Golden rules of a good UI design

Your complete guide to a good UI design: what to expect from UI designers and why we, personally, follow these rules. Approved by YCombinator and AirBnb.

Beauty is an exclusion

Let’s get to the point head first. The cornerstone question about design is, certainly, aesthetic-related. In other words, does such a thing as a “good design” even exists? Naturally, in our era of mutual tolerance we have to presume that “good” and “bad” characteristics should be totally subjective. That is, what some people think is a good design, might look horrific to others. And vice versa. Right? Wrong!

As YCombinator co-founder Paul Graham pointed out in his almost 18 years old essay, “Whatever job people do, they naturally want to do better. Football players like to win games. CEOs like to increase earnings. But if your job is to design things, and there is no such thing as [objective] beauty, then there is no way to get better at your job.” In other words, design (and UI design in particular) falls into this rare category, that excludes relativism. And thrives when unified rules and guidelines are vigorously followed.

Paul Graham “Good design rules”

Let’s summaries, what a good design according to YCombinator’s boss look like.
Good design is simple. Like in math it means that a shorter proof tends to be a better one.
Good design is timeless. Like in science, every proof is timeless unless it contains a mistake.
Good design solves the right problem. That’s why touch screen won the market
Good design is suggestive. Everyone makes up their own story about the Mona Lisa.
Good design is hard. If you’re not working hard on the design, you’re probably wasting your time.

Good design looks easy. But mostly this is an illusion. The easy, conversational tone of good writing comes only on the eighth rewrite.
Good design resembles nature.
Good design is redesign. It’s rare to get things right the first time.
Good design happens in chunks. Meaning, good design is the one following trends. At any given time there are a few hot topics and a few groups doing great work on them, and it’s nearly impossible to do good work yourself if you’re too far removed from one of these centers.
Good design is often daring. Today’s experimental error is tomorrow’s new theory.

UI design and usability

Mr. Graham’s vision, though being announced almost two decades ago, still remain in demand. However, there is one more thing that, perhaps was not so important in early 2000-s. But it’s definitely crucial now, when even good designers are numerous and competition field of products that had been properly designed is really packed.
We’re referring to usability. As Micka Touillaud (award-winning UI designer who contributed to how AirBnb and SoundCloud look like) points out, modern professional UI/UX designer should:

  • Consider product’s information architecture;
  • Run user research;
  • Make user needs evaluation;
  • Be all about usability;
  • Develop metrics and create deliverables based on conversion-related KPIs.

In other words, good UI design is a design that:

  • is focused on a particular customer persona (not on abstract “customers”);
  • had been tested by this particular customers ;
  • and been evaluated as efficient by these customers;
  • easy to use for these customers;
    helps customers to achieve a goal established by product owners in the fastest way.

To clarify, good design is not only beautiful, simple or trendy. It gets product owners exactly what they need in a most efficient way. And this efficiency is proved by customer studies. If a product owner needs more sales, good design is the one that brings sales in a fastest and cheapest way. If a product owner needs newsletter subscribers, a good design is the one that delivers just that. And so on.

Customer studies as part of UI design process

In Alsmark Studio the typical UI design process begins with understanding the problem. Customers usually come with briefs that sound like: “I want to improve design to look like a more mature firm”. Or “I want to make a product look better”. Our goal on this stage is to define, why the customer thinks, the product is not good enough. Or why exactly he needs to look mature. Based on customer’s answers, we create hypothesis.
Then we research actual users, trying to evaluate if the customer’s hypothesis is correct. For this purpose we create and run surveys and user tests, record result, use data analysis to get business insights. We come to the customer with this insights and discuss future roadmap.

Sometimes, we have to redefine problem. Sometimes, we don’t. But in any case, we have tested and measured understanding of things to improve. On the next stage, we create a buying persona and rebuild customer journey. Based on it, we now can come up with the ideas that will a) reflect this user journey, b) will solve the problem redefined on the previous stage.
First ideas are just lo-fi wireframes. But we prefer to test them with users as well. After a new set of result and iteration the hi-fi prototype is created. What do we do next? Certainly, you’ve already guessed it. We run user tests, analyze and iterate. What we get by the end of the day, we consider to be an example of a good UI design. Everything else, beautiful or not, is not it.


Usually, on this stage when we deliver the results to customer, our relationship with this particular design doesn’t end. What we’ve taken from Agile Scrum practices is Retrospective – a discussion that happens after the sprint is over, when parties involved discuss, what went wrong during the sprint and how to improve the experience next time. We’d be exited to invite customers to take part in these discussions, but they usually are too busy with new releases to spend time helping us to improve. However, if you’re designing in-house, we recommend you to include product owners in retrospective discussion. It’ll help you make your good designs closer to perfect.

Back to all Ideas

Previous article

Next article

Share This