The cheapest input in the store is the one you never measured.

When a senior leader asks what the return on in-store experience is, the deck that comes back covers the expensive things. New fixtures. A lighting refit. More bodies on the floor. A scent program. Each one carries a real capital line and a payback you can argue over in a planning meeting. All of it skips the cheapest input in the building, the music, which is already on, already reaching every person who walks in, and already paid for. The speakers are installed. The license is bought. The only thing left to decide is whether what plays through them is set to move people or set to nothing, and almost nobody has checked.

That makes the math unusual. Every other experience lever asks you to spend more to get more. The music asks you to spend close to nothing and pay attention. The hardware is in the ceiling. The cost of changing what plays through it rounds to zero against a refit. So when the upside is real, and the research says it can be large, the ratio of return to cost is higher here than anywhere else on the floor. That is the whole reason this input is worth a senior leader's time. It is the one place where getting it right does not show up as a number in the capital budget.

What the research puts on the table

The effects under the right conditions are not small. Ronald Milliman ran a supermarket in 1982 and changed nothing but the tempo. Slow music produced roughly $4,600 more in a single day than fast, and shoppers asked on the way out mostly could not say what had been playing. A restaurant that did nothing but turn the volume down saw average checks rise about 15 percent, and the genre did not matter. Another swapped fast music for slow and watched diners stay about eleven minutes longer, order more rounds, and lift bar revenue near 40 percent while the food bill held flat. None of those rooms spent money to get the result. They changed a setting that was already there.

You can read the full read on the research up in the field guide on store music and customer behavior, where the dwell and basket findings are laid out in order. The short version for a buyer is that the lever moves the lines already on your scorecard. Volume sets the pace people move at. Tempo sets how long they are willing to stay. Pace and dwell feed basket and average transaction, the numbers you already report.

The honest part, which is the part that matters

Here is the part most pitches skip, and it is the part that earns the rest. A 2025 conference paper from VU Amsterdam, looking at about 140 Dutch stores, found the overall tempo effect came back near zero. The Milliman result held in one supermarket in 1982 and dissolved at scale. So the move is not to promise a percentage, and I will not hand you one. The effect is real and it is conditional, and the conditions are specific to your room, your crowd, and the hour of the day. What the music should do depends on the brand it sits inside, how full the floor is, and who is actually standing there at eleven in the morning versus six at night.

Which is why the right answer to the ROI question is a method. The method is plain. Work out what a store's customers came in for, and fit the music to that. That is the work I do as a music behaviorist, reading the research that backs it and fitting the sound to your room. Then, if you want to know what it returns in your stores, run it against a control so you can tell what the music did from what you wished it had done. The research tells you the lever is real. Whether it pays in your stores is a question only a controlled test in your own store can answer, and the test costs almost nothing because the input is already running.

So the highest-ratio experience lever you own is the one nobody put on the plan. It is already playing to everyone who walks in, and right now it is set to a default that no one measured against anything. Pull up what is on across your stores and ask who chose it and what they chose it for.

Want a number on it? The talks make the case for a room. The door is here.