AI vs. Traditional In-Store Music: What's Changed

The in-store music industry has operated on the same model since Muzak's founding in 1934. License existing recordings, organize them into playlists, distribute them to stores. The technology for delivery has changed (satellite, streaming, cloud), but the underlying structure hasn't. The music was made for consumers. It's repurposed for retail.

AI-generated music changes that structure at every level. How the music is created, how it's targeted, how it's measured, who owns it. This isn't an upgrade to the existing model. It's a different model.

Where the music comes from

Traditional providers license recordings from labels and publishers. The catalog is finite. Even a library of two million tracks is a fixed set, and playlists built from it rotate and repeat. Employees hear the same songs dozens of times a month. Frequent customers notice.

Every track also carries licensing obligations. ASCAP, BMI, SESAC, direct label deals. For a multi-location retailer, the compliance burden and cost scale with every store.

A generative system produces original music from scratch. There is no catalog. Every composition is new, non-repeating, and built for the specific retail environment it will play in. No licensing fees, because there's nothing to license.

How targeting works

Traditional services target by genre, era, or mood label. A retailer selects "Upbeat Pop" or "Chill Acoustic" and the platform delivers a playlist matching that description.

Genre is a blunt instrument. The variation within any genre is enormous. Two tracks both labeled "jazz" can produce completely different behavioral effects in a store. A Miles Davis ballad and a Buddy Rich big band number have almost nothing in common from the perspective of what they do to a customer's arousal, pace, and purchase behavior. Genre tells you nothing actionable at the resolution that matters.

Entuned targets at a different resolution entirely. We take psychographic data about the retailer's target customer and translate that profile into a precise musical specification. The output might resemble jazz, or ambient electronic, or indie folk, or nothing categorizable. The genre is a byproduct. The specification is the point.

The practical difference: a traditional provider can give you "Jazz" or "Not Jazz." Entuned can give you music specified at the level the behavioral research actually operates at, built to produce a specific effect for a specific customer profile.

Whether anyone is measuring

This is the most consequential difference. Traditional music providers do not measure the behavioral impact of their playlists. They can't tell a retailer whether the music playing at 2 PM on a Tuesday helped or hurt sales. The audio is an input with no measured output. An expense line with no return calculation.

Entuned's system is built around measurement. We control every property of every track we generate. We log what played, where, and when. We correlate that data against retail performance metrics. That creates something traditional systems structurally cannot produce: a feedback loop between what's playing and what's selling. Over time, the music gets better at doing the thing you deployed it to do.

For anyone who has to justify line items, this is the difference between an overhead cost and an investment with documented returns.

What the licensing economics look like

Traditional providers charge $15 to $50 per location per month for the software, plus $30 to $100 or more per month in performing rights organization fees depending on store size and territory. For a retailer with 50 locations, the annual cost of licensed music can reach $50,000 to $90,000. That buys access to a catalog of songs that were never designed to move a business metric.

The cost comparison to Entuned isn't apples-to-apples. The relevant comparison is cost-per-outcome, not cost-per-location. Traditional music has no measured outcome. Entuned's model is built on the premise that the music should pay for itself through measurable performance improvements.

How to evaluate the decision

If a retailer doesn't intend to measure the impact of in-store audio on business outcomes, a $30-per-month playlist service is fine. The music fills silence. It probably doesn't actively harm the business. It costs very little.

If a retailer intends to treat audio as a performance variable, to measure its effect and hold it to the same standard as visual merchandising or digital spend, the catalog model can't deliver that. You can't measure what you don't control. The shift to generative, data-driven audio is about having any data at all.