UX has spent years asking for a seat at the table. We wanted to be involved before a solution was decided, to bring what we know about people into product strategy, and to have a say in what gets built.
In many teams, we are finally there. Then someone asks a fair question: What will this work change for the product or the business?
It is tempting to answer with the measures we know best. Task success went up. People completed the flow faster. The usability score improved. Those are useful signals for us, but they do not yet tell a product manager which problem to prioritise or help a business leader understand the return on an investment.
We need to finish the sentence.
From an experience measure to an outcome
Imagine we are improving an application form. In a usability test, more people complete it without asking for help. That is good news. But what happens when the form is used in the real product?
Perhaps more eligible customers submit an application. Perhaps fewer people contact support halfway through. Perhaps the team spends less time correcting incomplete submissions. Each of those outcomes matters to someone making decisions about the product.
The chain looks like this:
People understand the form → more people finish it → the business receives more complete applications with less support effort.
That last step is a hypothesis until we check it. A usability test can show that the form is easier to use. It cannot, on its own, prove that revenue increased or support costs fell. We should be clear about what we have observed, what we expect to happen next, and what we will measure after release.
Translate the work before you present it
The language changes with the problem we are solving. Here are a few translations I would use to start a conversation:
- Task completion rate → completed applications, orders, or other valuable actions. If more people can finish the job in a test, check whether completion in the live product improves too.
- Onboarding success → activation. Do new customers reach the point where the product becomes useful to them, or do they leave before they get there?
- Time on task → time to value or time spent serving a customer. A shorter process may help people get value sooner or help employees handle more work. Faster is only better if the result is still right.
- Error rate → rework, failed transactions, and support cost. How often does someone need to correct an input, contact support, or start again?
- Findability rate → product discovery and qualified enquiries. If people can find the right option, do more of them take the next useful step?
- Satisfaction or ease-of-use score → retention and renewal. Do people continue using the product, and do customers choose to stay? A good score alone cannot answer either question.
- Task completion for people with disabilities → reach and successful completion. Who is currently unable to use the product, and does removing the barrier let them complete the same job?
These are possible connections, not interchangeable numbers. A better task completion rate in a small study does not automatically mean a higher conversion rate in production. Agree on the outcome you expect, measure a baseline, and check what actually changes after release.
This does not mean turning every design decision into a revenue claim. Some work protects trust, reduces risk, or helps people who are currently excluded. The point is to explain why that matters to the product and to choose a measure that can show whether things improved.
Start with the decision at the table
Before presenting a research finding or a design proposal, I find it useful to ask three questions:
- Which product or business goal is this connected to?
- What are people struggling with that prevents us from reaching it?
- What would we expect to change if we solved that problem?
If I cannot answer those questions, I probably need another conversation with the team before making a recommendation. The goal is not to dress up UX work in business language after it is done. It is to understand the outcome early enough that research and design can help shape the decision.
We still need our own measures to learn whether an experience works. At the table, we also need to show how that learning connects to activation, retention, cost, risk, or whatever the team is trying to improve.
Getting a seat was an important step. Speaking the language of the decision is how we make that seat useful.