When designing with constraints, get your priorities straight
My favorite thing about UX design is that it’s not some open-ended creative exercise performed in a vacuum. It’s a mechanical dance: the synchronized push-pull tension of time, constraints, budget, technical limits, and opportunities. A lot of what we study in design school assumes some idealized version of design heuristics when the reality of our workplace experience requires us to navigate the everyday compromises of these tensions.
I work primarily with enterprise cloud security and data products; my users are a highly-skilled niche, their requirements are complex, and mistakes are costly (figuratively and literally). And because design teams don’t work in an idealized vacuum either, ideas and features are sometimes shoehorned in at the last minute if we can make them work. In one recent case, I had a request come in for one extra feature late in the design process. I wanted to say no because there wouldn’t be an elegant solution that could fit our design and development timeline. The solution that we had time to implement would be clunky and not easy to discover, but ultimately it would solve the customer’s need if they invested the extra effort to find it. And that’s precisely the kind of compromise we have to make all the time: when do we cut features because they’re not polished enough and when do we include a backdoor solution even if it’s not immediately obvious?
The truth is that we’re rarely releasing the Perfect Product™ but instead balancing that tension of all of our constraints to come up with the best step above the Minimum Viable Product. I propose a seven-step evaluation system for weighing these compromises. The idea isn’t to achieve a 7 every time but to realize that sometimes adding a Level 3 feature (works well, but isn’t intuitive) is better than omitting the feature entirely as in my example above.
The Seven Levels of Scrutiny for digital design:
- Does it work (at all)?
- Does it make sense?
- Does it work well?
- Is it intuitive?
- Is it visually appealing?
- Do the interactions feel natural?
- Does it create delight?
Does it work (at all)?
This one is a no-brainer. The absolute minimum Minimum Viable Product has to work at all. If it’s software, it has to compile. If it’s a webpage, it needs to load. And if it’s a physical object, it shouldn’t fall apart in your hand. In a realistic world, failure modes will still exist and should be handled accordingly, but error handling exceeds the threshold for this lowest bar: it should work.
Does it make sense?
Okay, so you’ve created a product that works. It exists and it’s made its way in front of some users. The next hurdle you need to clear is making sure your users can perceive what they’re looking at and understand how it will help them achieve their goals. This step is where the basics of branding and visual consistency factor in. Page text, headers, titles, and basic navigation should orient the first-time user around the product in a way that doesn’t leave them scratching their head.
Does it work well?
This shouldn’t come as a surprise, but the functionality of a product falls on a broad spectrum. Once you’ve achieved a product that works (at all), now you need to ascend the qualifiers of good, better, and best until your product is robust enough to work well under a variety of conditions. Does a text input break when an unexpected character is submitted? Do screens take an unreasonably long time to load? Are web pages responsive and accessible? Unlike the previous two levels, this one is nontrivial: even the best UX designers struggle to prepare for every edge case or provide peripheral user needs in an elegant manner. This guide isn’t intended to be a perfectionist’s manual, so the definition of “works well” only needs to be relative. If you’re working with personas, this would be a good time to apply it and determine if your primary persona would be able to accomplish a common task under standard conditions. If your users can complete their use cases, even if indirectly, then you’re ready to move on to the next step.
Is it intuitive?
There’s a reason information architects often have their own specialized job title: designing a complex system that can be navigated intuitively is no small task. Navigation paradigms depend and diverge based on the purpose of the product, whether it is for commerce, entertainment, or an enterprise tool. And intuitive product design extends much broader than a site map. Just like a door handle implies whether it should be pushed or pulled, we expect our digital products to indicate their behaviors before we start pushing. Some web design affordances are straightforward: Do your links have a hover state? Do your buttons look like something that can (and should) be clicked? Do you use common and easily-recognized icons for universal actions? When you start designing more complex products, it’s not always easy to indicate the sequence of actions your users will need to take to accomplish their tasks, but if you carefully manage a library of common controls and design patterns you can save your users from moments of confusion.
Is it visually appealing?
It’s important not to overlook visual aesthetics even when evaluating function. A product’s visual styling, graphics, semiotics, and branding all play important roles in guiding a user through tasks. It may be possible to have a functional and intuitive product even without visual appeal, but it’s impossible to create a truly great product without it. Even a small amount of visual polish can elevate a mediocre product into one of much higher esteem.
Do the interactions feel natural?
Building upon the levels below it, interaction design is both an ascension of intuitive design and a more abstract form of it. Just as visual design provides an intangible quality that makes the product look better, interaction design is a less tangible quality that makes the product feel better. This includes motion graphics, loading states, and timing, and it leans on those design affordances from the intuition stage like hover and click states. You can achieve a high-quality product without mastering interaction design, but you won’t create the highest quality product without it.
Does it create delight?
Delight is such an overused term in the UX world that I really hesitate to use it here, but it seems to fit better than any other alternative. Few things in this world are truly transcendent, but in the world of product design, those transcendent products are the ones with mood-altering effects. The ones that give you energy rather than drain you. The ones that can take a chore like paying your car insurance or managing your budget and leave you feeling optimistic. These experiences are not common, but they are achievable once you’ve climbed the ranks of a well-functioning, intuitive, artful, and interactive product. Now that we’re at the top of the pyramid, it’s hard to offer much advice; if I knew the secret to making any product a delightful experience, I’d have a very different job. The cliché of the art world is “I don’t know anything about art but I know what I like.” In that same vein, your user research subjects aren’t likely to hint at what this delight will look like. But once you’ve found it, they’ll make sure you know.
A visual metaphor
Originally I figured these Seven Levels of Scrutiny would serve as a sort of Maslow’s Hierarchy for design (and yes, I know I’m not the first to attempt that). The more I tried to visualize these questions all together, the more I realized they don’t stack very well. Every piece is necessary to achieve that “delight” product, but the others are relatively independent. Thus, we don’t have a hierarchy pyramid but rather a Scrutiny Mountain.
The exceptions to the rule here are the middle two verticals of the mountain. Out of my seven questions I first ask “does it work (at all)?” and then later “does it work well?” which naturally form a spectrum of functionality. Similarly, I ask “does it make sense?” and “is it intuitive?” which also create the ends of a coherence spectrum.
Little imperfect scrutiny mounds
The purpose of this entire exercise is to demonstrate that perfect products are rare and that most of these usability heuristics require compromises and tradeoffs. I abandoned the vertical pyramid metaphor when I realized the result should still stand even when we sacrifice one piece for another. Scrutiny Mountain is aspirational, but more realistically we wind up with products that punch above their weight on visuals but didn’t quite master the interaction design, or they’re robust but not so intuitive. When we evaluate each vertical independently, we more often end up with these Scrutiny Mounds.
Working with constraints is what UX design is all about. Somewhere between a minimum viable product and a blue sky design is the realistic product shaped by these constraints. I introduced seven questions that should be asked of a product, but these aren’t a checklist of what a product needs — these are intended to evaluate what a good product should do and help you decide which corners you’re willing to cut. If the product doesn’t work at all, you need to place your marbles there above all else. If it isn’t the most intuitive, try to make sure it’s at least meeting all of your functional needs. Smooth interactions are great, but they don’t add any value to a confusing product. Use these concepts to shape your goals. Not every product is meant to delight, but every product should triangulate the goals of its users, its builders, and its stakeholders.
