Experience Assessment
PropertyIntel is built on an unusually strong foundation — the opportunity now is to make the experience as inclusive and reassuring as the product is well-engineered.

The application already does the hard things well: a consistent design system runs through nearly every screen, modals and forms are built on shared, accessible components, and the interface communicates clearly in the moment with live toasts and staged progress. This is deliberate, mature work.

What's evolved is the surrounding experience. Expectations around accessibility, guided multi-step flows, and reassuring feedback have moved forward. The opportunities in this assessment map directly to those — most are small, high-leverage changes, and several can be made in one place and felt across the whole product.

Opportunities identified, each with a clear, practical path forward. Of the interface is already built on the Optics design system — a rare, strong starting point. Accessibility improvements that would bring the app toward full WCAG AA compliance.

Then.

Color, spacing, type, and shape run through a single set of Optics design tokens across nearly every screen — so the product reads as one consistent, intentional whole. Modals and drawers are built on one shared component, so focus trapping and screen-reader semantics come for free everywhere a dialog appears. Tool actions confirm themselves with a tidy three-state toast — searching, success, or error — that clears itself, so users always know what just happened. The bulk-order flow shows staged progress with per-step status and surfaces unverified addresses up front, so a large, slow task never feels like a black box. A clear sidebar with a visible "you are here" highlight on desktop becomes a thumb-friendly bottom rail on smaller screens, with breadcrumbs that gracefully collapse. Interactive controls are sized generously on mobile (48–56px), comfortably above the accessible minimum — the product was built with touch in mind.

Now.

Thirteen opportunities, grouped into three themes. Each is grounded in the actual application — and each builds on the strong foundation that's already there.

Experience.

Moments where a small amount of extra reassurance or guidance would let users move with more confidence.
When an action can't be completed, point the way forward. What we observed: when an action can't be completed, the app currently shows a single general message — "Something went wrong!" — without the specific reason or a next step.

What this means for users: people can't tell what happened or whether it's safe to try again, which turns a small hiccup into a dead end. Because this message comes from one shared button, giving it a clear reason and a recovery action improves every flow at once.
Show people how far they've come. What we observed: creating an estimate request moves through four steps — boundaries, features, details, review — but there's no visible "Step 2 of 4" indicator along the way.

What this means for users: without a sense of progress, a multi-step task feels longer and less certain. A simple step indicator sets expectations and pulls people toward the finish line.
Make required fields obvious up front. What we observed: required fields are signalled by their wording alone, with no asterisk or visual marker, and the form blocks you from continuing until they're complete.

What this means for users: people discover what's required by being stopped, rather than by scanning the form first. A consistent required-field marker lets them plan their input and avoid the friction of being turned back.
Guard against losing work in progress. What we observed: one key flow warns before discarding unsaved work, but most forms and dialogs close silently if dismissed mid-edit.

What this means for users: a stray click can quietly erase what someone was in the middle of. Extending the existing "unsaved changes" prompt to the other forms makes the product feel dependable everywhere.

A few smaller refinements.

Give system messages a clearer voice.
Notices currently share one neutral style. Distinguishing success, warning, and error at a glance — using the alert colors already in the design system — helps people read the message's importance instantly.
Preview what's behind "More".
On mobile, the navigation "More" control hides extra destinations with no hint of what or how many. A small count or labels would make it more discoverable.
Group the longer forms.
The material editor presents many fields as one long list. Grouping related fields into labelled sections would make it faster to scan and complete.

A system already working for you.

Nearly the entire interface is already drawn from the Optics design system, anchored to PropertyIntel's brand hue — so colour, type and shape stay consistent by construction. The opportunity here is mostly upkeep: folding the last few one-off colours into the same token scales so the product keeps feeling like one unified whole as it grows.

Modernization.

Accessibility expectations have become table stakes since the original build. These are concrete, well-understood improvements — and several are single changes that ripple across the whole product.
Let people zoom in. What we observed: the application currently prevents pinch-to-zoom on mobile devices.

What this means for users: anyone who relies on zooming to read comfortably can't, which is the one finding that rises to a true accessibility blocker. Re-enabling zoom is a single-line change — and one screen in the app already does it correctly, so it's simply a matter of bringing the rest in line.
Give every icon button a spoken name. What we observed: icon-only buttons currently lean on a hover-tooltip mechanism for their name, which can announce an internal label to a screen reader — and in a few places no name at all.

What this means for users: people using assistive technology may hear an unclear or missing label for common actions. Wiring a proper accessible name into the shared icon and button components fixes this everywhere from one place.

Make the keyboard a first-class citizen.

Keep the focus indicator visible.
Most of the app already shows a clear focus ring; two areas remove it without a replacement, so keyboard users lose their place there. Restoring the ring brings them in line with the rest.
Make every control keyboard-reachable.
A couple of clickable elements respond only to the mouse. Making them operable by keyboard ensures no one is locked out of an action.
Add a "skip to content" link.
A small, hidden-until-focused link would let keyboard and screen-reader users jump past the navigation straight to the page's main content.
Announce the current page.
The active navigation item is clear visually; adding the matching screen-reader signal — set once in the shared link helper — tells assistive tech which page you're on.

Small changes, system-wide reach.

The most exciting thing about this codebase is its leverage. Because so much already flows through shared components and a single design system, several of these improvements aren't screen-by-screen work — they're made once and felt everywhere: accessible names in the icon component, the "current page" signal in the link helper, clearer recovery in the shared action button.

That makes a focused accessibility pass — a sprint, not a redesign — the highest-value next move. It would take PropertyIntel from "visually polished" to "polished and genuinely inclusive," building naturally on a foundation that's already doing the hard part.

We evaluated PropertyIntel against core UX principles, based on what we observed in the actual application.

Standard, recognizable layout throughout — sidebar, breadcrumbs, and form inputs from a consistent component set. Users feel at home immediately. A single design system runs across nearly every screen, so the product looks intentional and cohesive. Actions confirm themselves promptly with clear toasts, and long tasks show staged progress. Dialogs and panels are well-structured and built on shared, predictable components. The multi-step request flow doesn't yet show how many steps remain. A failed action shows a general message, and most forms don't warn before discarding unsaved work. A small number of one-off colours sit outside the token system and could be unified. Pinch-to-zoom is currently disabled on mobile. A few controls are mouse-only, the focus indicator is removed in two spots, and there's no skip-to-content link. Icon-only buttons rely on a tooltip for their name rather than a proper accessible label.
PropertyIntel has the rare thing: a genuinely strong foundation. A focused, well-scoped effort brings the experience up to the same level — and we'd love to explore it with you.
A short accessibility pass — re-enable zoom, give icon buttons accessible names, restore focus indicators, add a skip link and the "current page" signal. Mostly single-point fixes with outsized reach. Strengthen feedback — a clearer recovery path on the shared action button and distinct styling for success/warning/error messages. Guide the key flows — a step indicator on the request wizard, obvious required-field markers, and unsaved-changes protection across forms. RoleModel Software · rolemodelsoftware.com · June 2026