Music Mentoring Dashboard for
Noise Solution
Noise Solution | Data ChangeMakers
Power BI
The Project
Noise Solution is a multi-award winning social enterprise delivering evidence-based digital music mentoring to young people across the East of England. Their mentoring sessions are more than just music, they are structured interventions. Each session is recorded and processed through a native AI model that scores the interaction across three core psychological needs: Autonomy, Competence, and Relatedness.
The result is a rich dataset sitting at the intersection of music, youth development, and behavioural science. But without a structured way to read that data, it was just scores.
I took part in the project to dig into these scores and surface insights for the program managers. The people responsible for understanding whether the mentoring was actually working.
Why It Mattered?
Program managers needed to understand:
Which psychological needs were being met and which were not?
Were session lengths connected to engagement, or just to age?
Were certain age groups being undeserved?
What were the patterns in how long young people stayed in the program?
These questions mattered because the answers had real implications. How mentors were trained, how sessions were structured, and how the program made the case for its own effectiveness to funders and partners.
Without a clear view of the data, those decisions were being made on instinct. The dashboard needed to change that.
Digging Into the Data
Before building anything, we needed to understand what the data was actually showing us. Some of what we found was genuinely surprising.
Competence stood out as the strongest psychological need and it shows up from session one. Music-making seems to create an almost immediate sense of "I can do this." That's a significant finding for a program trying to build confidence in young people.
But young adults scored lowest across all three needs, yet stayed in the program the longest. That tension raised some important questions:
-
what keeps them there?
-
Is it the relationship with the mentor or the music itself?
-
Something the scores aren't capturing?
Engagement patterns added another layer. Half of participants drop out after session 6, a consistent pattern that warranted attention. Session length also varied significantly by age pre-teens stayed in sessions the shortest, while young adults stayed the longest.
Whether shorter sessions reflect disengagement or simply what works for younger age groups isn't a question the data alone can answer but surfacing it gives program managers something concrete to investigate.
Structuring Around Questions
Rather than organising the dashboard around chart types or data tables, we structured it around the questions the program managers were actually trying to answer. This was a deliberate choice.
When stakeholders arrive at a dashboard, they don't want to interpret numbers, they want to find answers. Framing each section as a question gives them a path through the data rather than a wall of it.
The Results
Findings that challenge assumptions.
The low scores for young adults sitting alongside their long-term retention is the kind of finding that changes how a program thinks about its own impact. Competence isn't the only thing keeping people in the room. That's worth investigating.
A clear dropout signal.
The session 6 drop-off pattern is consistent enough to act on. Program managers now have a concrete point in the participant journey to examine and an evidence base for any interventions they might want to test.
Session length reframed.
Rather than treating short sessions as a problem, the data opens up a more nuanced question: are shorter sessions a sign of disengagement, or just what works for pre-teens? That reframe matters for how the program designs and evaluates its delivery.
Data that supports the program's case.
The finding that Competence registers from session one, gives Noise Solution something tangible to communicate to funders and partners, evidence that the music mentoring is creating meaningful psychological impact, quickly.
Closing Thoughts
What made this project interesting wasn't the complexity of the data. It was the nature of what the data was measuring.
-
Psychological needs
-
Music
-
Young people who often aren't well served by traditional support structures
The numbers only mean something if you understand the context behind them. The biggest challenge wasn't technical. It was learning enough about what Noise Solution does, and why, to ask the right questions of the data.
That process, sitting with the findings, interrogating the tensions, resisting the urge to smooth over the surprising results, is what turned a set of scores into a story worth telling.