Most music products have two sides to keep alive at once, creators and listeners, venues and acts, catalogs and the people who license them. The numbers that matter are the ones showing whether those sides meet. Luminate counted 158 million tracks with 1,000 plays or fewer in 2023, as Music Business Worldwide reported, so supply is plainly not the scarce side. Getting a track, a tool or a platform in front of people who keep using it is, and that depends on far more than the feature list.
Pitch decks still open on a product: a creator tool, a recommendation engine, an AI mastering chain, a fan community, a ticketing system. Software is visible and easy to demo. The harder story runs ecosystem by ecosystem. A team wins when its region connects talent, capital, venues, labels and universities, and when founders learn to measure, test and price before the money runs out.
Seen that way, a music-tech hub is a go-between. It makes it cheaper for people to work together in an industry where everyone wants something different. Below, Music Tech Hub Portugal is the worked example. It runs an innovation program, teaches product discipline and connects creative people with technology teams. One question runs through the whole piece: what should a team, or a hub, measure to know the ecosystem is working?
A great music-tech idea can still lose to complexity
The money is real. The IFPI Global Music Report 2024 put recorded music revenue at $28.6 billion, up 10.2%, with 667 million paid subscription accounts. Yet that market is hard to enter, because its value chain mixes four things that rarely sit together in one industry.
Fans buy on identity and emotion. Creators protect workflows they have refined for years. Labels, publishers and venues run business-to-business deals with long partnership cycles. Rights law decides what a product may do at all; a sampling feature needs clearances before a single user touches it. A clever product has to survive all four at once.
Mature ecosystems give teams supports for each layer: operators who have scaled multi-sided products, test environments such as venues, festivals and creator communities, distribution partners, data fluency, and legal guidance. Remove them and founders guess. In music, guessing is expensive. The mistakes take recognizable shapes:
- distribution assumptions that collapse at the first label meeting;
- monetization switched on before anyone has formed a habit;
- retention loops that only look like "growth" because acquisition is masking churn;
- AI features that delight once and never become routine;
- rights exposure that stops the business cold.
A hub earns its place when it turns these missing supports into shared infrastructure, the roads and power lines of a creative economy. A rival can clone a feature in one sprint. The supports around it take a region years to assemble.
Ideas are cheap in music; execution literacy is scarce
Creative industries have plenty of talent. What they often lack is a shared way of doing product work. Teams arrive with real taste, strong engineering and good industry contacts. Yet the way they plan, ship and measure is uneven. Call the missing skill execution literacy: knowing how to ship something, see what people do with it and decide what to change.
The symptoms repeat. Roadmaps are lists of features. "Engagement" is measured loosely and never tied to the value users get. Teams chase growth before users stick around. Pricing is treated as a switch to flip, when it needs designing like any other part of the product. Data gets collected, but nobody uses it to make a decision.
This is where Music Tech Hub Portugal becomes relevant. A hub built around product strategy, metrics and agile delivery is making a specific promise: it helps creative teams build like high-functioning product organizations without losing the culture that makes music products feel alive. The scarce resource is teams that can ship, learn and scale. Taste and relationships get a team to a promising demo. Surviving the second year takes execution literacy.
A quick test separates the two. Ask a team what its users did in week four after signing up. A team with execution literacy answers with a number and the event behind it. A team without it describes the launch party.
Ecosystems compound through loops, and events only start them
A common mistake is to treat ecosystems as a string of moments: a conference, a showcase, a pitch day, a festival activation. Moments help. Durability comes from loops. Each turn of a loop lowers the cost of the next, so progress accumulates instead of resetting every quarter.
The arithmetic is simple. Say each cohort of founders makes a validation test 10% cheaper for the next one, through shared templates, recruited user panels and mentors who have run the test before. After five cohorts a test costs 0.9 to the fifth power, about 59% of the original. A region that runs a showcase every year with no loop stays at 100%. Four loops decide whether a region becomes a real music-tech node.
1) The talent loop turns practice into mentors
People learn → they practice on real projects → they earn credibility → they mentor others → the local baseline rises.
It runs on training, apprenticeships and real shipping opportunities. Inspiration alone never turns it. One count shows whether it turns: how many of last year's participants mentor this year's cohort.
2) The validation loop kills weak ideas early
Ideas get tested quickly → weak assumptions are killed early → stronger ideas get refined → teams learn what "works here" → iteration speeds up.
It needs access to users (creators, fans, operators) and structured experiments. In music those users are easy to reach and hard to read. A room full of producers will praise a tool they never open again. So the loop works only when the test measures what people do after the session, for example the share who return within 7 days without a reminder. Set the pass mark before the session, say 20% returning unprompted, so the room's enthusiasm cannot move it afterwards.
3) The distribution loop needs trust before reach
Startups access partners → products reach real audiences → results attract more partners → partnerships get easier → distribution becomes less fragile.
It runs on network-building and credibility, which makes it the slowest of the four to start. A venue or label lends its audience to a startup only after someone it trusts has vouched for the team.
4) The capital loop follows evidence
Teams show traction → investors understand the category better → more capital enters → founders can iterate longer → better outcomes appear → the narrative strengthens.
It turns only when the evidence holds up under an investor's scrutiny. Retention curves persuade. Event attendance does not.
A hub can feed all four loops. It becomes powerful when it makes at least one of them repeatable. Music Tech Hub Portugal's positioning suggests a fit for the validation loop (product discipline, metrics, experiments) and the talent loop (training and capability building). The two feed each other. People trained in experimentation make validation cheaper, and every test that settles a question gives the next cohort a worked example.
Lisbon lowers the cost of assembling a coalition
Lifestyle explains part of Portugal's pull. The strategic reason is coordination. A region becomes a useful hub when it is cheap to gather talent across disciplines, run fast tests with real users, sign partnerships with operators and cultural institutions, and keep costs low enough to keep iterating.
Lisbon does well on that list. Web Summit has run there since 2016 and pulls international tech talent into the same city as an active cultural scene. The path from hypothesis to prototype to feedback gets shorter, with no long logistical chain in between.
The trade-off is market size. A cheap base for testing is not a large base for selling, so a team built in Lisbon usually plans for export from the first pilot: contracts, pricing and support in more than one language. The rule follows directly. Use the region for learning speed, and check willingness to pay in the markets that will bring revenue before adding headcount.
Cost works on the same clock as talent. Say a four-person team raises €300,000 in seed money. At a burn of €40,000 a month it has 7.5 months of runway. At €25,000 it has 12. If the team ships and evaluates one experiment every three weeks, the cheaper base buys about 17 test cycles instead of about 11. In an early product, those six extra cycles often decide whether retention is found before the money ends.
The less friction it takes to assemble the "minimum viable coalition" (creatives, engineers, distribution partners and operators), the faster a team learns. Geography hands out the ingredients cheaply. A hub turns a lucky concentration of people into a process that keeps running whether or not any single event happens that month.
One persona for the next 6 to 12 months
A hub earns its keep by spanning two gaps. The first runs from culture to product: turning creative insight into a hypothesis with a measurable value promise. The second runs from product to market: turning a working feature set into distribution, retention and revenue.
Teams fall into the first gap without noticing, because a demo that impresses musicians feels like validation. The second gap is louder. It shows up as flat revenue. Music Tech Hub Portugal is best read as an attempt to be that bridge institution. For its own framing, the public reference point is the initiative's site, techmusichub.com.
The most common way to fall into either gap is persona confusion. Music-tech products serve many groups at once: creators, fans, venues, labels, managers, publishers, brands. A product that tries to please all of them delivers weak value to each.
A ticketing tool built for promoters, artists and fans at once tends to ship three half-finished dashboards. One that wins promoters first earns the right to serve the other two. Structured product thinking forces four answers:
- Who is the primary user for the next 6 to 12 months?
- Who is the economic buyer, and is that the same person?
- What is the smallest segment where the value is painfully obvious?
- Which behavior proves the value was delivered?
"Market size" is deceptive in music. A product can address a huge market on paper and still have weak pull if it does not dominate one motivated segment. A hub's contribution here is mostly permission: an outside voice that makes it acceptable to serve one persona well before serving all of them poorly.
Two-sided music products live or die on liquidity
Many products in a music-tech pitch deck are marketplaces in disguise. A sample library connects producers with beatmakers, a booking tool connects venues with acts, a sync platform connects music supervisors with catalogs. For any of them, the first thing an analyst checks is whether the two sides actually meet. That is liquidity, and it can be measured directly:
- Fill rate: the share of requests (a search, a booking inquiry, a sync brief) that end in a match within a set window, say 7 days. A venue that posts a date and hears nothing back rarely posts a second one.
- Supply utilization: the share of listed creators or tracks that got at least one meaningful interaction in the period. The Luminate count is this metric at industry scale, and it is low.
- Creator retention: the share of creators active in month one who still upload, update or respond in month three. Creators who leave take their catalog's freshness with them, and fans notice later than the dashboard does.
- Time to first match: days from a creator's first upload to the first booking, sale or placement. Long waits tend to come before creator churn, so the metric works as an early warning.
The trap is reading these as totals. Say a beat marketplace grows from 2,000 to 5,000 producers in a quarter while buyer requests rise from 1,000 to 1,400 a month. Requests per producer fall from 0.5 to 0.28 a month, so the typical producer now waits more than three months for a single inquiry. The upload chart looks like growth, yet most of the producers who joined that quarter will leave before they see a sale. The fix usually sits on the demand side, or in narrowing supply to the genres buyers ask for, and the number that shows it is requests per active creator, tracked by joining cohort.
AI demos raise the bar for trust and measurement
AI cuts the cost of an impressive demo. It also raises the bar for a reliable product. Generating a striking output is now easy. Users, though, adopt workflows, not outputs, and in music those workflows are tied to reputation, rights and community trust.
The ecosystems that win the AI era will produce companies that can answer five questions without hand-waving. What does the AI feature do inside a real workflow? Which segment benefits most, and why? Does it lift activation and retention, or only novelty? How are provenance, permissions and user control handled? What is the safety and moderation story?
The novelty question is the one most teams fail. A mastering feature that 40% of users try in week one and 4% still use in week six is a demo, whatever its launch numbers say. A hub that pushes analytics rigor and early assumption testing helps teams avoid this kind of "AI theater": features that shine on stage and do not stick in real use. The cheapest discipline is to put week-six usage next to week-one trial in every demo-day deck.
The hub never works alone. A healthy region needs several other players. Universities train people who understand both music and code. Studios, venues and festivals let teams test products on real audiences. Labels, publishers and managers control access to catalogs. Founders and angels pass on what they learned. Investors need to know how long music businesses take to grow. Public institutions fund grants and export programs. Media carry the stories abroad. Each wants something different from a startup: a label wants control of its catalog, a festival wants a smoother night, a university wants jobs for its graduates.
No single actor dominates. A hub reduces the friction between them: a startup finds a pilot partner faster, a festival discovers new tools, investors meet credible teams.
Score a hub on what its cohorts did
Startups pick a North Star metric for a product. A hub can pick one for an ecosystem. Member counts and event counts are the easy candidates. Both miss the point.
A better North Star is how well the teams a hub works with actually execute. Do founders leave with clearer positioning and sharper segmentation? Do they track whether users come back? Do they run fewer tests that only flatter them, and raise money more easily because their evidence is stronger? One simple way to track this for each cohort is to multiply three shares:
Hub execution score = share of teams tracking 30-day retention × share that ran a test with the pass mark set in advance × share that signed a pilot partner
The multiplication is deliberate. A hub that excels at one thing and ignores the others scores low. Take a cohort of 12 teams. Nine track 30-day retention (0.75), six ran a test with a pass mark set in advance (0.50), four signed a pilot (0.33). The score is 0.75 × 0.50 × 0.33, about 0.12. The next cohort reaches 10, 8 and 6 teams: 0.83 × 0.67 × 0.50, about 0.28. That is more than double, and every input can be checked against a team's own records.
Hub operators will say they control inputs and that a bad funding year sinks any outcome score regardless of quality. For the capital loop they have a point. For the other three it holds less well: whether a team tracks retention or runs a clean test depends mostly on what it learned there. Compounding signs follow the same logic: more teams graduating from prototype to repeatable retention, more partnerships with measurable distribution lifts, more alumni founding second companies.
For a founder deciding whether to join a hub, Music Tech Hub Portugal included, the same score turns into a short diligence list:
- Which pilot partners can it introduce by name?
- What are the participation terms, in equity, fees and weeks?
- What did earlier cohorts reach on retention and pilots?
- How many alumni went on to found again?
A hub that answers with event photos is an events brand. One that answers with named partners and cohort numbers is doing infrastructure work, and that one is worth a team's scarce weeks.