Service · Software testing and QA

Usability and UX testing

We are not checking whether a button works but whether anyone can find it. A technically flawless form can lose a sizeable share of customers at a step the team considers self-evident. Sessions run online: the participant stays at home on their own computer or phone, so we see real conditions rather than a lab.

5 people
usually enough for one round
Recording
of the screen and spoken comments
Phone first
because that is where most traffic is
Priorities
ranked by effect on sales

Included in this service

The method is simple and inexpensive: the participant receives a goal, and the moderator stays silent and offers no hints, however tempting that may be.

Talk the scope through with an engineer

Tasks

We set realistic goals, for example “find a present under PLN 150 and order it to a parcel locker” or “book a GP appointment for next Tuesday”. No instructions on where to click.

Observation

We record where the participant pauses, backtracks, searches blindly or gives up, and ask them to think aloud as they go.

Devices

Most sessions happen on phones, because something perfectly legible on a 27-inch monitor often disappears below the fold on a 6-inch screen.

Accessibility

Keyboard-only use, the NVDA or VoiceOver screen reader, contrast and legibility against WCAG 2.1 AA. For public bodies this is a legal duty; whether the European Accessibility Act covers your shop is worth checking with a lawyer.

Forms and checkout

The most treacherous spot on any site: pointless fields, vague error messages, data wiped when going back a step, compulsory account creation.

Prioritised report

Problems ranked by their effect on conversion or task completion, each with a short clip showing it and an estimate of the effort needed to fix it.

How we work together

A study with five participants is a short project, and the plan gives the report date, task preparation included.

01

Research question

We agree what we want to learn, such as “why do so many customers drop out when choosing delivery”.

02

Recruitment and tasks

We find participants from your target group and write tasks that describe a goal rather than a sequence of clicks.

03

Online sessions

Calls on Teams or Google Meet with screen recording, during which we deliberately do not help.

04

Analysis and advice

We separate one-off quirks from problems that recur across several people and propose specific changes.

Analytics show the effect, observation shows the cause. Google Analytics will tell you half of visitors abandon the basket at the delivery step. It will not tell you they simply cannot see the “Next” button hidden under the pickup point map. A participant thinking aloud explains that in five minutes.

Questions and answers

Ideally from existing customers or people in your target group, reached for instance through a newsletter with a small reward. Failing that, staff from other departments who have never used the product will do. The people who designed the interface are no good: they click the right spot by reflex.

Yes, and it is the cheapest moment to do so. A study on a clickable Figma prototype lets you rework the structure while a change still means a few designer hours rather than weeks of development.

For one round, yes: the first participants stumble over the same obstacles, and then observations start to repeat. Two rounds of five with fixes in between reveal more than a single round with twenty.

Participants sign informed consent stating the purpose and retention period. Recordings are kept on servers in the EU, shared only with people who need them and deleted after the agreed time.

See your service through a customer's eyes

Tell us at which step your analytics show the biggest drop-off. We will run sessions and show you the real reasons.

Hours
Mon-Fri 8:00-18:00 CET, reply within one working day
Meetings
Online via Teams or Google Meet

We set strictly necessary cookies only: they keep the site running and remember the city you chose. Nothing here is used for advertising or tracking. More in our privacy policy.