Identity Portfolio website · Case study
For a filmmaker & visual artist
Portfolio redesign

Vladimir Zhdanov makes short films and paints. I designed his portfolio site, tested it with visitors and rebuilt the structure around what each of them came for.

Role: research, information architecture, UX and web design / Duration: 3 months

+78%
time on site, compared with the first version

In short
The problem

Film contacts and curators landed on the same homepage. Out of 60 visitors, nobody looked at both the films and the art. Half left within three minutes.

What I did

Watched nine people use the site. Then I split it into a film track and an art track, with one About page for the bio and CV.

The result

In the final test with ten people, everyone opened at least one project and nobody needed the back button to find their way.

The shipped site, desktop and mobile
01 The first version
The first version

Vladimir writes and directs short films, and he paints. He needed one site that shows both careers.

He had three conditions: nothing on screen should compete with the work, no bare list of titles, and no site-builder look.

I built the first version around how he sorts his own work: folders for short films, animations and artworks, each with a count. He liked it, and so did I.

Early layouts for the first version.

Brainstorm process  
02 Analytics
Nobody saw both careers

First I looked at how people used the live site.

60
visitors tracked
0
saw both the films and the art
1 in 9
left within three minutes

A producer could leave thinking Vladimir only makes films. A gallerist could leave thinking he only paints.

The analytics couldn’t show why people left. They also didn’t separate the film and art sides, so I couldn’t see where each group dropped off.

03 Testing
Testing the first version

To find out why, I ran nine moderated sessions on desktop. Each one had a short interview and three tasks, done thinking aloud. I logged the first click, wrong turns and time to the key piece of work.

I also compared filmmaker, artist and designer portfolios on where you enter, how you move around and how the work is captioned.

The three tasks
Film contact

Decide within two minutes whether to get in touch.

Curator

Find a series of four works for one room.

Recruiter

Find the CV, recent exhibitions and contact details.

Moodboard and the competitive audit
04 What I found
Visitors couldn’t tell where to start

The folders made sense to Vladimir and to me. Visitors had to decode them before they could see any work.

Nobody could tell what the site wasMost testers couldn’t say in the first few seconds whether this was a film practice or an art practice.One career hid the otherPeople saw one side and assumed it was the whole practice.Attention ran out at sixAfter six items, testers stopped opening work and scrolled to the end.
Film and art need different captionsArt viewers asked for medium, size and year. Film viewers asked for role and production.The CV had no homeCurators and recruiters looked for the bio, CV and awards on the homepage and gave up.
Audit of the first version
Four naming systems in one listFormats, years, series and moods sat side by side.
Color louder than the workThe palette set the mood before any work was visible, which broke Vladimir’s first condition.
Mobile copied desktopThe list turned into a wall of thumbnails.
05 Audiences
Each visitor comes with one question

I asked gallerists, studio recruiters and people in the film industry what they look for first.

Film contact

Where’s the showreel? Usually browses on a phone and decides fast.

Curator

What is this body of work? Reads slowly on desktop and needs series, medium and context.

Recruiter

Where’s the CV? Wants the bio, CV and awards without hunting for them.

Film contacts and curators each got their own track. Recruiters only need one page. They got an About page linked from every screen, which kept the homepage to two options.


Three personas & three first questions
06 Working with Vladimir
Working with Vladimir

The new structure meant changing a site Vladimir already liked. Before redrawing anything, we agreed on four metrics. After that, every proposal came with the session recording behind it, so our discussions were about what testers did.

He had the final word on the look. I owned the structure: research, the audience model, sitemap and navigation, caption rules, wireframes and the component set.

From the audit of the first version to a layout system Vladimir can extend.

Four metrics, agreed before the redesign
01Total time on site02Drop-off from home into projects03How many reach the CV04Whether visitors cross between film and art
Signed off before the redesign
07 Structure
Two tracks and an About page

I re-sorted all 31 pieces by audience and got two tracks. The homepage now asks one question: film or art?

FilmShowreel first, then projects by year.ArtGrouped by medium: traditional, digital, concept art, storyboards and comics.



WireframesI tested the structure in lo-fi before any styling: a homepage that asks film or art, and one project page layout shared by both tracks.






08 Second round
Four things broke in the prototype
The portrait took over

A centered portrait pushed the work below the fold. Testers called the page “a profile”. Now the name reads first, the portrait sits on the right and the first piece of work is on the opening screen.

The list was too long

The sidebar showed every title at once. I capped it at six projects, where attention ran out in the first round, and moved the rest into Everything else.

Both tracks looked the same

Identical blocks made film and art read as one feed. Each track got its own grid and a fixed caption line: medium, size and year for art, role and production for film.

Too much text up top

Long descriptions buried the work. Each page now opens with a short intro, and the context sits below the first screen.

The system underneath

The interface is black and white, so the color comes from the paintings and film stills. Merriweather for reading, Montserrat for titles, an eight-column grid on an 8-pt scale, and four reusable blocks: sidebar item, media block, text wrapper and two-column. Desktop and mobile share the type scale and spacing. A new film or series is built from the same blocks.

Type, color, rhythm, layout
09 Results
Results

Time on site comes from analytics on the live site, compared with the first version. Everything else comes from a final test: ten people, the same three tasks, on desktop and mobile.

+78%Time on site, compared with the first version.8 of 10Testers moved between the film and art tracks.0Testers needed the back button to find their way.
The four metrics
01Time on siteThe only metric with a clean before and after. Longer visits can mean people are lost. In the final test nobody needed the back button, so I read it as people looking at more work.02Drop-off into projectsBetter, but I don’t have a clean figure. Everyone opened at least one project. The old analytics didn’t separate the two tracks, so there’s no baseline.03Reaching the CVSeveral testers opened the bio without being asked. In the first version it sat inside the homepage scroll, and nobody found it during the tasks.04Crossing between tracksThe two who didn’t cross went back to the homepage and continued from there. The tasks covered both sides, so real visitors may cross less often.
The first version and the final one

First version: color first, centered portrait, work below the fold.

Final: name first, portrait on the right, film and art as two tracks.
10 Reflection
Reflection
The hardest partChoosing what to show from Vladimir’s portfolio. 
A producer or a curator needs a few strong pieces and a quick sense of how he could help them. So the film track opens with the showreel, the list stops at six projects and the rest sits under Everything else. 

What I enjoyed mostTurning a pile of films and paintings into a structure I could test with real people. As well as our collaboration proccess with Vladimir, whome works I 
Let’s talk Back to top