UX research and usability testing: how to find out why people are not buying

The curse of knowledge

You know your site by heart: where everything is, what to click, what each icon means. That is why it seems clear to you. The customer sees it for the first time in their life, on a phone, in passing — and where you see “obvious”, they see a wall. This is called the curse of knowledge, and it is cured by only one thing: watching how real people cope with the site.

Hallway testing: the cheapest research

You do not need labs or budgets. Find 5 people similar to your customers (not colleagues who are in the know), give them a task — “find the price and submit a request” — and watch in silence. Do not prompt. Note where they freeze, where they click the wrong thing, what they ask about. Five people reveal about 80% of usability problems — this has long been proven by research. In one evening you will learn more about the site than in a year of guessing.

The key rule: stay silent. The moment you start prompting, the test is ruined, because the customer will not have you next to them.

What to look at in analytics

The quantitative picture comes from the tools we wrote about in the breakdown of Yandex Metrica:

• Session Replay (Webvisor) — recordings of real sessions: where they abandon the form, where they click in vain.
• Click map — what they click, including non-clickable elements (meaning they expected a link).
• Scroll map — how far down the page they get. If 10% see the main offer, you are showing it to one in ten.

In-depth interviews

When you need to understand not “where they stumble” but “why they do not buy at all”, interviews with customers — and especially with those who left — help. Half an hour of conversation: how they chose, what bothered them, why they picked a competitor. Three to five interviews uncover objections that are in no analytics: “you had no reviews”, “it was unclear whether delivery was included in the price”, “I was afraid they would start calling constantly”.

Typical findings

What comes up almost always:

• people do not find the price and leave — the classic killer of conversion;
• the form is too long, they abandon it at the “Patronymic” field;
• it is unclear what to do next — there is no obvious button;
• the mobile version is inconvenient: tiny buttons, broken layout;
• no trust — stock photos, reviews without names, no visible address or real people.

From findings to hypotheses

Research gives a list of problems. Next comes prioritization (first what affects more people and is closer to the money) and testing the changes. Large edits on high traffic should be validated with an A/B test; small and obvious ones — just implement and watch the trend. This way UX research turns not into a pretty report but into more leads.

Conclusion

UX research is not expensive science but the habit of regularly looking at the site through the customer’s eyes: a hallway test on five people, Session Replay, a couple of interviews with those who left. Start today — give a friend the task of buying from you and watch in silence. We will run a full usability audit and show where leads are lost — get in touch.