Why Raising Venture Capital in Music Tech Is So Rare w/ Max Shafer (Just for Noise) | Ep 37

Three Things You Can Take Away From This Episode

• Validate the problem before you build the hard technology.
• Work your investor list backwards.
• In AI, your data is the real moat, not the model.

‍

Closing a venture round is uncommon in the plugin world. The margins are tight and the market is niche, which makes audio software a hard sell for investors who expect outsized returns. So it is genuinely notable that Max Shafer, co-founder of Berlin-based Just for Noise, managed to close a round for his audio AI startup.

In this episode, Max joins us to trace the full journey: the origin of the company, how he found his earliest testers in our own community, the design decisions that shaped the product, and the months he spent raising capital. If you are an audio developer curious about the business side of building a product, there is a lot here worth paying attention to.

1. The Idea Started With a Real Moment of Friction

Just for Noise began with a broken drum machine. During one of his morning jam sessions, Max reached for a simple kick and hat, and it took him twenty minutes to dig through his library to find the sounds he already had in his head. That was the spark. He knew exactly what he wanted and could describe it clearly, so the difficulty of actually getting it into his project felt like a problem worth solving.

What is interesting is how the team approached it from there. Rather than immediately building a generative model, they started with an empty prompt box connected to a basic search over Max's own samples, with no machine learning behind it at all. They wanted to learn whether people actually wanted this and whether it still made them feel creative. That early validation shaped everything that followed.

2. The First Testers Came From the Community

Max did not run a formal study to find his early users. His first 15 to 20 testers came straight from The Audio Programmer Discord, with a simple and honest ask to try the thing and tell him what they thought. From there he stayed close to his own circle of music-making friends, a few other Discord channels, and Reddit.

Here is where it gets meaningful. People who join a product that early, when it is still rough, tend to stick around and help shape it. Some of those original testers still receive every new version and still give feedback that makes it into the product. That kind of early community involvement is something many founders overlook, and it creates a bond that is hard to replicate later.

3. Designing for the Non-Expert Changed the Product

The empty prompt box taught the team something they did not expect. It turned out to be intimidating. Facing a blank field in the middle of a creative flow, musicians had to stop and put a sound into rich descriptive words, which was even harder for people who do not speak English as a first language.

That insight pushed the team in a new direction. Instead of asking people to describe sounds from scratch, they built a familiar, game-inspired interface with a spider graph of tunable characteristics like saturation and decay. What many people do not realize is how much the interface, rather than the underlying model, determines whether a tool actually feels usable. For Just for Noise, getting that layer right was one of the biggest focuses.

4. Taking Investment Is a Decision Worth Sitting With

Max is thoughtful about the choice to raise money at all. As he puts it, capital is a tool that takes you down one particular path, and once you take it, some of your other options quietly close. He and his co-founder spent a long time deciding before they committed.

He compares bringing on an investor to a marriage. You should only do it with someone you are genuinely ready to work with as a partner. For Just for Noise the reasons lined up: after a year working full-time on savings and grants, they needed to start paying themselves, and they had come to believe their core model could grow into something much bigger than a single drum generator. It is a helpful reminder that raising is not a goal in itself.

5. Starting With the Least Likely Investors Sharpened the Pitch

One of the most useful parts of the conversation is how Max structured his fundraising. He built a ranked list of around 1,000 possible investors, from dream partners down to unlikely fits, and deliberately started at the bottom. By taking those lower-priority meetings first, he could absorb feedback and refine his pitch before reaching the people he most wanted to impress.

Within a couple of weeks of ten to twenty calls a week, he had heard the hard questions and understood the risks, so his top conversations went far more smoothly. He treated the whole process like product development, iterating against real feedback, and he ran it as a focused three-month sprint where he did nothing but talk to people. In the end that came to roughly 170 conversations and seven yeses.

6. The Biggest Gaps Were Storytelling and Differentiation

When Max reflects on what almost held him back, two things stand out. The first was storytelling. His early pitch jumped from having a fun, well-used plugin to selling a model to large companies, with a gap in the middle where the bigger vision should have connected the two.

The second was differentiation, or what investors often call the moat. He was not clear enough at first on why Just for Noise was better than the alternatives and why others could not simply copy it, and he heard that feedback consistently. These two areas, a coherent story and a defensible advantage, turned out to be central to earning investor confidence.

7. The Data Became the Real Asset

The technical side of the story carries a lesson of its own. Open-source models got the team to around 80 to 85 percent audio quality and then stalled, largely because of the data. High-quality audio data is hard to come by. So Max spent a month recording in studios across Los Angeles, New York, Berlin, and London, capturing his friends' drum machines and running everything through analog and digital processing to build a dataset of his own.

That dataset ended up being more valuable than expected. It now underpins the licensing side of the business and lets the company add new instruments by partnering with other sample libraries rather than recording everything from scratch. For anyone building in this space, it is a reminder that the data deserves as much care as the model itself.

Closing

What ties this conversation together is a consistent way of working: validate ideas with real people, iterate honestly on feedback, and stay clear about what you are building and why. That same grounded approach shows up in Max's view of the industry. He is realistic about AI in music, acknowledging that prompt-based tools are here to stay while making the case that the real opportunity is in technology that supports the creative process rather than replacing it. It is an encouraging note to end on, and a useful perspective for anyone thinking about where audio and AI go next.