Resonate: Stream-to-own

Founded in Ireland in 2015 and established as a cooperative in 2016, Resonate attempted something structurally different. Its stream2own model charged listeners a price that doubled with each successive play of the same track, starting at approximately 0.002 credits for a first listen and reaching 1.022 credits on the ninth play - at which point the listener had paid the equivalent of a full download price and owned the track permanently (Resonate, 2022). The average pay-per-play across all listens worked out at roughly 0.0096 euros, more than double the streaming market average at the time (Resonate, 2020).

The model was ethically coherent and economically interesting. It incentivised discovery by making first listens cheap, rewarded genuine fans by letting repeated listening convert into ownership, and paid artists at a rate significantly above the market average. As of its last reported figures, Resonate had attracted around 2,600 artists and 340 labels (Resonate, 2022).

That figure tells the story. Despite a genuinely novel model and consistent press interest, Resonate never achieved meaningful consumer scale. The stream2own mechanic, however logical, requires listeners to hold a mental model of their listening behaviour that most people simply do not maintain. How many times have you played this track? How much have you paid so far? How close are you to owning it? These are not questions that listeners naturally ask. The subscription model succeeded in part because it made those questions irrelevant. Resonate's model made them central.

There is also a structural problem. Resonate's cooperative governance model, while philosophically aligned with its values, made it dependent on grants and crowdfunding to fund development (Resonate, 2022). Without the capital to build the catalogue depth and product quality that mainstream listeners expect, the platform could not attract the mainstream listeners it needed to demonstrate the model at scale. It remains a functioning platform with a committed community, but it has not grown beyond that.

Bandcamp: Direct purchase model

Bandcamp never positioned itself as a streaming service. It is, and has always been, a direct-to-fan marketplace where artists set their own prices and listeners pay to download music, with Bandcamp taking a 15% commission that drops to 10% after an artist surpasses $5,000 in sales (Bandcamp, 2023). There are no subscription fees, no royalty pools, and no per-stream rates. When a fan buys an album on Bandcamp, the artist receives approximately 85% of the sale price.

By the measures that matter for artists, Bandcamp works. It is the most commercially successful non-pooled model in the western music market. It has supported a sustained ecosystem of independent artists and labels for well over a decade. Bandcamp Friday, launched in 2020, on which Bandcamp waives its revenue share entirely, became a significant cultural moment - fans actively planning their purchases to maximise the money reaching artists directly.

But Bandcamp also illustrates the limits of the purchase model at scale. Its audience is, by definition, people who have made a deliberate choice to seek out and financially support music they care about. This is a valuable and committed audience. It is not a mainstream one. The friction of the purchase decision - the same friction that ultimately limited iTunes - is if anything higher on Bandcamp, because the platform explicitly frames buying as an act of support rather than consumption. That framing is powerful for the audience it attracts. It does not scale to casual listeners.

Bandcamp's recent ownership history is also relevant. Acquired by Epic Games in March 2022, it was sold to Songtradr in September 2023 following Epic's decision to cut 16% of its workforce (Songtradr, 2023). Roughly half of Bandcamp's staff were not offered roles in the transition. The editorial platform Bandcamp Daily, which had been a significant part of the site's cultural identity, was effectively gutted. The platform continues to operate but its future trajectory under a B2B licensing company with very different commercial priorities is uncertain. Even the most artist-aligned model in the market has proven vulnerable to the economics of platform ownership.

Deezer: Artist-centric experiment

In September 2023, Deezer and Universal Music Group announced the Artist-Centric Payment System, first rolled out in France later that year and extended to publishing rights via a partnership with Sacem in January 2025 (Universal Music Group, 2023; Music Business Worldwide, 2025). By early 2026, Deezer reported that 85% of its label partners had adopted the model (Lanternier, 2026).

The ACPS does not eliminate the pool. Streams still aggregate and are divided by share. What it changes is the weighting applied to different types of streams. Artists who can demonstrate an active fanbase, defined as at least 1,000 streams from at least 500 unique listeners in a given month, receive a doubled weighting on those streams (Universal Music Group, 2023). Streams categorised as non-artist noise, including functional audio and suspected artificial streams, are excluded from the pool entirely.

Deezer's own modelling suggests the system redistributes slightly in favour of mid-tier and independent artists with engaged fanbases, at the expense of the very largest acts and long-tail content (Universal Music Group, 2023). No independent audit data has been published.

From a listener's perspective, the ACPS is invisible. Users interact with Deezer exactly as before. The change is entirely in the back-end allocation. This is both its strength and its limitation: it asks nothing of listeners behaviourally, which means it faces no adoption friction. But it also means it cannot claim to have aligned listener behaviour with artist payment, because listener behaviour has not changed at all. What has changed is an algorithm.