01 About

What is Konext?

Konext is a marketplace connecting artists with creative opportunities through geographic discovery.

01 Context

The world this tool lives in

This is a case study about a two-sided marketplace, not a single-user tool. To understand why geographic discovery mattered, you need to understand who’s actually looking for whom, and why the usual search bar fails both of them.

The scale

A two-sided marketplace covering 6 disciplines. Every artist profile and every client brief passes through the same discovery layer first.

The users

Freelance artists chasing their next paid brief, and the studios and clients trying to find them by style and location not by algorithm luck.

How discovery moves through the system

1. World

   A client or artist opens Konext and lands directly on the live globe: discovery, not a dashboard.

   
2. 

   Filter

   Discipline, location, and availability narrow the globe down to a real shortlist.

3. Profile

   Verified work history and mutual reviews replace a cold, unverified portfolio link.

   
4. 

   Apply

   One brief, one application no mass-pitching into the void.

5. Track

   The 4-step tracker shows Applied → Viewed → Shortlisted → Hired. Silence stops being the only signal.

   
6. 

   Hired

   Messaging opens once both sides have signaled real interest through the tracker.

What is world discovery?

Keyword search

Search “illustrator” and get 3,000+ results, ranked by an algorithm no one can see, blind to where anyone actually is.

Geography discover

A live globe filtered by discipline, city, and availability a shortlist you can actually reason about.

Every marketplace treats discovery as a search bar. Geographic discovery treats it as a map, because where an artist works and where a client is hiring matters as much as what they make. It’s the one decision the rest of the product builds on.

03 Research

Finding out the problem

To find out what was actually broken, I used four approaches:

What the audit actually showed

Findings from the research

• Every platform I audited treats discovery as a single search bar. None combine location, discipline, and availability into one filtered view an artist has to run three separate searches to get what one geographic filter could show instantly.
• Upwork’s paid Connects cost money just to submit a proposal, but give no more visibility into the outcome than a free platform does. The only signal an artist gets back is silence, or a rejection days later with no explanation.
• Behance and ArtStation solve hosting and browsing beautifully, then stop. There’s no path from “someone viewed my work” to “someone hired me” two disconnected products a freelancer has to run in parallel.
• Trust signals barely exist. Reviews are either absent or one-directional a client rates a freelancer, never the reverse so nothing holds both sides of the same deal accountable.

Business & user goals

The product needed to stop being a list of profiles and start being a decision-making tool one that tells an artist whether it’s worth applying before they spend the time.

05 Solution

Fixing what was broken

Before designing anything, I needed to sort out something more fundamental:

Choosing not to hold the money

No matter how well the discovery flow worked, if Konext held payments between artists and clients, it would need escrow licensing, dispute resolution, and compliance overhead more than a solo project can carry. A better UI on top of that model would still be the wrong foundation. I had to remove the problem, not design around it.

Konext stays connection-only. Verified profiles, mutual reviews, and a 3-tier subscription do the trust-building work an escrow layer would otherwise handle at a scale a solo designer can actually build and run.

The key interaction (status that actually means something)

The four states below are the core bet of the whole product each one replaces a moment that used to be pure silence with a clear, visible signal.

Three decisions, in the actual screens

Each one is a specific interface decision, shown where it actually lives not described in the abstract.

06 Final design

Four screens, four solutions

Not a screen dump four places where a specific decision is visible.

07 Impact

What I’m projecting at launch

Konext hasn’t shipped publicly, so there’s no live data yet. These are targets set against the documented industry baseline not results I’m claiming.

These aren’t results they’re the targets the design was built against, each traceable to a specific decision above.

The figures above are projections, not measured outcomes. “Before” values come from the Leapers 2025 ghosting survey and the hands-on competitive audit. “After” values are design targets, not collected data Konext has no live users yet.

08 Learnings

Takeaways from this project

• Silence is a UX problem, not just a business one. The single biggest design decision in this project the 4-step tracker exists purely to replace “no response” with something visible. That reframe came from research, not instinct.
• Shipping changes what you’re allowed to skip. Once I was directing real code instead of static mockups, decisions like “no escrow” stopped being theoretical they had to hold up against a database schema and a deployment.
• A two-sided marketplace needs two-sided empathy from day one. Designing Marco and Sarah together, not sequentially, caught mismatches like the profile needing to sell trust, not just skill that a single-persona process would have missed.
• Removing a feature is still a design decision. Choosing not to show an earned-revenue counter took longer to defend than adding it would have and it’s one of the calls I’d make the same way again.