UX/UI · 2026 · CONFIDENTIAL (NDA)

UX Audit

following users from the first click

CLIENTunder NDA.
INDUSTRYunder NDA.
RESULTSunder NDA.
CLIENT
INDUSTRY
RESULTS

CONFIDENTIAL

CONFIDENTIAL

ROLE

UX auditor in a three-person audit team

MY SCOPE

Entry paths and the elements that support decisions

TIME

10 days

DEVICES

Mobile 80% · desktop 20%

DELIVERABLES

Client presentation + recommendations

01 / 02

THE BRIEF · THE QUESTION

TRAFFIC

TRAFFIC

↑

↑

Traffic was there.

SIGN-UPS

SIGN-UPS

↓

↓

Sign-ups weren’t.

Where do people get lost?

The audit started with one question: at which step do people drop off on their way to the sign-up form — and what would help them keep going?

03

MY PART

Every way in, every nudge forward.

I mapped the complete path for every way a user can arrive — step by step, all the way to the form, on the phone and on a computer — and checked what helps people decide and what stops them.

01

Entry points

Google ads · Google search · Facebook ads · with a location chosen · without one

02

Navigation

can people go back and still move on?

03

Messages

does the site say what is happening and what comes next?

04

Buttons (CTA)

is it clear what to click — and which matters most?

ADGOOGLELOCATION SETNO LOCATIONNAVIGATIONMESSAGESCTASIGN-UP FORM

04

TEAMWORK

How we worked.

0+
0+
0+

checks on desktop and 130 on mobile — each issue ranked high, medium or low priority

0
0
0

days from brief to the client presentation

0
0
0

auditors, each owning a different part of the path

My part: I clicked through the site the way users do — from Google ads, Google search and Facebook ads all the way to the sign-up form, on desktop and on mobile. Along the way I downloaded reports from Semrush and SEOptimer, looked up the ads in Meta Ad Library, checked link previews, every button and how the site picks a location, and tested pages in the browser console, with and without cookies. Each problem went into Figma as a card — where it happens, what goes wrong, its impact and a recommendation — ranked high, medium or low.

05

METHODS & TOOLS

How I looked at the site.

01

User paths

from Google ads, Google search and Facebook ads — step by step, all the way to the sign-up form

02

Navigation & labels

does the menu follow what people came to do — and do the names say what’s inside?

03

Link previews & ads

what people see before they even click: ads and shared links on social media

04

Search visibility

how the site shows up in Google — and in AI answers

05

Competitor comparison

traffic, keywords and links next to a competitor, plus a benchmark of other platforms

TOOLS — Figma · Figma Slides · Semrush · SEOptimer · Meta Ad Library · browser DevTools · Claude · ChatGPT

Claude and ChatGPT worked as my AI assistants along the way.

06

WHAT WE FOUND

What the audit showed.

Path to sign-up

03

Broken paths to sign-up

→ Every step needs a clear way forward.

Dead ends on the way to sign-up

→ Every step needs a working way forward — and feedback while it loads.

Too many buttons with the same weight

→ One main action per screen — the rest steps back.

Navigation & labels

03

The menu followed the organisation, not people’s goals

→ Make the menu an advisor: build it around what people want to do.

Catchy names instead of clear labels

→ Name every tab so people know what’s inside before they click.

Too many ways of browsing, competing with each other

→ One main way to browse — the rest supports it.

Content & context

03

Key information hidden where nobody looks

→ Put key information where the decision is made.

The site showed the wrong location — so the offer didn’t match what people expected

→ Show people where they are and let them change it.

Readability and contrast on the phone

→ Phone first: that is where most people are.

Visibility

03

Technical issues affecting visibility in Google

→ Fix the basics, so people can find the site at all.

Shared links looked empty or wrong

→ Give every key page a clear preview, so traffic from social media isn’t wasted.

A competitor got more from a similar-strength site

→ Answer real questions in the content — for Google and for AI.

07

THE REPORT

Observation → recommendation.

OBSERVATION · SAMPLE

After choosing a location, the main button drops below the fold on mobile.

HIGH

→

↓

RECOMMENDATION

Keep the button visible in a sticky bar, so the next step never disappears.

PRIORITY

HIGH

MEDIUM

LOW

The outcome: a presentation for decision-makers, with a clear recommendation for every finding.

08

WHAT I LEARNED

Write for the client, not for designers.

“hero section”

“hero section”

→

→

“the first section”

“the first section”

The full audit is under NDA.

NEXT PROJECT →

PMI Silesia