Definition of Done for UI Screens
If users have to guess, you’re charging them attention tax. Treat cognitive load like performance: a constraint you design against with clear hierarchy, plain
contents
Cognitive load used to be treated like polish.
A nice refinement. A UX “extra”. Something you handle after the interface works.
That framing is outdated.
Cognitive load is now a constraint – like performance, privacy, and security. If your UI is cognitively exhausting, it does not “mostly work”. It fails quietly, then shows up later as drop-off, errors, support tickets, refunds, and distrust.
And yes, the legal side is catching up. Accessibility standards explicitly include cognitive, learning, language, and neurological disabilities, not just vision or mobility.
This article is not legal advice. It is a design reality check: treat cognitive load as a risk surface and a cost center, and your work gets clearer, safer, and more defensible.
Why cognitive load is no longer optional
1) Cognitive load is a commercial constraint
When cognitive load is too high, three things happen:
Errors increase.
Users choose the wrong plan, misunderstand pricing, miss steps, or submit wrong information.
Support load increases.
Every “Where do I click?” message is the interface charging users for confusion.
Trust drops.
Confusing interfaces feel unsafe. Users assume hidden costs, poor service, or a scam, even if your product is solid.
Teams often don’t label these as “cognitive load problems”. They label them as “conversion issues”, “bad traffic”, or “users are careless”. But the pattern is consistent: the UI is asking users to work too hard.
2) Cognitive load is an accessibility constraint
WCAG 2.2 is explicit: accessibility involves a wide range of disabilities including cognitive, language, learning, and neurological.
Important nuance: WCAG does not cover everything cognitive users need. That is why the W3C has additional guidance focused on making content usable for people with cognitive and learning disabilities.
Translation for designers: if your interface relies on memory, dense reading, confusing language, time pressure, or surprise state changes, you are building barriers.
3) Cognitive load is moving into compliance reality
In the EU, accessibility is not just “best practice”. The European Accessibility Act is widely described as applying from 28 June 2025 and covers a range of products and services.
Even if your project is not directly in scope today, the direction is clear: accessibility requirements are increasingly tied to procurement and market access. When that happens, teams that treat cognitive load as optional create avoidable risk.
What cognitive load actually is
Cognitive load is the mental effort required to:
→ understand what is happening
→ decide what to do next
→ recover when something goes wrong
It becomes dangerous when the interface assumes the user has unlimited:
→ attention
→ memory
→ comprehension speed
→ emotional bandwidth
Real users do not show up with unlimited bandwidth. They show up tired, distracted, stressed, multitasking, in a second language, on a bad connection, in bright sunlight, with a baby in one hand, with anxiety, with neurodivergent processing, or all of the above.
Your interface should not require stamina to use.
The five overload patterns
These are not “bad UI” in the obvious sense. Many of them look modern and clean. That’s why they’re so common.
1) Decision stacks
When a screen presents too many choices at once, users do not “choose carefully”. They freeze, guess, or leave.
Common causes:
- multiple primary CTAs
- too many options with equal visual weight
- features listed before the user knows what matters
Design signal: if everything looks important, nothing is.
2) Memory traps
Any flow that requires remembering information from earlier steps is taxing users.
Examples:
- showing pricing terms on one screen, then asking users to choose later
- hiding important rules in a tooltip or modal the user has to recall
- requiring the user to compare options without allowing side-by-side viewing
Design rule: critical information must be present at the moment of decision.
3) Jargon nesting
“Enterprise-ready omnichannel workflows” might be accurate internally. It is not usable language for a user trying to choose a plan.
Jargon increases cognitive load because users must translate before acting. That translation is work. Work feels risky.
Design rule: if a user must decode your label, the label has failed.
4) Surprise states
When the UI changes and does not explain what happened, users lose orientation.
Examples:
- unexpected modals
- confirmations that do not confirm anything
- silent errors
- “success” states with no next step
Surprise states create a specific kind of cognitive load: uncertainty. Uncertainty is where trust dies.
5) Visual noise disguised as “premium”
A minimalist UI can still be exhausting if hierarchy is weak.
Visual noise is not just “too many elements”. It is:
- inconsistent emphasis
- headings that do not tell the story
- blocks with equal weight and no clear path
- dense layouts with no breathing room
If a user has to hunt for meaning, you are charging them attention tax.
How to design against cognitive load
“Make it simpler” is not a design instruction. Treat cognitive load like a constraint you can test.
Constraint A: First-time comprehension
Can a first-time user explain what this screen is for in 5 seconds?
If not, the screen is not clear enough to be safe.
Constraint B: Primary action clarity
Is the next step obvious without reading everything?
If users have to read the entire screen to find the action, the interface is fragile.
Constraint C: Decision count
How many real choices exist on this view?
Most screens should have:
- one primary action
- a small set of clearly grouped secondary actions
Constraint D: Recovery
If the user makes a mistake, does the UI tell them what happened and how to fix it?
Recovery is cognitive accessibility. It is also customer support prevention.
Constraint E: Memory independence
Can the user complete the task without remembering previous content?
If the answer is no, you are building a memory-based interface. Most people cannot afford that.
Cognitive Load QA:
Use this before handoff. Not as a vibe check. As a gate.
Clarity
- The screen has one main job, stated in plain language.
- The primary message is visible without scrolling.
- Headings tell the story even if body text is skipped.
Decision load
- One primary action per view.
- Secondary actions are grouped, not scattered.
- Options are not repeated in multiple places “just in case”.
Memory load
- The UI does not rely on the user remembering earlier steps.
- Key info is visible at the moment of decision (price, terms, next step).
State change safety
- The UI explains what changed and why.
- Errors state what went wrong and what to do next.
- Success states confirm the result and show the next step.
Sensory sanity
- Motion is minimal and avoidable.
- Dense blocks are broken with spacing and progressive disclosure.
The 10-minute Overload Trim exercise
Pick one screen. Set a 10-minute timer. Do not overthink. Just cut.
- Write the screen’s job in 8 words max.
- Circle everything that does not serve that job.
- Remove one entire cluster (not one button, a whole block).
- Rewrite the primary action label in plain language.
- Add one recovery hint: “What happens next” or “How to undo”.
This exercise is intentionally brutal. Cognitive load is not reduced by polishing. It is reduced by removing work.
Closure
Here’s the uncomfortable truth: most teams don’t “choose” high cognitive load. They drift into it.
A few extra features. One more CTA. A label that makes sense internally. A modal that felt efficient. A “premium” layout that looks clean but hides meaning. Then suddenly the product is beautiful and exhausting.
And the user does what users always do when they feel uncertain: they stop. They abandon. They hesitate. They ask support. They blame themselves. They leave quietly.
So treat cognitive load like what it is: a constraint, a cost, and a trust issue.
If you want a simple standard to carry into every project, use this:
If a tired person, on a small screen, in a second language, can still complete the task – you’re designing something that holds up.
Everything else is decoration.
Share this post
Andre Santoro
Join the Telegram community chat
Our Telegram community chat is where the Circle actually happens. We keep it friendly and low-pressure: weekly prompts, questions, small wins, and feedback when you want it. No gatekeeping. No judgement.
Get Circle Notes
Short emails with frameworks, breakdowns, and prompts. Decision Making, Information Hierarchy, Quality Control, UX Ethics, Design Systems, Career Practice, News.
Scope Guardrails – How to Say No Without Losing Trust
Guardrails don’t make you cold. They make you reliable. Here’s how to control scope without conflict.
When Everything Is Important, Nothing Works
A practical triage method to cut clutter, choose what wins, and make every screen feel obvious at first glance - without redesigning the whole
Cognitive load as a legal and commercial constraint, not a UX nicety
If users have to guess, you’re charging them attention tax. Treat cognitive load like performance: a constraint you design against with clear hierarchy, plain
How to Stay Stylish While Keeping It Comfortable
Discover simple tips to effortlessly blend comfort and style. From versatile wardrobe staples to chic accessorizing tricks, achieve a look that feels as good
Transform Your Space: Minimalist Home Décor Ideas
Create a serene and clutter-free environment with minimalist décor. Explore timeless design tips to transform your space into a stylish and peaceful retreat.
Quick and Healthy Meals for Busy Weeknights
Whip up delicious and nutritious meals in no time. Discover quick recipes that are perfect for busy weeknights and packed with flavor.
Simple Ways to Make Your Lifestyle More Eco-Friendly
Embrace sustainability with small changes that make a big impact. Discover easy ways to reduce waste, save energy, and adopt eco-friendly habits in your
Join the Circle
Circle Notes is where we share the frameworks we actually use, plus prompts you can reply to.
If it’s not useful, we don’t send it.
By subscribing, you agree to receive Circle Notes and accept our Terms of Use and Privacy Policy.