The short version
- A design fail is the gap between what a product offers and what a user can actually find.
- Three products that hundreds of millions of people use daily (Spotify, Facebook and Google Analytics) each carry one that has survived years of iteration.
- They survive because scale absorbs the cost. The same fail is enough to sink a six-month-old product.
- The four costs worth putting to a budget-holder: abandoned journeys, support volume, early churn, and a late redesign (commonly cited as up to 100× the cost of fixing it at design time).
- Use the Four Questions UX Audit at the end to catch yours.
Design fails are the small breaks between what a product offers and what a user can actually find. A playlist buried three levels down. A privacy setting nobody can locate. A report nobody can build. Even the biggest companies ship them, and this article shows three, what each one costs, and how to catch yours.
It is written for whoever signs off the roadmap: a founder, a product owner, a budget-holder. The examples are consumer products because everybody has used them, but the argument is about what the same fail would do to yours. We have seen before how good UX can be determinant in some markets; here I want to look at the other side. What does bad UX actually look like, and how are we affected by it? From smaller businesses to high-end companies, nobody is immune.
These mistakes will probably never trouble companies at this level. Most users keep going, quietly annoyed, stepping over the obstacles the UX designers left behind. The same mistake is enough to seal the fate of a new product from a small company.
Scale is what buys the tolerance. Someone who cannot find Discover Weekly stays anyway, because everything else in the library holds them there. Someone who cannot find the equivalent feature in a six-month-old product has no reason to stay at all.
So the bottom line is to understand your ideal user, then make sure everything in your product behaves the way that user expects. It takes time. You get the time back later. We saw this first-hand with NotaryCam: rebuilding the experience around how people actually used it, rather than how the product had grown, helped the platform go on to power over a million completed notarizations.
The commercial case rarely gets made, which is exactly why these problems survive so long. Four costs are worth putting in front of anyone approving a budget.
Abandoned journeys. Baymard Institute's meta-analysis of 50 studies (updated September 2025) puts the average online cart abandonment rate at 70.22%. Among shoppers who abandon for a fixable reason, the biggest triggers are unexpected extra costs like shipping and tax (39%), being forced to create an account (19%) and a checkout that is simply too long or complicated (18%). Much of that is friction, not price. Every step a user cannot complete is revenue that does not arrive.
Support volume. A setting nobody can find becomes a ticket. A report nobody can build becomes a request to the data team. The cost simply moves from the product budget to the operations budget, where it is harder to see and easier to keep paying.
Retention. The tolerance that protects Spotify does not exist for a product someone tried last week. Churn caused by navigation is churn that happens before the value has ever been felt.
The cost of a late redesign. Fixing a structural design fail after launch means unpicking flows, data and content at once. The software quality literature has put a rough multiplier on this for decades: the widely cited 1:10:100 rule, most often attributed to the IBM System Science Institute, holds that a problem costing one unit to fix at design time costs around ten in development and around a hundred once it is live. The exact ratio is arguable, and its precise origin is disputed. The direction is not.
Not everything that goes wrong comes down to bad UX design. But there is one thing running through every example here: things are not working the way users intend them to. The path is paved one way, and people walk another.
The list of conditions for a good user experience is long. Every one you meet removes a reason for somebody to leave. So assume less and research more. Watch five real users attempt the journey that matters most to your business, and fix whatever makes three of them hesitate.
When we review a product at Imaginary Cloud, we start with four questions, in this order. They are deliberately blunt, and the first two catch most design fails on their own.
We ask each question against a recording of a real user, never against the product in the abstract. That is why the answers tend to be uncomfortable.
A found fail is not a fixed one, and the order of the repair matters more than its size.
Now, is every design fail an accident? No, of course not. Sometimes what looks like a mistake is a way of driving users towards a certain page or option. That is a dark pattern, and it is a business decision rather than a design error. People do not like being herded, and they do not react well once the reasoning behind those decisions comes out. The lack of clarity around Facebook privacy was a minor issue until it became a public one.
Most companies are making innocent mistakes rather than deliberate ones. Either way, the consequence of misjudging UX is the same: users get annoyed by a product that was supposed to be helping them, and sooner or later they take that annoyance somewhere else.
Burying a wanted feature under unpredictable navigation, making a critical setting harder to reach than it needs to be, using internal vocabulary in user-facing labels, and leaving no recovery path when somebody takes a wrong turn. All four appear in the examples above.
Because scale absorbs the cost. Features accumulate faster than the navigation that holds them, and the users who would otherwise leave are held in place by everything else the product does well. The fail survives because nothing in the numbers forces the fix.
It shows up as abandoned journeys, support tickets, early churn and the price of a late redesign. Baymard Institute's aggregated studies put average cart abandonment at 70.22% (September 2025), much of it caused by checkout friction rather than price, and the widely cited 1:10:100 rule puts the cost of a post-launch fix at around a hundred times the cost of the same fix at design time.
A design fail is unintended friction. A dark pattern is friction placed on purpose, to steer somebody towards an outcome that suits the business rather than the user. From the outside they look identical, which is why intent matters so much once you find one.
Watch a new user attempt the journey you care about most, and do not help them. The point where they hesitate is the fail. The Four Questions UX Audit above turns that into something repeatable.
That is where it pays best. A large company survives a design fail. A new product often does not. The earlier a structural problem turns up, the cheaper it is to unpick.
If any of the three fails above sounded familiar, it is worth an hour of somebody's attention now, before it becomes a quarter of your roadmap. We run the Four Questions UX Audit against your own product and finish with a prioritised list of what to fix and what it is costing you. Tell us about your product and we will take it from there.
Content manager, text editor, and presenter of strategic ideas, along with being an avid fan of the cinematic arts and visual storytelling.
People who read this post, also found these interesting:
