Designing Interfaces
Transcript
Made automatically from the video's sound, so it may contain mistakes. Click a line to play from there.
Welcome back. Let's jump straight into this explainer. Following on from our last session, where we anchored our demo site, Canopy, with authentic focal cropped photography, consistent aspect ratios, and unified 24 pixel icons, we're now focusing purely on the interactive controls.
I'm talking about buttons, links, cards, navigation, and forms. Users need to instantly perceive what's clickable, what's editable, and what's primary. So we're skipping the foundational typography and grid setups today and diving headfirst into the deep end of interactive signifiers.
Now, if you take nothing else away today, remember this stat. Stripping those interactive signifiers from a flat UI template results in a massive usability drop. We're talking a greater than 20% task time penalty for users.
That's a huge speed bump. According to eye tracking studies by the Nielsen Norman Group, when we remove traditional clickability clues, like a simple underline on a text link or a shadow on a button, users actually have to stop and spend significantly more time just figuring out where on earth they're allowed to click.
And this brilliantly illustrates our core philosophy for today. Modern interfaces are basically ripped out all those perceptible cues users rely on, which skyrockets the interaction cost and creates a ton of uncertainty. Don Norman famously summed this up perfectly.
Signifiers, not affordances. He pointed out that whilst affordances define what actions are actually possible, it's the signifiers that communicate where that action should take place. Without clear signifiers, your users are essentially navigating in the dark.
So let's look at section one, the cost of ambiguity, and exactly why flat UI templates fail users. Let's break down a typical AI generated SAS page because it perfectly captures why these generic templates are just so frustrating.
We've all suffer through these, right? You get three competing solid fill primary buttons basically screaming for attention side by side. You've got borderless ghost pills that are completely indistinguishable from static category tags.
Body links, stripped of their underlines, vanishing entirely into a sea of prose. Oh, and my personal favourite, the Matrioshka doll nested cards. You know, an out of order wrapping a bordered badge, wrapping an inner box, wrapping a bordered button.
Combine all that with two column forms featuring invisible one-pixel inputs and disappearing placeholder only labels, and you've engineered an absolute masterclass in cognitive friction. But it actually gets worse.
Weak signifiers force people to constantly scan the page because they just lack the confidence that the element they're looking at is what they actually want. And when error messages inevitably occur, or visual focus is lost, users navigating with keyboards or screen readers are left completely stranded.
Just imagine using the tab key to move quickly through a form. If a placeholder vanishes and there's no persistent label, you have absolutely no idea what field you're currently on. These weak cues don't just slow down the average user, they actively block people who rely on assistive technologies.
Fixing the interactive hierarchy. Let's build a structured system for clear actions. Alright, let's fix this live by transforming the canopy demo site. First, we'll streamline the navbar, stripping it back to four scannable links, one quiet login link, and exactly one primary call to action, drawing inspiration from the US web design system.
We're going to demote all those competing solid hero and pricing buttons into a strict four tier visual hierarchy. Tier 1 is primary, a solid filled background limited to one per view, with verb object copy, like star 14-day trial.
Tier 2 is secondary, giving medium visual emphasis using outline or tonal styles for things like a cancel action. Tier 3 is tertiary, text-only buttons for optional actions like few more. And tier 4 is destructive, using high contrast warning styles, usually red, so users know a destructive action is completely irreversible.
That's right out of playbooks like Material 3, Gov.uk, and Shopify Polaris. Now, a button's style tells you its importance in the hierarchy, but you also absolutely need these five essential button states to communicate interaction status in real time.
Enabled is your default high contrast ready-to-go state. Disabled is visually muted, letting users know they can't click it yet. Hover offers a subtle visual shift when the pointer rests on it. Focus is critical.
That highly visible focus ring is literally the only way keyboard users can navigate your site. And finally, pressed. If you don't design a pressed state to give visual feedback, the button feels completely unresponsive, and users are gonna wonder if the system even registered their click.
Perfecting cards and forms. This is where semantic HTML meets accessible design. Okay, let's sort out those awful, overly bordered Matroischkadol nested cards on Canopy's feature section.
We need to stretch the click target over the entire card. But wrapping a whole block element in a single anchor tag creates incredibly messy output for screen readers, which will just read out every single piece of text as one massive link label.
Instead, we use Hayden Pickering's absolute genius CSS pseudo-element trick. First, you give the main card container element position relative. Then you target the primary link inside the card using the after sudo element. You give that pseudo-element position absolute with inset values of zero.
By doing this, we preserve clean document structure and text selection, whilst providing that massive ergonomic click target users expect. And just ensure your interactive Hoffer States updates the entire card visually. Next up, let's rebuild Canopy's sign-up block.
We're binning the broken template with its confusing two-column layout, one pixel low contrast invisible borders, and disappearing placeholder text. Instead, we're adopting the W3C and Beimard Institute standard, a single column flow.
This prevents users' eyes from having to zigzag all over the place. We'll use persistent, top-aligned labels with a tight 4-6 pixel gap above the input. We're setting the input heights to a generous 44 pixels, width matching the fields, and ensuring a minimum 3 to 1 border contrast against the background to meet WCAG standards.
Top it off with calm, on-blur inline validation, and you've got a brilliantly accessible form. Sizing your interactive elements, like those inputs and buttons, to a minimum touch target area of 44 by 44 CSS pixels is not just some arbitrary aesthetic choice.
It is a fundamental accessibility standard. Whether a user has motor control issues, hand tremors, or is simply tapping with a fat finger on a mobile screen, giving them a generous target area prevents accidental clicks and endless, completely preventable frustration.
Three distinct visual archetypes, showing how you can be polished without looking generic. You might be thinking that following all these strict usability best practices is going to make your interface look incredibly boring, but that's simply not true.
You can follow Jacob's law, creating interfaces that behave in wonderfully familiar ways while still injecting an unmistakable brand personality. Let's look at three examples. For a warm editorial vibe, we can use high-contrast editorial CTAs, restore crisp offset underlines on prose links, say a 0.2EM offset with a one-pixel thickness, and use white space to group borderless article cards.
For a crisp Swiss or technical utility, we might use sharp tabular pricing toggles, unmistakably high contrast states, and precision 40 to 44 pixel inputs with sharp keyline borders. And for a tactile craft brand, we can use beautiful warm tonal surface cards, leveraging that pseudo-element click trick we just learned, paired with calm single column booking forms.
The underlying accessible structure remains identical across all three. It's just the visual skin that changes. Clarity never requires sterile sameness. Finally, section 5, the interactive spec sheet.
Let's quantify these component workflows. To wrap up this explainer, we've anchored all these exact component specifications into a reference sheet. Feel free to screenshot this, because you can take these specs directly into your Figma component variants, Webflow, or your code editor.
Notice the semantic distinction. Always use a button tag for on-page actions and an A anchor tag for URLs. Look at the sizing specs. Aim for a 44 by 44 pixel target based on the Apple Human Interface Guidelines, rather than just scraping by on the 24x24 pixel WCAG 2.2AA floor.
Observe the spacing rules, a tight 4-6 pixel gap binding the label to its input, contrasted with a much larger 16-24 pixel field to field gap. And of course, ensure you've explicitly designed all six interactive states default, hover, focus visible, active, loading, and error.
This is your definitive blueprint for clarity. And that sets us up perfectly for our upcoming 15-minute sign-up card cleanup exercise, as well as our next explainer, where we'll be tackling depth, surfaces, and motion.
So, leaving you with a final thought: with your Interfunx elements now clearly signifying exactly how they function, how will you apply these tools to design surface depth and motion in your next layout?