When UX design is actually worth it
UX research and prototypes aren't always necessary. For a small website with a few pages, the logic is usually obvious, and research would cost more than it returns.
UX design pays off when:
- people use the product every day, so every friction point is multiplied by hundreds of repetitions;
- the system has several roles with different tasks: customers, managers, administrators;
- users have to complete a multi-step process: sign-up, submitting a request, configuring settings;
- the product is already live, but analytics show that large numbers of people drop off at a specific step;
- user errors are costly: wrong orders, lost data, support tickets.
Research: what we find out
The goal of research is to replace guesswork with facts. Depending on the product, we:
- talk to a few real users about their tasks and frustrations;
- interview support and sales staff, who know what customers complain about most;
- review analytics data and session recordings in Hotjar or Clarity to see where people get stuck;
- look at how other products your users already know handle similar tasks.
If you don't have behavioral data yet, set up analytics first. To learn how to read that data, see our web analytics and audit page.
User flows, structure, and wireframes
After research, we map out the key user flows: what a person wants to do, where they start, which steps they take, and where they end up. For each flow, we count the steps and ask whether every one of them is really needed.
Next, we build screen wireframes: simple black-and-white layouts with no colors or photos. At this stage, changes are cheap: moving a block or removing a step takes minutes, not a rework after hours spent on detailed design.
Prototyping and user testing
A clickable prototype in Figma looks almost like the real product, but there's no code behind it. We give it to a few people from the target audience and ask them to complete specific tasks: find a product, submit a request, change a setting.
Even a handful of these sessions uncovers most of the major problems: unclear button labels, hidden features, fields people don't know how to fill in. Fixing them in a prototype is far cheaper than after development.
Importantly, we don't ask testers "Do you like it?" We watch what people do, where they hesitate, and what they say out loud. Behavior tells you more about usability than any rating.
Engagement options: pros and cons
Full cycle for a new product. From research to final screens and a design system. Pro: the product is built around real user tasks from day one. Con: the largest amount of work before development begins.
UX audit of an existing product. We analyze where users get lost and deliver a prioritized list of issues with recommended fixes. Pro: fast results and a clear improvement plan. Con: without follow-up design work, your team will have to implement some of the fixes on its own. And if there are so many changes that it's easier to start over, you're looking at a redesign.
Designing a single flow. For example, just the checkout or a new user's first steps in the product. Pro: focus on the biggest pain point. Con: you need to make sure the new solution stays consistent with the rest of the product.
Design systems for large products
When a product has dozens of screens and several designers or developers, the interface quickly drifts without shared rules: the same actions look different, and five shades of the same button appear.
A design system is a component library in Figma plus rules for when and how to use each component. It speeds up the release of new features and makes the product predictable for users. A small website only needs a simple UI kit; a full design system pays off only on large products.
Website, portal, or mobile app
The design approach is the same, but the details differ. For a website, navigation and search matter most; for a customer portal, it's tables, filters, and handling large volumes of data; for a mobile app, it's gestures, iOS and Android platform guidelines, and one-handed use. We cover app design separately on our mobile app design page.
What we'll need from you
- Access to analytics and, if available, support tickets.
- Contacts for a few users willing to talk or take part in a test.
- Your business goals: what should improve after the interface redesign.
- Any technical constraints we should know about: what can't be changed in the existing system.
- A point person on your side who knows the product and can quickly answer questions about business rules.
Common mistakes
- Designing for yourself. Product owners know their product too well and don't notice what's confusing to a newcomer.
- Skipping empty and error states. The product looks great with perfect data but falls apart when a list is empty or the server returns an error.
- Testing on coworkers. Employees already know the product and company jargon, so they won't notice what trips up a new customer.
- Treating UX as a one-time job. Products evolve, with new features and new users. Revisit user flows after every major update.
- Copying big-product solutions. What works in a tool with a million users can be overly complex for your audience.
What affects the cost of UI/UX design
- The number of roles and user flows to be designed.
- Research depth: data analysis and a few interviews, or full-scale user testing.
- The number of screens and states, including complex tables and forms.
- Whether you need a design system or a component kit is enough.
- Platforms: web only, or tablet and mobile versions too.
FAQ
Can we skip research and go straight to design? Yes, if the product is simple or you know your users well. But for a complex product, saving on research often leads to rework after launch.
How many people do you need for testing? To find the main problems, a few participants from each key role are usually enough. What matters more is that they're real members of your audience.
How do we know the new interface is better than the old one? Before making changes, we record baseline metrics: how many people complete the key flow, how long it takes, and how many support tickets come in. After launch, we compare them with the new numbers.
Do you work on products built by another team? Yes. We hand off mockups and components in Figma and consult the developers during implementation.
Why Promodex
Promodex has been in business since 2015, with 750+ projects in the agency's portfolio. Products we've worked on include IKURA, Crypto Monitoring, Unibox Warehouse, and Pro ERP. Our designers work on the same team as our developers, so prototypes account for real technical constraints from the start. We explain every decision and, after launch, help improve the product based on data.
