This is a brief summary of some of the more egregious design mistakes I try to avoid with every project. Though I might not give the recipe to my secret sauce in HOW to avoid these mistakes, I hope this is enough information to serve as guidelines that improve your next design.
Some of these mistakes are near fatal, meaning that, if your team makes many of these mistakes, you can guarantee your design will never reach its potential. Others cumulatively contribute to poor successes. As a rule of thumb, if your design includes 5 of these mistakes, then you can be certain your design is crippling your profits.
It isn’t so much how bad your design is, but how easy it would be for your competitors to leapfrog your design merely by avoiding your design’s mistakes.
To experience how these guidelines apply to your design, I invite you to take advantage of my free UX design review. There is no obligation, this is not some cheesy sales pitch, and you WILL get something useful out of this. For Free! Request your Free review by clicking here.
1. Avoid Solving the Wrong Problem, Very Well!
Most designs I’ve seen are typically great solutions… but for the wrong problem. Before rushing headlong into a solution make very sure you understand the problem, from the users’ perspective. Be able to describe the point of pain or the trigger, identify what the desired outcome should look like. Design for the desired outcome and work backwards.
For instance, users don’t want to buy something, they want a solution to their problem. Help them identify the problems they are having then show them solutions that fit that problem.
2. Delighting the User is NOT a Strategy
While a desirable side effect, merely delighting the user is not a viable UX strategy that improves performance. Solving their problem is. I’ve seen too many websites so focused on delighting the user that they neglected to address the users’ primary needs. Users come to a website to solve a problem. The easier you make it to solve their problem, the more delighted they will be. Once you’ve solved their problem, you can worry about improving their delight. Until then, you are just polishing a turd.
3. 10% Conversion Rates
Many websites are happy to yield 3% conversion rates. A used car salesman has a closing rate of 20%, so most websites are 7 times worse than a used car salesman. And they find that acceptable? 100% of users visit a site on purpose and are looking for a solution. They WANT to find something on your site, and you can only satisfy 3% of them?
If you don’t have double-digit conversion rates, there’s something terribly wrong with your design. It costs as much to design a site that converts at 10% as it does to create one that converts at just 3%. It doesn’t take more effort to convert at 10%, just a better effort. You should wonder why you aren’t converting at 10%. The most common reason is that you are solving the wrong problem, very well.
4. No Report Generators
Think about it, is printing a report the last thing a user does? Of course not, they will always do something else with that report. Instead of creating an elegant design to generate a report, eliminate the report and take the user to a screen where they will take the next step with that interesting data.
Users run reports to do identify two basic things, trends or exceptions. Of a million different things to look for based on thousands of parameter combinations, how is the user supposed to know what to look for?
5. Create Actionable States
Don’t just provide data, present the user with an action item. Users are using your product or site to accomplish something. Try to do as much of the work as possible for them, leaving them with only the most obvious button that achieves their immediate task or solves their problem.
6. Avoid Pretty (Useless) Dashboards
Most companies provide some kind of dashboard to show how well things are going, but the real purpose of a dashboard is to identify problems. Dashboards typically include a number of graphs, charts, or gauges that display the static value of some metric, but they never indicate what to do about poor values.
7. Don’t Rely on Website Analytics
Google is well known for their website analytics, and many people rely on them, but from a UX design perspecitive, they are pretty much useless. They are like footprints in the sand. They tell you where someone went, but can’t tell you who they were or why they went that way. A well-designed site converts at 10%. If website analytics are so good, why do most sites still convert at less than 3%?
8. Avoid Feature-oriented Designs
Many products are organized around the data or features they provide instead of organizing the features into task groups. If your design has only one instance of each feature and requires users to navigate to each feature to use it, then you have a feature-oriented design. These designs force the users to know which features to use for each task. They rely on individual skill and knowledge in order to use correctly and therefore are prone to user error and frustration. A task-oriented design focuses on grouping features by task, even duplicating some features, providing them in each of the tasks where and when it is needed.
9. User Knowledge is Fallible
Your users’ understanding the problem domain is limited by their own knowledge and experience. Your success is dependent on your users individual levels of skill and knowledge. You, as a company, have insight into how ALL of your customers perform their tasks. You can learn from their best practices and design your site or product to guide users down those best practice paths, thus avoiding the dependency on individual skill and knowledge.
10. Don’t Get Stuck on the Obvious
If 1,000 users do the same task 1,000 different ways, the obvious design approach is to design screens that allow anyone to perform the task any of a 1,000 ways. Which usually results in screens no one can figure out. The reality is that if 1,000 people do the same thing 1,000 different ways, it means that no one knows how to do it right, that there is no best practices approach. This is your opportunity to go beyond the obvious and establish a best practices approach that helps most of your users do the right thing.
11. Hide the Technology
For almost every product, the users are not engineers and are NOT enchanted or even impressed by the technology. They just want to get the job done with as little effort as possible. Therefore, it is better to downplay the technology and make it as transparent, as possible. Moreover, automating some tasks in an obvious and acceptable manner does more for the users than does exposing the technology.
12. SME’s are not Real Users
Though Subject Matter Experts help you better understand the problem domain, they are still so expert that they tend to misrepresent your average users. SME’s are usually far more knowledgeable about both the tasks and the technology than the typical users. If you design for the experts then only the experts can use your product.
13. You are not Thy User
Your design team is far more familiar with your technology than most any user, so much so that they forget what its like to NOT know every detail or nuance about the task. Designers often think in terms of the technology and not in terms of the task flows. A common indication is a function-oriented design approach instead of a task-oriented approach.
14. They are Developers, NOT Designers
This is the number 1 cause of poorly performing designs. It’s not because developers are stupid, quite the opposite. They are actually too knowledgeable about their product to be able to think like an average user. Developers are more apt to design functions, not tasks, and thus create very tech-oriented designs. Don’t force your developers to design the UI. That’s NOT their key skill set.
15. Hidden Improvements
A common design mistake is to change a design so little that its indiscernible from the current interface. When users see something that looks familiar, they tend to expect the new interface to act like the old one. Users will transfer the baggage of a previous interface to the new one, which usually only confuses them and increases their frustration. The more dramatic the change the more different it must look. Users will notice the change and react differently with the new design.
16. Afraid to Start from Scratch
As the say in Maine, “you can’t get there from here.” If your product is solving the wrong problem, not matter how well it solves it, it’s still the wrong problem and changes to the existing product will not take it where it needs to be. Starting from scratch actually enables your team to make truly significant and game changing designs.
17. Technology Drives the Design
The design should drive the technology, not the other way around. Your team will identify new opportunities when they don’t limit their ideas to what they think the technology can do. Start your design with blue-sky concepts and work towards making them feasible. If your design doesn’t require new technology, then it isn’t a new design; it’s just a regurgitation of existing designs.
18. Too Many Bad Experiences
People are conditioned to pull on handles. A door that requires you to push on a handle will frustrate everyone. Replacing the handle with a flat panel won’t delight the users, but it will eliminate that frustration. You don’t necessarily need a great design, just make sure you avoid invoking any user errors or bad experiences. Users will often overlook a few bad design issues, but they are cumulative and can eventually turn users off to your design.
19. Training is a symptom, not a solution
If your design requires training, then your design has already failed. Training is NOT a sustainable solution. Users forget most of what they learn in training. Oftentimes, users go through training and are comfortable with the examples, but are completely confused once they try to perform actual work. Training is a symptom of a design that does not fit the users’ perceptions of their task and requires that the users adapt their thinking to serve the technology instead of adapting the technology to the way the users think. It is far easier to adapt the technology to the user than to force the users to adapt to the technology.
20. Probabilities, instead of Possibilities
To paraphrase Antoine St. Exupery, “Perfection in design is achieved, not when there is nothing left to add, but when there is nothing left to take away.” All too often, designers justify a feature with some faint possibility that the user might want to do something. Though this might seem like an enabling approach, it usually has the opposite effect. The Paradox of Choice is a well-documented phenomenon proving that more choices actually reduces user success. Limit your design to support the probable tasks and avoid the merely possible ones.
21. Automating Current Frustrations
Thomas Edison didn’t invent the light bulb by improving the candle. Too many designs satisfy the users’ current needs, resulting in merely automating bad processes. Oftentimes, user processes are a manifestation of the tools they have available at the time, which were probably poorly designed in the first place.
22. You Can’t Lead by Following
If all you do is copy your competitor’s features and functionality, the best you can hope to achieve is to merely be as good (or bad) as the rest of the herd. What makes you think your competitors are doing everything right? While conventional wisdom suggests that breaking from the herd limits your chances, the reality is that breaking off in the RIGHT direction leads to those blue water opportunities that leave the competitors fighting over chum. Observational user research that is technology agnostic is the most successful method to identify truly innovative ideas that disrupt markets.
23. One Task per Screen
Allowing users to do many things on one screen is not as efficient as you might think. Users get confused about which fields and which controls relate to each other for a given task. Focus on one task per screen. A task may require several screens, but limit each screen to a single task.
24. More is NOT Better
Just because you can put something on a screen doesn’t mean you should. Ask yourself how will this data or UI control change the users’ behavior? If it doesn’t directly support the task at hand, then it doesn’t belong on the screen. While most designers busy themselves with finding things to add, you should be looking for things to remove.
25. Number of Clicks is NOT a Good Metric
This one of those overly prescribed folks remedies of UI design. Unfortunately, like so many other folk remedies, this one results in trying to fit too many controls on each screen, thus overly complicating each screen. Perception of progress is far better than blindly adhering to some contrived metric.
26. Users Don’t Customize
There’s a reason most computer screens have the default desktop images. Users have better things to do than customize your UI. Avoid making your success dependent on users having to customize a screen.
27. No Intelligent Defaults
Users don’t always know how to make the best choices regarding what action to take in each situation, but you do. So instead of depending on the users to figure it all out, limit the choices to appropriate actions and then default to the MOST appropriate action.
28. Empty Forms
Of the three options regarding definable objects (forms, details, etc.) – 1) use as is, 2) edit an existing version, 3) create one from scratch – the most difficult option is to create something from scratch (fill in a blank form), yet this is the most common design approach. The easiest option is to use something as it already exists, but this is also the least likely one to apply to users. Users are most successful when they can simply modify an object that most closely approximates their desired object.
Rather than ask you users to enter parameters before creating a result, show them a result and let them see what happens when they adjust the parameters.
29. Design for IKIWISI
IKIWISI: I’ll Know It When I See It. Users often are not certain what they are expecting as a result of some action but are very likely to recognize a good result when they see it. A design should not depend on the users knowing what to look for up front, but should still return a successful result.
30. Control-Shift-Left Elbow
If there’s a trick to using a control, the users won’t know the trick, and your design will fail. Apple is actually starting to show signs of over-designing their interactions. The latest OS design comes with a number of instructions to introduce these various interaction modes.
Users don’t perform well with non-obvious controls. A good example of custom or hidden functionality is the Infragistics grid. It includes a second header row that allows users to perform a slew of complex functions on each column. Even developers who use this grid can’t figure out what all it does, so how are the poor users supposed to figure it out?
31. Cascading Menus
Though developers like them, users don’t perform well with cascading menus – those menus that break out into submenus. Mega-menus are more successful, but those have limited application.
32. No Instructions Required
Like the door handle that has a PUSH sign above it, if you have to include instructions on a screen, there’s something inherently wrong with the design.
33. Death by a 1,000 Cuts
Tim Ash (SiteTuners.com) uses this term to describe how designers start off in the right direction but, over time, they make one subtle change to the design, then another, then another, ultimately each minor change compounding to create a major design change that adversely affects the UX. No single change creates a failure, but the cumulative effect alters the users’ behavior to the point that the design fails.
Summary
While you may challenge some of these guidelines, they have been responsible for some of the greatest design successes on the internet, in enterprise software and in medical devices for 25 years. I don’t recall ever deviating from these guidelines to create a better design. To experience how these guidelines apply to your design, I invite you to take advantage of my free UX design review. There is no obligation, this is not some cheesy sales pitch, and you WILL get something useful out of this. For Free! Request your Free review by clicking here.