For years, UX design has focused on interactions: buttons, menus, forms, dashboards, onboarding flows and all the small decisions users need to make to get something done.
But what if the next step in UX is not designing better interactions, but removing some of them completely?
Good UX has always been about reducing friction. Increasingly, that means products should understand enough about the user, their context and their previous actions to avoid asking unnecessary questions in the first place.
From reactive to proactive UX
Most software is reactive. The user clicks, searches, configures, filters or confirms, and the interface responds. Every interaction adds a small amount of effort. A more proactive product can use context to reduce that effort.
- A project management tool could notice that a deadline is approaching while several dependent tasks are still blocked and surface the risk automatically.
- A SaaS dashboard could detect an unusual change in usage and explain it instead of waiting for someone to discover it inside a report.
- A form could remember information the user already provided instead of asking for it again.
The interface does not disappear. The amount of work required from the user simply becomes smaller.
Sometimes the best feature is one less interaction
There is a natural tendency in product development to solve problems by adding things.
- Add another notification.
- Add another dashboard.
- Add another filter.
- Add another button.
But sometimes the better question is: does this interaction need to exist at all?
If users repeatedly apply the same filters, perhaps the product should remember them. If there is an obvious next step in a workflow, perhaps it should already be prepared. If the system already knows the answer with high confidence, perhaps it does not need to ask again. Individually, these improvements look small. Together, they can completely change how a product feels.
Users stop operating software and spend more time doing the actual work.
AI makes this much more interesting
Traditional interfaces depend heavily on predefined flows. Designers anticipate what people might want to do and create screens and controls for those scenarios.
AI gives software a much better ability to interpret context. That means products can gradually move from:
“What do you want to do?”
to:
“It looks like this needs your attention.”
For enterprise SaaS, that could mean automatically identifying project risks, summarizing activity, detecting blockers or preparing likely next actions. And importantly, this does not mean putting an AI chatbot on every screen. Some of the best uses of AI may be almost invisible, working behind the interface to reduce the amount of information users need to process and the number of decisions they need to make.
Proactive UX still needs control
There is an obvious danger. Software trying too hard to be helpful can quickly become annoying. Bad proactive UX means excessive notifications, unwanted recommendations, incorrect assumptions or automatic changes users do not understand. The difference between helpful automation and frustrating automation often comes down to confidence, transparency and control.
Low-risk actions can happen automatically. Higher-risk actions should usually be suggestions. Important changes should be understandable and reversible.
A better question for product teams
Can we infer it? Remember it? Prepare it? Reduce five decisions to one?
Sometimes the best interface is not the one with the most polished components. It is the one that quietly gets out of the way.
When designing a feature, we usually ask:
How should the user do this?
Maybe we should ask another question first:
Does the user need to do this at all?
As software becomes better at understanding context, great UX may increasingly be measured not by how easy an interface is to use, but by how many unnecessary interactions it manages to eliminate.