Time for UX research to catch up
Editor’s note: Sara Faatz is the senior director of strategic awareness, digital experience at Progress Software. With 20 years of experience leading corporate marketing and community building for multimillion-dollar international organizations, Faatz has a proven track record of conceptualizing and orchestrating campaigns that evolve the brand and positively impact the company’s image and revenue. She has run both large and small marketing departments, built community programs from the ground up, created partner programs and acted as a brand ambassador for various organizations. Find Faatz on LinkedIn.
For decades, users adapted to digital experiences by learning navigation structures, search syntax and predefined workflows. AI is beginning to reverse that relationship. As people grow accustomed to tools that understand intent, respond conversationally and anticipate what they need, they increasingly expect technology to adapt to them, not the other way around.
That shift is changing the definition of a good digital experience. If Netflix can recommend the right show before users know what they’re in the mood to watch, why should they tolerate endless menus or cumbersome workflows elsewhere? The more often people encounter experiences that remove unnecessary work, the more noticeable it becomes when another experience makes them do all of it themselves.
For UX and CX researchers, that creates an important challenge. A user can successfully complete a task and still walk away dissatisfied. Traditional usability measures are good at answering, “Can the user do this?” Increasingly, researchers also need to ask, “Why did the user have to do this?” An experience can function exactly as designed and still feel unnecessarily manual, slow or unintuitive against the standard users now bring to it.
AI is changing what users consider usable, even when AI is not part of the experience. That means research must capture more than whether an experience works. It also has to reveal where users expected the experience to do more of the work for them.
The new expectation gap
Customers are no longer comparing a digital experience only with those of direct competitors. They are comparing it with the best experiences they encounter anywhere. Whether they are using ChatGPT to summarize information or Spotify to recommend a playlist, people are getting used to experiences that understand context, anticipate needs and reduce effort.
That comparison changes what success looks like. A participant may complete every task in a usability study and still leave dissatisfied because the experience demanded more effort than they believed should be necessary. This creates an expectation gap: the distance between what an experience enables users to do and what users increasingly expect the experience to do for them.
The more often customers encounter experiences that anticipate what they need, the more noticeable it becomes when another experience feels unnecessarily difficult, manual or outdated. The research challenge is to identify where that shift is happening and where an experience that is technically usable is no longer meeting the standard users bring to it.
Researching beyond usability
Closing that gap requires researchers to look beyond whether users can complete a task and examine the effort they expected the experience to remove. Traditional measures such as navigation success and satisfaction remain essential, but they no longer tell the whole story. Researchers also need to understand where users expected the system to recognize context, provide guidance, personalize the experience or simply make a step unnecessary.
During usability studies and customer interviews, that may mean asking different questions:
- When did you expect the system to understand what you were trying to accomplish?
- What recommendations or guidance did you expect to receive?
- Was there a point where the experience felt more manual than it should have?
- If you could have asked the website a question in plain language, what would you have asked?
These questions can surface insights traditional usability metrics often miss. They reveal whether users achieved their goals as well as where the experience required work that they no longer expected to perform themselves. Those moments can provide early signals that expectations are changing, giving organizations an opportunity to respond before the gap begins to affect satisfaction, loyalty or brand perception.
Search behavior offers a window into changing expectations
Search is one of the clearest places to see this shift. For decades, digital experiences were designed around navigation menus and keyword searches, requiring users to understand where information lived and how to phrase a query. Conversational and AI-assisted search is beginning to shift that burden. Increasingly, users expect the system to understand what they mean without first requiring them to understand how the system is organized.
Research from the Pew Research Center found that users are less likely to click through to traditional search results when Google’s AI-generated summaries appear. The Wall Street Journal has also reported declining referral traffic as conversational search becomes more common. For marketers, these trends signal a changing discovery landscape. For researchers, they point to something just as important: People are becoming less willing to translate what they want into the language and structure a system requires.
That is a meaningful change in user behavior. The question is no longer only whether someone found the right answer, but how much work they expected to do to get there. Did the experience feel effortless, or did it require more interpretation, navigation or query refinement than anticipated? If participants instinctively look for a conversational assistant before using navigation menus, or become frustrated after repeatedly refining search queries, the interface may not be unusable. Their definition of usable may have changed.
Researching for an AI-first world
None of this means every digital experience should become AI-powered or conversational. In some situations, customers will prefer a familiar interface, greater transparency or more direct control. In others, they will want a person. For example, a study from Kinsta found that 94% of consumers preferred interacting with a human rather than AI in a customer service setting. That is an important reminder that reducing effort is not the only thing customers value. Trust, empathy and human judgment still matter.
The challenge for researchers is to understand where AI genuinely removes friction, where it creates new frustrations and where customers still value human interaction. Qualitative interviews, moderated usability sessions and journey mapping remain invaluable. But research also needs to explore how customers define a “smart” experience, what builds trust in AI-assisted interactions and when personalization crosses the line from helpful to intrusive.
And “smart” does not have to mean AI-powered. A well-designed conventional interface that anticipates what someone needs may meet these expectations better than an unnecessary chatbot or recommendation engine. The goal is not to add AI for its own sake. It is to research the expectation, not the technology.
In an AI-first world, we must understand if the experience asks consumers to do work they no longer believe they should have to do. That is the shift UX research needs to catch up with.