2019 / FreeBird

Digital jukebox for venues

Our goal was to provide an easy way to introduce music-on-demand in pubs, bars, restaurants, venues and cafés.

FreeBird was a startup created by John Weiss (based in Chicago) and Robbie Swaddle (based in Barcelona). We’ve teamed up in order to make our way to disrupt the digital jukebox market.

Our goal was to provide an easy way to introduce a music-on-demand in pubs, bars, restaurants, venues and cafés for:

  • venue owners (low cost, easy maintenance and profit-making)
  • musicians (providing extra earnings from music-on-demand)
  • venue guests (people who’d like to hear their favourite song in their favourite bar)

As for a startup, the resources and financials were limited. Different time zone and our regular daily jobs were also w “slowing down” factor in this project. We all knew what’s the vision, what is the goal and how we all can benefit from it as well.


Market research

We’ve put a lot of effort in order to recognise the market of jukeboxes (music-on-demand in public spaces). We’ve found that the whole industry is held by just a few players (we took mostly US market under consideration). The main offering from those company was to put. a massive machine inside a venue (cost of $4,000), then monthly fees for music licenses and cost of the machine maintenance (software updates, fixing damages, etc).

We already knew that our solution has already a lot of advantages.

For venues:

  • Low monthly subscription for the venue owners,
  • Easy maintenance (everything controlled through the web or mobile app),
  • No-cost entry (jukeboxes were redundant),

For users:

  • Low fee on ordering a song ($1),
  • Gamification (you could “battle” with other people in the venue on which song will be played next),
  • A convenient way of finding a bar or pub based on your music taste (you could preview the currently played song and see the venue’s Catalog of music),
  • You pay only after your song is fully played.

Business plan

Most of our weekly Sunday catch-ups we spend on making a business plan. This included pricing models, predictions of earnings after rolling out in specific regions. We believed that what we’re trying to achieve makes a lot of sense and bring benefit for multiple parties: earnings estimate for FreeBird (service fee, subscriptions), venue owners (a chunk of the fee for allowing people playing music), musicians (earning extra from the songs being played).

This wasn’t my favourite part during the entire process but teamwork is teamwork. We’ve created a complex business plan with supporting its pitch deck.

Defining personas

As for the design process, we’ve started with creating our personas:

  • John, the punk rock music lover, hangs out with his friends in a pub in his neighborhood somewhere in Chicago
  • Mark, pub owner

User story mapping

Everything starts with a story. I’ve organised a remote session with John and Robbie and we practised storytelling. Each of had a task of writing down a real-life story about going to a place they like and how they picture playing their favourite songs in a venue. We had a bunch of funny stories put together in a crappy way in MS Word. We voted which of these stories are the closest ones to our business idea and the user needs (we couldn’t afford that time to set up any proper user research, however, when I’m thinking about this now, it was just about going to the bar and talk to people).

From these stories, we’ve extracted a backbone of the app: activities and tasks.

Endless (re)design

This project was one of these that you spent too much time on redesigning and polishing the UI without validating on the go. I wanted to tell the entire story with a single prototype (check Version 1 below) and find out if the entire flow makes sense to people.

One of the mistakes we did was to keep redesigning and fixing the UI. The flow was alright but there was the “glitter” missing. After a few iterations, we’ve realised that the bidding flow is crap. Literally crap. Browsing through venues, maps, currently played songs, the search was doing just fine, but the core of the app was very unsatisfying.

That’s why I love Figma. I’ve set up accounts for my teammates, we spend a couple of hours just doodling different ideas, moving things around and basically doing a live design with the “stakeholders”.

The final result was very close to what we have pictured.

Onboarding allows quickly set up a bid preferences, location permission and payment method.
Browsing flow allows to find near by venues with your favorite music (based on the Spotify history)
Checking in at the venue and browsing for music to play.
Bidding flow with notifications when the song is played or overbid.

Family, friends and fools

Now was the time to test it out with people who had no idea about our startup, about the project and had never ever anything to do with using a jukebox. Lockdown was disabling us from testing on location. Happily, we have friends and family (and fools, as in the startup jargon). I shared the prototype with my closest friends and some family members, so did John and Robbie. We’ve received a lot of positive feedback, kudos and wishes to have this product up and running (which translates to “funded by fools”).

Positive feedback can be less valuable than negative feedback. Based on the suggestions we’ve received I took some time to redesign the venue browsing flow. We found out that not only playing your favourite song could be a selling point for the users but also the way of looking for a perfect place for you and your friends, where they play your favourite music.

This is why we have a Version 2 (see below).

Unfortunately,  this was the last approach to make the app happen. Covid-19, limitations on the API from Spotify and huge licensing costs we couldn't afford forced us to get back to our regular lives and focus on other projects.

Exploring technicalities

As a startup, we couldn’t even think about affording to pay all the licenses to the big entertainment corporations. That’s why we spend a lot of time researching APIs available from Spotify. We’ve figured out that there might be a way of controlling a venue’s queue of songs via Spotify Premium on the user’s phone (like a remote control). I’ve personally taken a look into the entire documentation of it and I figured out that we might be able to build a very efficient backend using Python. The code was clear, not too many endpoints to maintain.

The next obstacle was to find out how we can use the Soundtrack Your Brand (a sub-brand owned by Spotify, which uses licensing arranged by Spotify). Then the shit hit the fan... SYB doesn’t allow to pick a track. There are themes and genres you can queue up but you don’t have much of control on the tracks. That was a huge disappointment for the entire team. We’ve already prepared an impressive business plan and pitch decks. We’ve already sent about 100 of them to the VCs to get noticed and at least have a meeting with anyone.


As we were closer to wrap up the first version of the UI and the core flow for the bidding process Covid-19 hit the world. Our entire business model was based on public venues. During the lockdown, we weren’t able to test it in a real setting. Adding to that impossibilities on the technical side and lost motivation with the newest global developments we killed our darlings kept it safe for better times.

My role as Head of Product at FreeBird

Chicago (remote) / Nov 2019 — Mar 2020

A bunch of friends teaming up and trying to build a startup from a simple need. I was in charge of the visual and experience direction of the app and system for venue owners.

I also worked on: