Picture this. You’re cooking your famous chilli for your best friend. You’ve pulled out the ingredients, you know the recipe by heart, and you’re ready to go. Now imagine a helpful stranger walks into your kitchen, looks around, and just... starts doing things. They preheat the oven, and they dice your onions . Great, that’s helpful! But then they start mixing in their own spices, they decide to use ground turkey instead of beef, and they substitute kidney beans for pinto beans.

Sure, the chilli will be made, but it doesn’t feel like yours anymore.

This is what bad AI design feels like. Not because the AI made bad choices, but because it made choices that it wasn’t supposed to make. And as designers building AI-powered products, this is one of the most consequential decisions we face: not what the AI does, but what it should be allowed to do on its own.

Automation Should Be a Spectrum

The instinct in most AI product teams is to frame automation as binary. Either the AI does it, or the user does. But that’s not how it works in practice, and designing as if it were leads to products that can be frustrating and sometimes even dangerous.

Automation is a spectrum. Researchers have known this since 1978, when Thomas Sheridan and William Verplank at MIT published what became the foundational framework for human-machine interaction: a scale from full human control to full computer autonomy, with meaningful gradations in between. It was built for undersea robotics and aviation. The principles apply just as well to AI product design.

On one end of the spectrum, the AI suggests, it offers recommendations and waits. Think of the Smart Compose feature in Gmail, when you type something, it suggests a sentence; you can press tab to accept the suggestion or simply keep typing to ignore it. On the other end of the spectrum, the AI acts autonomously, silently, and without asking. The interesting design decisions don’t live at either extreme, but somewhere in the middle.

The first is how. How should the AI approach the task? What tone, what format, what level of confidence should it project? These are questions of execution, and they’re the ones most product teams spend their time on.

The second is if. Should the AI complete this task at all? And if so, in what way, and at what point does it stop and hand back control? These are harder questions, and more consequential ones. A feature that answers how beautifully but gets if wrong can still cause real harm; it can send the email you didn’t mean to send, make the edit you wanted to review first, or take the action you weren’t ready for.

At the very heart of this is the real trade-off that designers need to make when it comes to designing for AI automation: Speed vs Responsibility.

The Trade-Off

Every step toward more automation is a step toward faster outcomes and reduced user accountability. Every step toward more control is a step toward slower outcomes and higher user ownership. This is the fundamental trade-off across every AI feature. The question designers need to answer is where on that spectrum any given feature should sit, and why.

The research bears this out. A 2014 meta-analysis of automation studies found that while higher automation reliably improves routine performance and reduces workload, it creates a serious problem when things go wrong: users who’ve been out of the loop struggle to take back control precisely when it matters most. Psychologists call this the “out-of-the-loop” phenomenon. The faster the AI has been operating, the harder the handoff back to the human. A product that automates completely is implicitly saying: we own the outcome. A product that keeps the human in control is saying: you do. Most AI features sit somewhere in the middle, which means designers need to be honest about how much responsibility they’re quietly shifting away from the user and whether that shift is appropriate for what’s at stake.

And what’s at stake depends on two things: who’s using it, and what they’re doing with it.

Novices Want Speed, Experts Want Control

The right position on the automation dial isn't the same for everyone. It can change depending on the user’s skill level, goals, and responsibilities.

Novice users tend to want more automation. They’re trying to close a gap between what they know and what they need to produce. An AI that fills in, simplifies, and handles the hard parts feels genuinely useful. The cognitive cost of manual control — understanding the options, making the call, taking responsibility for the outcome — is a cost they’d rather not pay yet. Automation lowers the floor and gets them to a result.

Expert users tend to want less. They already know how to do the thing. What they want is to do it faster, with less friction. The moment an AI overrides their judgment, produces output they didn’t sanction, or makes a decision they would have made differently, it creates frustration. It’s in the way. Experts have what psychologists call a strong internal locus of control: a deep sense of agency over their work. Automation that bypasses them doesn’t feel like a shortcut.

This has a practical implication that most AI products ignore: the optimal place in the spectrum of control is a moving target. A useful design pattern is to start new users with higher automation defaults and progressively surface manual controls as they grow in confidence. Let the product earn the trust and right to automate. Provide users with the ability to control their level of automation during onboarding, so novices and experts can start exactly where they want.

High Stakes Always Mean High Control

When the outcome of an AI action is consequential: medically, financially, or legally, these are the situations where the human in the loop is a design requirement.

There’s a well-documented cognitive phenomenon called automation bias: our tendency to over-trust automated systems, even when we have information that contradicts them. The more confident an AI appears, the more likely users are to defer to it uncritically. In low-stakes contexts, this is tolerable. In high-stakes ones, it can be catastrophic.

Consider AI diagnostic tools in healthcare. These systems can analyse imaging, flag anomalies, and identify patterns a human eye might miss. These tools can be extremely useful and valuable. But a radiologist who accepts an AI’s recommendation without scrutiny has abdicated judgment, and the AI cannot bear that responsibility in their place. These functional high-stakes AI have a defining characteristic: the output is factual, verifiable, and the error has a clear victim. In these types of applications, transparency and override control are absolutely essential because a single error can cost someone their health, money, or freedom. The AI’s job in these cases is not to make decisions, but to provide accurate information to inform the decision.

High-stakes creative or personal AI (writing a eulogy, designing a brand identity, producing music for release) is different in a subtler way. There’s no objectively right answer. If an AI writes your eulogy for you, it might be a perfectly good eulogy. but it isn’t yours. The emotional weight of that piece comes from the fact that you found those words. Automate it, and you’ve solved the wrong problem. The design implication here is about augmentation over generation. The AI should sharpen your thinking, surface possibilities, remove friction, not produce the finished thing. Control matters here not because mistakes are dangerous, but because the process is the point.

The Essential Feature: Undo

Wherever your product sits on the control spectrum, one principle is non-negotiable: users must always be able to go back.

The more autonomous the AI, the more essential undo becomes. It builds trust, encourages experimentation, and most importantly, provides a failsafe for when the AI gets it wrong. And it will get it wrong, because that’s how LLMs are by design.

Most AI tools today are powered by large language models, and LLMs are non-deterministic by design. This means if you run the same prompt twice, you’ll likely get two different responses. At its most benign, this produces useful variability. At its most disruptive, it produces what the industry calls hallucinations: outputs that are confident, fluent, and factually wrong. The model didn’t malfunction; it has no concept of right or wrong, only probability. It just landed on the wrong part of the distribution.

This is the thing that makes undo categorically different in AI products than in traditional software. A deterministic system makes predictable mistakes, the same bug triggers the same error, and you can design around it. A probabilistic system can fail in ways nobody anticipated, including the people who built it. If the AI can act, the user must be able to reverse it, cleanly, immediately, and without penalty for having trusted the system in the first place.

Good AI is a Great Sous Chef

The best sous chef in the world knows two things: how to cook, and when to stay out of the way. They prep, they assist, they catch things you’d miss. But when you’re plating the dish, they hand you the spoon.

AI works exactly the same way. The question isn’t how much can we automate, its how much does this person want automated, right now, for this specific task. And the honest answer to that varies by user, by task, by stakes, and by how much someone’s own judgment and voice are wrapped up in the outcome.

Design for the spectrum of control thoughtfully, default to augmentation over automation, and let users adjust their own level of trust. And always, always let them go back.

That tiny animation wasn’t just decoration, but an intentionally designed interaction. It was doing two things at once: confirming that your action registered, and making the confirmation feel good. Most microinteractions only do the first part. The best ones manage both, and the difference between those two outcomes is what separates products that feel functional from products that feel alive.

The Job Microinteractions Were Built to Do

A microinteraction is a single-task interaction built around one contained moment, a tap, a toggle, a swipe, or a form submission. The concept was defined by product designer Dan Saffer, and at its core, it describes something simple: every action a user takes deserves a response from the system.

The reason this matters goes back to basic psychology. Our brains are prediction machines. We constantly scan our environment to understand cause and effect, and when we take an action and get no visible response, something feels off. Not dramatically wrong, just slightly unsettled. It’s the same feeling you get pressing a crosswalk button with a dead light, or submitting a form and staring at a screen that hasn’t changed. You find yourself asking: Did that actually work?

Microinteractions exist to answer that question before it forms. They close the feedback loop, the psychological mechanism that makes systems feel predictable and trustworthy. When the loop is open, users hesitate, second-guess, and sometimes abandon. When it’s closed, they move forward with confidence.

Feedback First, Always

The foundation of any microinteraction is feedback: some signal that the system heard you. This can be visual (a button changing color on press), auditory (the iOS “sent” whoosh when an email leaves your outbox), or haptic (the subtle vibration when you toggle a switch on your phone). Often, the most effective designs layer all three together, because redundancy ensures that even if one channel is missed, another catches it.

What’s easy to underestimate is how much the absence of feedback costs you. Research from Nielsen Norman Group shows that delays of even 100 milliseconds can disrupt a user’s sense of direct manipulation, that feeling that they are controlling the interface rather than waiting on it. The implication for designers is that feedback doesn’t just need to be present, it needs to be immediate. A spinner that appears half a second after a button press still creates doubt. A state change that happens the moment your finger lifts does not.

Progress and loading states are a particular place where designers can either earn trust or quietly lose it. A blank screen while content loads feels like nothing is happening. A skeleton screen (the placeholder layout that mimics incoming content) feels like the product is already at work on your behalf. The difference in actual load time is zero. The difference in perceived wait time is significant, because our brains interpret visible activity as progress, even when it’s only a representation.

The same logic applies to form validation, error states, and success confirmations. A "Message sent" banner that slides in after you submit a contact form isn't just a nice touch; it's closing a loop that would otherwise stay open. I’m sure you’ve had times where that confirmation was missing, and you’ve found yourself asking, “Did it work? “Do I need to resubmit?” That’s the cost of leaving the loop open.

Where Feedback Becomes Delight

Here’s where it gets more interesting. Once feedback is doing its job reliably, you have an opportunity to make it do something more: to make the interaction feel good, not just correct.

This is the realm of motion design and delight, and it’s one of the most misunderstood areas of product design. Delight doesn’t mean playful animations everywhere, or confetti cannons on every button press. It means using motion with intention, to communicate something that a static state change cannot.

Consider Duolingo’s correct answer animation. The green glow, the cheerful sound, the way the screen seems to briefly celebrate with you. Each of those elements is feedback in the technical sense (the system is confirming you got it right), but the animation goes further than acknowledgment. It creates a small emotional reward. Psychologists call this positive reinforcement, and it’s exactly what Duolingo is engineering: a conditioned association between getting an answer right and feeling good about it. It’s a deliberate design decision that keeps people coming back.

Or take the iOS rubber-band scroll effect, the way the screen stretches and snaps back when you reach the edge of a page. It serves no functional purpose in the strict sense. Nothing new happens when you hit the bottom. But the elasticity communicates something important: you’ve reached a boundary, and the system is acknowledging it. Without that feedback, hitting the scroll limit feels abrupt and disorienting. With it, the interface feels physical, like something with real edges and weight.

What both examples share is that the animation isn’t just decorating a state change. It’s communicating a relationship between the user and the system. Duolingo’s animation says: We noticed what you did, and it mattered. The scroll effect says: you pushed against a limit, and the product pushed back. That kind of communication builds a subtle but cumulative sense of trust.

What This Means When You’re Designing

Every interaction in your product is a conversation between the user and the system. Microinteractions are the responses in that conversation, and how you respond shapes how the user feels about the relationship over time.

The baseline is always feedback: every action needs an acknowledgment. A button press without a state change is a question with no answer. A form submission without a confirmation is an email left on read. Once that baseline is solid, you have room to go further, to use motion not just as confirmation but as communication, as an emotional cue, as the thing that makes a product feel like it was designed by someone who cared about the experience of using it.

The best microinteractions are the ones users never consciously notice. They’re the reason an app feels smooth, but nobody can explain why. They’re the millisecond animations, the haptic nudges, the state changes that arrive before doubt has a chance to form. Get them right, and your product doesn’t just work. It feels like it’s on your side.

Why the Principle of Reciprocity Matters

In social psychology, reciprocity is the principle that when someone gives us something of value, we feel an innate obligation to return the favor. It’s why free samples work, it’s why lead magnets are effective, and why a small gesture of goodwill can spark long-term trust. In UX, reciprocity is just as important. Have you ever really considered when the best time to ask users to sign-up or subscribe is? Understanding this principle will help you design more effective user flows. In our example of free samples, there is a physical exchange, but in the digital world, it’s more often about providing something of value before expecting user commitment. When users feel like a product has already given them something of value, they’re much more willing to return the favor, whether that’s subscribing to your app, signing up for a newsletter, or providing you with contact information.

Reciprocity in UX Design

Reciprocity isn’t just an abstract theory; it is well-researched psychology, and some of the most successful products utilize it. Let’s look at two examples and see how powerful reciprocity can be.

Free Trials - Figma

If you’re reading this article, you’re no doubt familiar with Figma. Figma implements one of my favourite forms of free trials: there are no time limits, and you have full access to core features. You might be wondering how on earth this could be effective; they’re literally giving their product away for free! Let’s think about it from a product strategy perspective. By letting you access its full feature set, you get to experience the product exactly as you would if you were a paid user. This builds immediate trust with the user because they can see firsthand the value your product provides. As you use Figma, you’re investing your time, effort, and creativity on the platform, which makes them much more likely to stick around.

When does Figma actually gate you? When you require more than 1 project or more than 3 pages per project. By this time, you’ve already fully invested and you’re perfectly primed to reciprocate the value they’ve provided by purchasing a subscription.

There’s another benefit to this approach: it removes the barrier to entry. Whether you’re a student, a freelancer, or in the early stages of a startup, you can start without spending a dollar. Eventually, you grow, scale, and continue getting value from Figma; now paying for it feels like a cost and more like a natural next step to your growth.

All of this revolves around Figma providing you with value before it asks for your money.

Gradual Engagement - Duolingo

Duolingo also employs a similar strategy of reciprocity called gradual engagement. Like Figma, they let you jump into the core product experience right away, but they go a step further: you don’t even need to create an account to get started! Instead, they turn this barrier into a reciprocal action at the end of the user flow (we’ll get to that later).

Duolingo is a language learning app, which makes understanding the value of the app much more direct than something like Figma. Within seconds of launching the app for the first time, you’re already learning words from a different language. Baked into this process is gamification, visual indicators of progression, positive reinforcement, and many other strategies to make you feel like you’re getting value from the app. Similar to Figma, you get the full core experience with nothing gated behind a paywall. Unlike Figma, there’s no hard limitation for when you need to upgrade; instead, Duolingo uses different criteria to prompt you to subscribe. Some of those moments include: reaching certain milestones in your learning journey (which also plays into the peak-end rule), offering a premium free trial after multiple days of usage, when you frequently run out of hearts, etc.

By the time you get prompted to create an account to “save your progress”, it’s no longer a barrier; instead, it’s additional value! Now signing up is no longer about locking content; it’s framed in a way where they’re removing friction, preserving your progress, and it’s really an additional feature. You’re already invested, you’ve already learned a bunch of words and phrases, and you’re ready to sign up.

These strategies work because there is a psychological investment; people don’t like to abandon things they’ve started or invested time into.

How to use Reciprocity

How can you apply the Principle of Reciprocity in your designs? Just remember the core concept: provide the value first and ask for their loyalty later. This can take many forms depending on what your product or goals are. If you’re creating a lead magnet, provide a preview of your eBook or Whitepaper rather than completely gatekeeping it behind a form. If your product can provide a free model like Figma or Duolingo, that’s great! If you can’t, substitute it by providing smaller moments of delight, welcome discounts, or some other form of instant value that will trigger reciprocity. The timing of your ask is also important. It should always be after you provide users with value or removing a pain point. This heightens the sense that they’re making a fair exchange of value.

As you’re designing with reciprocity in mind, remember that you’re not trying to trick the user; always provide something of actual value, not the illusion of it. Because if it’s done in a way where it feels inauthentic, reciprocity can backfire, and you can lose the user’s trust.

Conclusion

Reciprocity is a fundamental social norm that shapes how people interact with your product. When you build trust, reduce friction, and create a fair exchange, you foster a user base that is loyal to your product. The most effective call to action starts with a gift of value.

You’ve just had a productive workday and you’re ready to go home. You get to the elevators and press the button, but nothing happens. No light, no ding, no change in the state of the world to signal that your request was received. Within seconds, uncertainty sets in. Did the button work? Is the elevator out of service? Do I need to press the button again? I’ve no doubt many of you have experienced something similar; it’s not just mild annoyance, it’s a deep-rooted human need being ignored. From birth, we’ve always relied on feedback to understand cause and effect: when a baby cries, the caregiver responds. One of the very first lessons we learn is the fundamental feedback loop of actions and reactions. In psychology, this is called Operant Learning (aka Operant Conditioning). Without proper feedback, behavior becomes random, user interfaces and buttons lose meaning, and the user gets frustrated.

Why Feedback Matters in UX/UI

When it comes to UX/UI in the digital world, feedback is the proof that the system is listening and responding. Feedback in design isn’t just about aesthetics or usability, it’s about addressing one of the most deeply embedded psychological needs humans have: certainty.

Our brains are prediction machines. We constantly scan our environment to understand and analyze cause and effect. Uncertainty and the fear of the unknown is a primal fear, and it is likely why when we break the feedback loop, we feel a sense of unease. It is in some ways a micro version of our primal survival mechanisms, and it triggers a “something is wrong” signal. Your brain tells you: If you can’t predict what’s next, you need to stay alert.

In the modern digital world, the dangers we used to face for survival are gone, but the primal circuitry is still there. When you press the “Pay Now” button and nothing happens, the same “something is wrong” signal still triggers. Neuroscience research indicates that uncertainty activates the amygdala, the brain’s threat detection center, while predictable actions activate the prefrontal cortex. This suggests that when things are clear and consistent, our brain feels more secure, safe, and in control. The solution to this problem is simple: Provide users with appropriate feedback and complete the loop. Every action in the UI should have an accompanying reaction. This could be a small animation, a loading bar, a spinner, a message, anything that acknowledges the user’s actions.

Types of Feedback

There are many ways you can implement feedback in UX/UI, but they all share the same purpose: to let the user know their action as been acknowledged, understood, and is leading somewhere. The most effective designs often layer multiple feedback types together, creating a richer experience.

Visual Feedback

Visual feedback is the most common form in digital design as it appeals to our dominant sense and is often the fastest to register.

Examples:

• Button states: Hover highlights, click depressions, or color changes.

• Progress indicators: Spinners, bars, percentage counters.

• Success/confirmation states: Checkmarks, “Message Sent” banners, order confirmation pages.

• Psychological impact: Visual changes give immediate proof that the system “heard” you, calming the brain’s uncertainty alarm. A well-designed animation can also add a touch of delight, rewarding the action and encouraging continued engagement.

Auditory Feedback

Sound can convey meaning instantly and efficiently. This can also be an excellent choice when considering accessibility and multitasking applications. It can also be used to train users to associate sound with positive outcomes, like Duolingo’s famous correct answer sound.

Examples:

• Success chimes: Sending an email in Outlook or iOS “sent” whoosh or Duolingo’s da-ding effect when you answer correctly.

• Error alerts: Low-pitched beeps, buzzer sounds.

• Status cues: Notification pings, incoming message sounds.

• Psychological impact: Humans are wired to respond emotionally to sound: high tones can feel positive and confirming, while harsh tones trigger alertness. This taps into primal survival wiring: certain sounds historically signaled safety or danger.

Haptic Feedback

Tactile cues provide a physical sensation that reinforces interaction, particularly in mobile and wearable devices. It provides an excellent way of feedback without the user needing to actively pay attention.

Examples:

• Micro-vibrations: A gentle “tick” when a virtual dial turns.

• Press feedback: Subtle vibration when tapping a touchscreen button.

• Gesture confirmation: A pulse when you swipe to complete an action.

• Psychological impact: Haptics close the sensory loop in a uniquely visceral way; they make the digital feel tangible. This can increase confidence, especially when visual confirmation is delayed.

Contextual/Informational Feedback

This type focuses on guiding the next step, not just acknowledging the current one. It provides additional information to the user, helping them understand the current situation. You may have noticed these with the rise of the AI bots like ChatGPT. These LLMs take time to analyze and generate responses and many of them have started displaying notices like “Thinking…” or even displaying each step of the process its going through to assure users that they’re working in the background. Sometimes, UX designers will add progress bars or contextual information even if nothing is really happening in the background; this is known as the Labor Illusion, but we’ll talk about that in another post.

Examples:

• “Your file is uploading… 30% complete.”

• “Password is strong.” or “This username is already taken.”

• Smart tooltips that appear after an error.

• Psychological impact: Reduces cognitive load by telling the user exactly what’s happening and what to do next. It keeps the mind from wandering into uncertainty and prevents repeated trial-and-error.

Multi-Sensory Feedback

The most powerful feedback loops often combine visual, auditory, and haptic cues for redundancy and reinforcement.
Example: Sending a message in a mobile app might:

• Show a “sent” checkmark (visual)

• Play a soft chime (auditory)

• Give a micro-vibration (haptic)

• Psychological impact: Redundancy ensures that even if one sense misses the cue, another catches it. This mirrors how, in nature, important survival cues often come through multiple channels like seeing, hearing, and feeling a storm approaching.

Conclusion

Every digital interaction is a conversation between a human and a system. In human conversation, silence after a question feels awkward, even unsettling. In UX, silence after an action feels the same, and it leaves us with a primal discomfort rooted in the fear of the unknown. Feedback is essential to every part of the UX/UI, and it builds trust between the user and the system. The next time you’re designing something, remember to always close the feedback loop and ensure the user feels confident and acknowledged.

At its core, Gestalt psychology explains how our brains instinctively group elements, recognize patterns, and simplify complexity. Once you understand these principles, you'll start noticing them everywhere: in branding, user interfaces, and even everyday objects. I think it’s important for every designer, whether you’re a graphics designer or a UI designer to understand the very basics of Gestalt principles.

While there are more Gestalt principles, there are typically six core principles that are particularly essential to design, so let’s go through them!

Proximity

The principle of proximity states that objects placed close together are perceived as part of the same group; creating a visual relationship between them. This principle is widely used in UI design to enhance usability and reduce cognitive load. For example, navigation menus group related links together, and connectivity controls like WiFi and Bluetooth settings are always clustered together for quick access. By leveraging proximity, designers guide users intuitively, making interfaces more efficient and easy to navigate. Design decisions like these are often taken for granted because they’ve become standard practice. A good designer can follow conventions and make things look right, but a great designer understands why those conventions exist, and knows when breaking them can lead to something even better.

Similarity

The principle of similarity states that visually similar elements are perceived to be part of the same group. This is a fundamental concept in UI design, helping users quickly recognize related functions. Lets use the navigation menu as an example again, they often share the same typography, color, and spacing to show they are a part of the same system. Similarities in typography and iconography create a feeling of cohesion and unity. This also extends into functionality where a designer will use a certain color for primary actions, and often red for anything destructive. By using similarity properly, designers can create more intuitive and cohesive experiences, and even convey meaning like functionality.

Closure

The principle of closure states that our brains will fill in gaps to perceive incomplete shapes. This is one of the most commonly used principles in logo design. Many recognizable logos use closure to create visually interesting designs such as the famous WWF panda or IBM’s logo. Beyond branding, closure is also used in icon design. By strategically omitting parts of a shape, you can create minimalistic icons that improve clarity while reducing clutter. Closure invites users to interpret designs in interesting ways.

Continuity

The principle of continuity states that our eyes tend to follow lines or curves. This is to say that we have a perception bias to perceive things as connected forms rather than disconnected. This is used in fundamental UI design almost everywhere you look, from the way navigation menus are laid out to progress bars and onboarding sequences, continuity helps subtly guide the user’s eye.

Figure-Ground

The Figure-Ground principle describes how our brains instinctively separate objects (figure) from their backgrounds (ground). This technique is used extensively in visual design to create emphasis and direct attention. In marketing, for example, product images and logos are often placed against high-contrast backgrounds to make them stand out. In UX/UI design techniques like lightboxes, where the background is dimmed while the main content remains bright and sharp, ensure users stay focused on key content. Drop shadows, overlays, and layered UI elements also help establish hierarchy, making important components more visually dominant.

Pragnanz

The principle of Pragnanz states that our brains prefer to interpret complex visuals in the simplest, most organized way possible. The most common example of Pragnanz is The Olympic Rings. The Olympic Rings is actually an intricate design made up of overlapping and intersecting lines, however, our brain naturally perceives the symbol as 5 distinct rings that create a sense of order and symmetry. Pragnanz is often used in logo and icon design to reduce visual complexity and improve clarity.

Conclusion

Gestalt principles play a fundamental role in how we perceive and interact with design. By understanding how our brains naturally group, organize, and simplify visual information, designers can create experiences that feel seamless, intuitive, and visually harmonious. Once you start recognizing them, you'll see Gestalt psychology at work everywhere, from the apps you use daily to the logos of your favorite brands. When we design with these principles in mind, we can make our experiences feel natural and intuitive and sometimes, even a little magical.

Before we dive in, think about a product you’ve recently used. Hold onto that memory, we’ll come back to it soon.

When users interact with a product, they aren't processing every moment equally. Certain experiences like the emotional highs, the standout moments, and the way things end, matter far more than others. If you can design around these moments, you can leave a lasting, positive impression.

Today, we’ll focus on four key cognitive biases that influence how users remember an experience:

• The Peak-End Rule

• The Serial Position Effect

• The Zeigarnik Effect

• The Von Restorff Effect (Isolation Effect)

Understanding these biases will help you craft experiences that don’t just work, they stick. Let’s dive in!

The Peak-End Rule

Not every moment is equal, and according to the Peak-End Rule, users recall the emotional peak and the end of an experience the most. This principle, coined by psychologist Daniel Kahneman, has been supported by studies across healthcare, entertainment, and everyday experiences. For example, patients undergoing medical procedures rated their experiences based not on the total pain endured, but on the most painful moment and the ending. The same logic applies in UX.

As designers, these are the perfect moments to create emotional impacts and make the experience more memorable for users. For example, when a user accomplishes a task/goal in your app you can play a celebratory animation, play a positive sound cue, provide visual affirmations, or provide social reinforcement. Think of apps like Duolingo or Finch, which uses many of these tactics to give users a sense of delight when they answer a question correctly or complete a daily goal. These moments are also great for softening the effect of negative experiences, like providing a quick path to resolving an error.

The Serial Position Effect

Similar to the peak-end rule, the serial position effect has to do with moments. The serial position effect states that people tend to remember the first and last items in a sequence the most. Have you ever noticed CTAs are almost always placed at the top or bottom of a navigation menu? That’s on purpose. You might be wondering why this is the case, and the answer lies in the two “sub effects”:

• The Primacy Effect: People tend to remember the first items the best because they’re processed first and given more cognitive attention.

• The Recency Effect: People recall the last items better because they’re still held in their short-term memory, making them easily accessible.

Make sure to always put the most important elements, like CTAs, at the beginning or the end of sequences. There are many ways you can take advantage of this and here are a few examples:

• For onboarding processes, make the first screen simple, motivating, and aligned to the user’s goals, and make the final screen a success state that creates delight.

• When listing features for your product, place the highest-value proposition features at the beginning and end of the list.

• If you have an e-commerce website, research shows that the first and last product in a carousel or search result list gets the most clicks.

Zeigarnik Effect

The Zeigarnik Effect states that incomplete or interrupted tasks are remembered more often than completed tasks. Have you ever had moments where you were writing an email, got interrupted, and kept thinking about getting back to that email? That’s the Zeigarnik Effect. Our brains crave closure, and unfinished tasks create a mental tension that keeps tasks at the top of our mind.

If we want to help users complete their task there are a few effective methods. Visual cues like progress bars and checklists trigger a psychological urge to complete a task. This is commonly done with onboarding checklists in SaaS dashboards. E-commerce websites often provide consistent reminders about having items left in your cart, prompting you to checkout. You can also implement gamification and habit-forming strategies to create engagement loops; this is often found in educational apps like Duolingo.

While the other biases we’ve talked about so far clearly revolves around timing, the Von Restorff Effect is more about contrast. The Von Restorff effect is a cognitive bias where people are more likely to remember an item that stands out from its surroundings. Imagine a table with a line of oranges, but there’s an apple right in the middle. That apple is going to stick out like a sore thumb.

There are many ways we can take advantage of this in terms of contrast, but I will save those ideas for a post dedicated to contrast. Let’s think about how this bias works in conjunction with timing. The reason why providing moments of delight during peak-end events works, is because of the Von Restorff Effect. If we always provided fun animations and positive reinforcement for everything a user does, it would dull its effect. These moments of delight are effective because of temporal contrast; if everything delights all the time, nothing stands out. With proper timing, emotions significantly boost memory. It is essential that these moments should occur immediately after emotional effort or anticipation, not before.

Conclusion

What these 4 cognitive biases teach us is this: not every moment is experienced equally, users remember specific moments. The highs, the endings, the unfinished tasks, and the standout elements all shape how people feel, recall, and ultimately return.

As designers, understanding these crucial moments helps us create experiences that delight and impact the users. And it is often these key moments that convince a user whether they’ll continue using your product or not.

In the world of user experience, the term “heuristics” likely brings to mind Jakob Nielsen’s 10 Usability Heuristics. These are foundational guidelines for designing user-friendly interfaces (and if you haven’t read them, I recommend starting here).

But today, I want to go one level deeper and look at heuristics through a psychological perspective, and explore why they’re so effective when applied to UX and UI design.

First, a Quick Detour: Cognitive Load

Before diving into heuristics, we need to understand the concept of cognitive load. Cognitive load refers to the mental effort required to process information in your working memory. When cognitive load is high, users have to work harder to complete tasks. When it’s low, things feel intuitive and seamless. As UX designers, our mission is to reduce unnecessary mental effort, which is where heuristics shine. To learn more about Cognitive Load, check out my previous blog here.

Heuristics reduce cognitive load because they operate at a subconscious level. They draw from our mental models and lived experiences, allowing us to make rapid judgments with minimal thinking.

The Psychology Behind Heuristics

Our brains naturally categorize the world around us. This is how we manage the overwhelming amount of information we take in every second. By sorting things into these mental buckets, we can quickly make assumptions, assign meaning, and most importantly, respond to these stimuli. This is the foundation of heuristics, the mental shortcuts we rely on every day to guide our decisions and behaviors.

Take a dog, for example. You don’t need to see every breed to know what a dog is. You’ve built a mental model: four legs, barks, and wags its tail. So when you see a new breed of dog, you recognize it instantly and act accordingly; maybe give it a pet, or if you’ve previously had a bad experience, avoid approaching it altogether.

The same principle applies to digital interfaces.

When users encounter a familiar element like a magnifying glass icon, they immediately associate it with “search,” even if they’ve never used the product before. This is an example of Jakob’s Law: Users spend most of the time on other websites and applications, and expect yours to function the same as well. The magnifying glass is a perfect example of a Universal Icon, which means the vast majority of users will immediately recognize it and associate it with a particular function. You should almost never replace these icons as swapping these out for something creative or novel will go against a user’s expectations.

By properly leveraging heuristics, you ultimately accomplish one thing: lower cognitive load. It saves your users’ time, mental effort, and makes your UI feel intuitive.

The Flip Side: Cognitive Biases

Here’s where it gets tricky. Heuristics deal with categorizations based on assumptions and learned behaviors; however, assumptions are not always correct. Because of this natural vulnerability to false assumptions, heuristics are responsible for many of our cognitive biases, which are systematic errors in reasoning..

Let’s consider a simple example of confirmation bias. If a user signs up for a task management app, they’ve likely used something similar like Trello or Asana, for example. The user may have expectations for how it is organized: maybe they expect to see drag-and-drop boards, color-coded labels, or a “+” button to add a task. If your app organizes tasks in an entirely different way, like a calendar-first layout or timeline view, then they may have automatically assumed this is not as "good," even if your solution is actually better. Why? Because the new solution does not confirm what they believe "good" task management should look like.

This is the essence of confirmation bias: people look to find information that matches what they already have a mental model for, and ignore/dismiss what doesn't. If users never had any previous assumptions about the product, there wouldn’t be a mismatch in expectations. It all stems from heuristics, and most of the time, the user isn’t even aware that any of this is happening at all.

But this isn’t necessarily bad news. If we understand how biases work, we can design with them in mind. As designers, it’s our job to understand and leverage these heuristics and biases to create experiences that truly resonate with the users.

Conclusion

Heuristics are one of the most primal and powerful tools of our minds. They help us quickly understand our environments, make snap judgments, and come to decisions, all without us thinking about it at all. By understanding and intentionally using heuristics, we can create more than just usable interfaces; we can create intuitive experiences. As designers, we can guide users with familiar cues, put things where they are expected, and make their journey as effortless as possible.

You can think of cognitive load as your brain’s “bandwidth.” If you’re trying to process too many things at once mistakes happen and frustration creeps in. But when cognitive load is well-managed, even complex tasks feel doable, sometimes even enjoyable, and dare I say, even magical.So, what exactly are the types of cognitive load we deal with in UX? Let’s break it down using a real-world example: tax software. (Stick with me and I promise it won’t be painful).

1. Intrinsic Cognitive Load (ICL)

This is the inherent difficulty of a task. Some things are just naturally complex (Tesler’s Law). Filing your taxes, for example, involves legal forms, personal details, income breakdowns, and more. You can’t really eliminate that complexity, but there are ways we can design around the complexity to make it feel a lot more manageable.

One of the best strategies? Progressive Disclosure: showing only what’s needed, when it’s needed.

If the app throws every question and input at the user all on one page, it’s overwhelming, and users will make more mistakes. But if that same information is separated into logical groupings like Personal Info, Income, Deductions, and Tax Credits, and revealed one chunk at a time, it’s suddenly much more manageable. Now the user can focus on just one category at a time with a more manageable amount of information.

This not only helps reduce cognitive overload, but it also gives users a clear sense of progress and direction. When the design mirrors the way people naturally think and process information, it makes even complicated tasks feel more intuitive and a lot less stressful.

2. Extrinsic Cognitive Load (ECL)

This is the mental junk we don’t need, the stuff that distracts or confuses. Think cluttered layouts, inconsistent navigation, or long-winded copy. All of it pulls users away from their actual goal. And as designers, this is the kind of load we can and should eliminate.

Let’s go back to our tax software. You already have a mentally heavy task — so now imagine that on top of that, the interface is a chaotic mess:

• Five different fonts

• Buttons in random places

• Menus that make no sense

• Inconsistent color usage

How do we prevent this? Ensure we have clean layouts, clear visual hierarchy, and well-structured information architecture. When the task itself is already mentally demanding, we must do everything we can to ensure the design does not add to it:

• Use white space to reduce visual noise

• Group related items together

• Label elements and use simple language

• Maintain consistency in your design

• Use common patterns and universal icons

The goal is to ensure the user feels in control and focused on the task at hand.

3. Germane Cognitive Load (GCL)

Finally, there is the Germane Cognitive Load, this is the good kind of mental effort. It’s what helps users understand, learn, and feel empowered by what they’re doing. You actually want to encourage this, as long as it’s well-balanced. Continuing with our tax example, here are some examples:

• A short animated explainer showing how tax brackets work

• Visual charts breaking down your income sources

• Quick notes explaining why a certain deduction matters

• Tooltips explaining line items in plain language

Germane cognitive load is all about helping users learn and integrate information. These design elements don’t just help users complete a task; they help them get better at it. And when you apply germane cognitive load properly, you can make users feel a real sense of confidence and satisfaction.

Conclusion

At the end of the day, UX isn’t just about slick interfaces and pretty buttons. It’s about managing how people feel as they interact with your product. To recap, the three types of cognitive loads:

• Organize intrinsic cognitive load, so even hard tasks feel manageable.

• Reduce extraneous cognitive load, so your users don’t get frustrated.

• Encourage germane cognitive load, so people walk away feeling smarter.

If you get these right, even filing taxes can feel smooth, simple, and maybe even a little bit magical.

A brief introduction about myself — my journey into UX design was a little unorthodox. I originally studied Microbiology and Immunology, with a minor in psychology. Even though psychology was a small part of my classes, they were always my favourite. I loved understanding how we think, make decisions, and form behaviours. After graduating and working in medical labs for a few years, I decided to go back to school for software development. While I was there, I realized something: UX and product design is essentially a combination of psychology and technology, two things I truly loved. Fast forward to today, I am creating a space to share my passion for thoughtful and human-centered design. The UX Forge is a space where we break down the why behind good design by exploring the cognitive science of user behavior and psychology.

Here’s what you can expect from this newsletter:

Every 2 weeks or so, you’ll get thoughtful posts across a few key categories:

• UX Laws – Psychological principles and heuristics (like Hick’s Law or the Von Restorff Effect) that help explain how users think and interact.

• UX Principles – Best practices and design ideas like feedback, clarity, empathy, and reciprocity — the softer, essential side of human-centered design.

• General Posts – Editorials, industry observations, and other news.

• Product & Tool Spotlights – A look at the software, tools, or even physical things that help me design and work more effectively.

• Case Studies – We’ll take a look at real products and dissect aspects of it from a psychological and UX perspective.

Whether you're a designer, product manager, or curious designer, I hope this newsletter gives you fresh insight into how people think — and how we can design better because of it.

Let’s forge better experiences together :)

Share TheUXForge