The Milliman Study Is Wrong. Sort Of.

You have probably heard the stat. Slow music, 38% more sales. It shows up on every pitch deck from every in-store music vendor you have ever talked to. It comes from a 1982 study by Ronald Milliman. One grocery store. Slow tempo versus fast tempo. The slow condition won.

If you run retail locations, this number has probably crossed your desk at some point. Maybe your current music provider cited it. Maybe you found it in a trade article. The question worth asking is whether it holds up, because 42 years is a long time to run on one data point.

The study was straightforward. Milliman played slow-tempo and fast-tempo music in a single grocery store and tracked how long customers stayed and how much they spent. Slow tempo produced longer visits and higher total sales. The 38% number is real. For its era, the study was well-designed.

But it was one store, in one market, with one customer base. The shoppers who walked in during 1982 in that town were responding to a specific set of conditions that the study could not fully isolate. The question that matters for anyone running stores today is simple: does it replicate?

The answer is: partially. Slow the tempo, people stay longer. That finding has held up across grocery, restaurant, and general retail contexts for decades. Researchers keep testing it and keep getting the same result. The dwell time effect is solid.

The sales number is a different story. When later researchers scaled up to more stores and tighter controls, the direct tempo-to-sales link got weaker. A 2025 study across 140 stores found no overall sales effect from tempo manipulation alone. A 2024 restaurant study found that tempo controlled dwell time perfectly while producing zero change in spending.

The 38% was probably real in that store, for those customers, at that time. But something else about the slow music in that particular store aligned with that customer base in ways the study did not measure and could not isolate.

If the Milliman finding were the whole picture, every retailer would just slow the music down and collect the revenue lift. Some have tried. The revenue lift does not reliably appear.

Tempo is one piece of what a customer hears when they walk into your store. Slowing the beat changes the pace of the room. It does not change whether the music sounds like something your customer would choose to listen to, or whether the style of the music belongs in your space. If all of that is wrong for the person walking through the door, a slower tempo just produces a longer exposure to the wrong music.

That is the part the industry skipped over. For four decades, vendors have cited the tempo finding without asking what else was right about the music in that grocery store in 1982. The slow songs Milliman played were not just slow. They carried a set of qualities that happened to fit the people who shopped there. Nobody measured those qualities because nobody had a reason to think they mattered more than tempo.

Walk into your best-performing store on a Saturday afternoon. Listen. Does the music sound like it belongs to the people shopping there? Not to your staff. Not to your brand guidelines document. To the actual customers spending money in the room right now.

Then walk into your worst-performing location and do the same thing. Most operators notice something off within thirty seconds, even if they cannot name exactly what it is.

The Milliman study proved that music changes how long people stay. That finding is real and durable. The part it got wrong, or at least incomplete, is the assumption that tempo is the variable to control. Tempo is one input. The music has to be right for the customer in ways that go well beyond beats per minute.

Measuring whether your music is working is possible. So is building music that fits a specific customer in a specific space. The question is whether anyone in your current setup is doing either of those things.

For more on how tempo interacts with other variables, see the science of tempo in retail and Does Your Store’s Music Actually Matter?.

For the broader research record on this, see The Science Behind the Soundtrack.