The Invisible Obvious, Part 5: Sound

If you’ve read the first four posts in this series, you’ve seen the same pattern play out across handwashing, typography, body language, and lighting design. Something sits in the environment, treated as background. Someone measures it. The measurements are dramatic. The world resists, then adopts, then forgets it was ever any other way.

Music in retail is in the early part of that sequence right now. Most store operators treat it the way doctors treated hand hygiene in 1840, the way building engineers treated lighting in 1985. Something that’s obviously present, obviously “handled,” and therefore not worth serious attention.

In most retail stores today, the music situation looks like this: someone has a Spotify account or a licensing service. There’s a playlist. It was probably assembled based on personal taste, a vague sense of brand identity, or whatever the streaming algorithm suggested. It plays on a loop. Nobody is measuring what it does. Nobody is connecting the sonic characteristics of what’s playing to the behavioral outcomes the store is trying to produce.

If you ask the store manager whether the music matters, they’ll say yes. If you ask them what it’s doing specifically, at 2pm on a Tuesday versus 11am on a Saturday, for the customer who just walked in versus the one who’s been browsing for twenty minutes, they’ll look at you the way a 19th-century doctor would have looked at Semmelweis. The question doesn’t make sense to them yet. Music is on. That’s the category. It’s handled.

The academic literature on music and consumer behavior is not thin. It’s been accumulating for decades.

Milliman published a foundational study in 1982 showing that the tempo of background music significantly affected the pace at which grocery store customers moved through the space. Slow music, slower movement, more time in the store, 38% higher sales. Fast music, faster movement, less browsing. The effect was measurable at the cash register.

Milliman followed up in 1986 with a restaurant study showing the same pattern in a different environment. Slow background music increased dwell time and spending. The diners didn’t report noticing the music.

Areni and Kim (the same researchers who later studied lighting and wine) found in 1993 that playing classical music in a wine store led customers to purchase more expensive bottles. Pop music produced no such effect. The genre of the background music was shifting perceived product quality and purchase behavior without the customer’s conscious awareness.

North, Hargreaves, and McKendrick ran a study in 1999 that should have been a turning point. They played French accordion music at a wine display and measured purchases. French wine outsold German wine by a factor of 3.3 to 1. When they switched to German music, the ratio reversed. When asked, customers denied that the music had influenced their selection. The music was doing the work. The customers didn’t know it.

This body of research spans over 40 years. Tempo, genre, mode (major vs. minor key), volume, familiarity, lyrical content, rhythmic density. All of them have demonstrated measurable effects on dwell time, browsing behavior, perceived product quality, spending, and pace of movement. The academic case is as strong as the early lighting research was in the 1990s. Probably stronger.

So here’s the question the first four posts in this series set up: if the research is this clear, why hasn’t music become a retail design discipline the way lighting did?

Part of the answer is that music is harder to spec than light. A lighting designer can point to a Kelvin number, a CRI rating, a lux measurement. The vocabulary for specifying music at the production level, tempo, harmonic complexity, rhythmic density, frequency distribution, lyrical sentiment, is technical in a way that most retail operators have never encountered. There’s a translation problem between what the research says and what a store owner can act on.

Part of the answer is the licensing model. The infrastructure that exists for putting music in stores is built around catalogs of existing recordings. A store picks from what’s available. The idea of composing music specifically for a retail environment, the way a lighting designer specifies fixtures for a specific space, barely exists as a category. You can choose from the menu. You can’t write the recipe.

And part of the answer is the same identity cost that shows up in every post in this series. If the music matters at the level the research suggests, then every store that’s been running a Spotify playlist has been leaving that variable uncontrolled. Every day. For years. That’s the Semmelweis problem applied to the retail floor. The implication is uncomfortable.

Lighting made the transition from infrastructure to design discipline in about fifteen years. The research accumulated, the technology caught up (LEDs made precise control affordable), and the early adopters demonstrated enough ROI that the rest of the industry followed.

Music is at the beginning of that same arc. The research exists. The technology to compose and deploy store-specific, outcome-aware music exists now. The measurement infrastructure to close the loop, connecting what’s playing to what’s happening on the floor, exists now. The missing piece was always the same one that was missing in lighting before LEDs, in nonverbal communication before video analysis, in typography before digital A/B testing. Someone had to build the tool that makes the knowledge actionable.

Every example in this series follows the same path. The phenomenon was always real. The evidence accumulated for years before anyone acted on it. And when the transition finally happened, it happened fast and the amnesia was total.

Five years from now, the idea that a retail store would play music without understanding what it’s doing to customer behavior will sound the way “we don’t need to wash our hands between the autopsy and the delivery” sounds today. Obviously wrong. Hard to believe anyone ever thought otherwise.

We’re building for that moment.