Run Demos for the Repeat, Not the Day

A demo feels like a win in real time. Four hours in the store, samples disappearing, units moving off the shelf, a tally at the end that looks like proof. Founders leave a good demo day energized by a number that, on its own, means almost nothing. Forty units moved this afternoon is not the result. The result is how many of those forty people come back next month and buy you again without a sample in their hand. That number won’t show up for weeks. It’s also the only one that decides whether the demo was worth running.

Sampling is the most expensive marketing a young brand does. Product given away, labor for hours, travel, and the spoils on everything opened and unsold. Spend that much, measure it by the applause of the moment, and you’ll happily repeat a program that loses money every time it runs.

The demo doesn’t sell units. It buys trial. Whether trial was worth buying is a question only the repeat rate can answer.

The Day’s Number Lies

The units that move during a demo are the most flattering and least informative number in the building. Of course product moves. You’re standing there handing out free samples and a discount, applying social pressure in a high-traffic aisle. A person who tastes something free and pleasant, on a deal, in front of a smiling human, will often drop one in the cart. That purchase tells you the demo worked as a demo. It tells you nothing about whether you bought a customer or rented a transaction.

The number that matters arrives later and quietly. Of the people who tried you and bought once, how many came back and bought again at full price, with no sample and no person in the aisle. That’s the repeat rate, and it’s where the economics actually live. A demo that moves forty units and produces five real repeat buyers spent a lot to rent a one-time bump. A demo that moves fifteen units but turns twelve of those into repeat buyers built something durable. The second demo looks worse on the day and is worth far more, because repeat buyers are the only ones who pay back the cost of acquiring them.

This is the trap of optimizing for the day. Push hard enough on the discount and the pitch and you can juice the afternoon’s units almost arbitrarily, teaching yourself nothing while you spend real money to do it.

Do the Math You’re Avoiding

Sampling has a unit economics, and most brands never compute it. Start with the cost: the product sampled, the labor for the shift, the travel, the spoils, the units given away. Divide by the number of genuine repeat buyers the demo produced. Not the units moved that day, the people who came back. That’s your cost to acquire a repeat customer through sampling. Now compare it to what a repeat customer is worth to you over the time they keep buying. If acquisition costs more than the customer returns, the demo program is underwater no matter how good the day felt.

Compute it for one reason. The math changes which demos you run and how. The store with the lower day-count but higher repeat rate is your best venue and deserves more of your budget. The high-traffic store that moves units to people who never return is quietly draining you. You can’t see any of that if the only number you track is the register tape from that afternoon.

Build the Demo to Create the Second Purchase

Once the goal is the repeat instead of the day, the demo changes shape. You’re no longer trying to move maximum units in four hours. You’re trying to convert the right shoppers into people who will come back. So you demo where your actual buyer shops, not where the foot traffic is highest, because a sample handed to someone who’ll never return is product poured down the drain. You make sure the demo connects the taste to the shelf. The shopper has to know where to find you next time, at what price, with no one there to point. And the sample has to sell the product the way it’ll actually be bought: full price, off the shelf, alone.

Then you follow the number afterward. Did velocity in that store lift in the weeks after the demo and stay lifted, or did it spike and fall back to where it was? A durable lift in baseline velocity means the demo created repeat buyers. A spike and a return to baseline means it rented a crowd. The weeks after the demo are where you learn which one you ran. Most brands never look, because they already cheered for the day.

A demo’s only honest job is the second purchase. The first one, the one you stood there and coaxed out with a free bite and a smile, was never the point. It was the cost of finding out whether anyone would make the second one on their own. Run demos for the repeat, not the day, and measure the thing that arrives after you’ve gone home. The afternoon’s tally was never the prize. Who comes back is.