins, including effects on autonomy, privacy and quality.
  5. Daniela Kubesch: The Impact of Web Accessibility Overlays (opens in a new tab) — master's research with technical evaluation and a usability study involving 21 people with visual disabilities.
  6. Lainey Feingold: Why Accessibility Overlays Don’t Work (opens in a new tab) — results of overlay tests on US state government websites.
  7. Scope: Why accessibility overlays do not improve site accessibility (opens in a new tab) — widget limitations and possible conflicts with assistive technologies.
  8. Sheri Byrne-Haber: Overlays are not the solution to your accessibility problem (opens in a new tab)
  9. Sheri Byrne-Haber: More lawsuits filed against companies using accessibility plugin/overlays (opens in a new tab)
  10. Sheri Byrne-Haber: The four pillars of accessibility overlay marketing (opens in a new tab)
  11. Siteimprove: A guide to accessibility overlays (opens in a new tab)
  12. Karl Groves: Selling Accessibility – Negative Factors (opens in a new tab) — misconceptions, fear and practices that hinder support for accessibility within organizations.
  13. FTC: settlement with accessiBe over accessibility claims (opens in a new tab) — announcement of a $1 million settlement to resolve allegations of deceptive advertising.

Videos

  1. The Underlying Truth about Overlays (opens in a new tab)
  2. Overlays Underwhelm / Adrian Roselli #id24 2022 (opens in a new tab)
  3. 5 False Claims 1-Line "AI" Accessibility Script Vendors Make (opens in a new tab)
  4. What are accessibility overlays? (opens in a new tab)
  5. Avoid companies claiming quick & easy AI accessibility (opens in a new tab)
  6. Accessibility Overlays: What You See You Already Have (opens in a new tab)
  7. ICT2020 Overlays Panel (opens in a new tab)
  8. Can a website be made accessible with just one line of code? (in Portuguese) (opens in a new tab)
  9. Do not trust "miracle" accessibility tools (in Portuguese) (opens in a new tab)

Further reading

  1. Digital Government: VLibras documentation (opens in a new tab) — reference on automatic translation and the caveat that it does not replace a human interpreter.
  2. Scheibehenne, Greifeneder and Todd (2010): Can There Ever Be Too Many Options? (opens in a new tab) — meta-analysis of choice overload (PDF, in English).
  3. Chernev, Böckenholt and Goodman (2015): Choice overload (opens in a new tab) — factors influencing choice overload (in English).
  4. W3C: Avoid Too Much Content (opens in a new tab) — cognitive accessibility guidance, supplementary to WCAG (in English).

Open-source adjustment tools

Some open-source options for your projects

Preferências de Acessibilidade

Based on the same options available in operating systems and browsers, we built a free, open-source ADJUSTMENT tool that you can freely use, recreate and adapt in your projects.

Just remember that it should be described as an ADJUSTMENT tool, not as an accessibility solution.

Preferências de Acessibilidade website with its panel open, showing zoom, contrast, font, text size and spacing adjustments.

Explore Preferências de Acessibilidade (opens in a new tab)

When unchecked, the tool will be hidden and adjustments will be reset.

Other free, open-source options

We do not endorse the use of tools, but there are reasons why some people may want to use them.

Just remember that offering adjustment tools is different from making digital products or services accessible.