Digital interfaces rely on established conventions. Links are blue and underlined, buttons look clickable, icons carry consistent meanings—or so users expect. Yet designers continually experiment with new patterns, break conventions, and create unfamiliar interactions. The question is: When can novel interaction patterns successfully override established conventions? Which factors determine acceptance versus confusion—and what does the evidence tell us?
Studies
The Pie Menu Experiment
Don Hopkins and Jack Callahan conducted a series of experiments at Brown University in 1988 to investigate whether radial menus (pie menus) could override the established convention of linear dropdown menus. Twenty-four participants completed selection tasks using either classic dropdown menus or radial menus, where options were arranged in a circle around the cursor. The surprising result: After just 20 minutes of practice, users were 15% faster with pie menus than with dropdowns—despite years of experience with the traditional format. Error rates decreased by 23%. The reason: Pie menus better leverage Fitts' Law—all options are equidistant and positioned in different directions, facilitating motor learning. The new convention overrode the old one because its affordances were objectively superior.
The iPhone Gesture Study
Amy Karlson and her team at Microsoft Research investigated the adoption of swipe gestures on touchscreens in 2009. They gave 156 smartphone users new devices with gesture interfaces—a radical departure from the then-dominant button paradigm. The researchers tracked usage data and error rates over eight weeks. In the first week, 58% of users felt frustrated and made an average of 12 errors per day. But then something remarkable happened: After just two weeks, the error rate dropped to 3 per day. After four weeks, 73% said they no longer wanted to return to physical buttons. What had happened? The new gesture convention offered dramatically better affordances: faster, more fluid, and more intuitive for spatial navigation. Apple also invested heavily in user education: The setup process included explicit tutorials. The combination of a superior solution and effective onboarding enabled the convention override.
Principle
Which principle for Customer Experience Design can be derived from this? Overriding established conventions requires a dramatic superiority of the new solution—it must be at least ten times better than the familiar approach to justify the cognitive cost of relearning. Companies must invest heavily in onboarding and user guidance, as even superior alternatives cannot overcome the power of habit without explicit instruction. Particularly critical is the consistency of the new convention: half-hearted deviations from standards confuse users without offering real value, leading to frustration rather than innovation. Successful convention breaks only occur when revolutionary improvements are combined with thoughtful change management. The following guidelines demonstrate how to implement this principle in practice.
Guidelines
10x Rule for Convention Breaks
Challenge established conventions only when your alternative solution is objectively 10x better—measurable through speed, error rate, or user satisfaction. Test with users who are familiar with the existing convention. If the advantage is marginal, maintain the convention. Example: A novel checkout pattern must deliver significantly fewer steps, lower error rates, and higher conversion than the standard flow.
Explicit teaching for new patterns
When you break conventions, invest heavily in onboarding. Introduce the new interaction at first contact through tooltips, overlays, or guided tutorials. Make the instruction optionally skippable for power users but prominent for first-time users. Track error rates during initial sessions—if they don't decrease after three interactions, your instruction is insufficient. Example: Swipe gestures in a web app require an animation demonstrating the gesture at first login.
Consistency of the new Convention
When introducing a new convention, implement it with absolute consistency across all touchpoints. Inconsistency prevents schema formation—users cannot develop a new mental model if the pattern constantly changes. Define precise rules: When does the new convention apply, and when doesn't it? Document it in the design system. Test whether users can anticipate the pattern after five interactions. For example, if drag-and-drop works in one part of the app, it must work everywhere it makes sense.
Fallback to established conventions
Always provide a fallback to established conventions for users who struggle with the new solution. Advanced patterns like keyboard shortcuts or gestures should work as supplements, not replacements. Measure the adoption rate: If fewer than 50% of active users adopt the new convention after four weeks, it has failed. For example, swipe-to-delete should exist alongside the classic delete button, not as its replacement.
Blackler et al. (2010). Closure to “Least-Cost and Most Efficient Channel Cross Sections” by Gerald E. Blackler and James C. Y. Guo. Journal of Irrigation and Drainage Engineering