How to Measure a Location's Foot Traffic Before Installing a Machine

The most expensive mistake in the coffee business is made before the equipment is bought. A weak location can hardly ever be fixed with marketing: if a point does not reach plan in the first weeks, it is cheaper to move the machine than to rescue it. A move costs money, downtime and nerves, so it is cheaper to spend a week on measurements in advance.

Below is a method you can carry out yourself, with no analytics services, just a notebook.

If you have not yet chosen a location, start with the article "Where to place a self-service coffee shop": it covers the types of location. Here we cover how to check a specific spot with numbers before installation.

Step 1. Separate the flow from potential buyers

Foot traffic past a point and traffic that turns into a purchase are different quantities. The gap between them decides the fate of the location.

You need to count not everyone passing but those who have a coffee scenario. In the morning on the way to work, there is a scenario. On the way to the exit of a shopping mall with two bags, almost none. A parent taking a child to a class has one: they have forty minutes of waiting.

A practical trick: during the count, note along with the number of people their direction of movement, speed and hands. A person walking fast and purposefully buys less often than one walking slowly or standing and waiting.

Step 2. Count by hand on the right grid

Counts are done with a stopwatch and a table, in specific time windows. The minimum scheme that gives usable data:

Twelve counts of 15 minutes is three hours of work. Nobody gives you better data for that price.

An important detail: the count is done at the place where the machine will stand, not at the entrance to the building. The difference between "past the building" and "past a specific corner of the lobby" can be threefold.

Step 3. Apply a realistic conversion

This is where people most often lie to themselves. Benchmarks by site type to put into the calculation:

Site typeReasonable conversion of traffic into a cupFuel station, basic coffee zoneabout 10%Fuel station with deliberate work on the zone20-25%Grocery retail1-3%Business centre (of the number of employees, not passers-by)15-30% of employees per dayTransport hubfractions of a percent of total flow

The figures for fuel stations are an estimate by Alexander Kuzmin, head of RusHolts (RBC Saint Petersburg, March 2026): a normal coffee zone sells coffee to 10% of those who have refuelled, and with serious work on the zone, 20-25%.

In a business centre you need to count not passers-by but people working in the building. One employee buys 0.15-0.3 cups per working day, depending on price and alternatives nearby.

Step 4. Check the surroundings

A list of questions, each of which changes the forecast:

Step 5. Check seasonality and schedule

A point that lives only on weekdays loses a third of its potential. At universities, traffic falls almost to zero in July and August, and in June and January it is cut by exam sessions: the model must allow for two empty months and one or two weak ones. In office buildings the dip falls on the New Year holidays and the holiday season.

Opening hours of the site are a separate question. Round-the-clock access gives a noticeable gain at fuel stations, railway stations and hospitals. In a business centre that closes at nine in the evening, the machines' advantage of 24-hour availability does not work.

Step 6. Pull it into a forecast and compare with the threshold

Take the average traffic across the measured windows, apply the conversion, adjust for seasonality and you get the plan in cups per day. Compare it with the break-even threshold of your model.

A sensible rule is to build in a margin. If the forecast gives exactly the threshold, the point is loss-making: in the first weeks a point usually sells less than after a couple of months, and some of the assumptions always turn out optimistic. A margin of one and a half times the threshold is the minimum comfortable one.

When to walk away at once

Walking away from a bad location saves the investment in a machine and half a year of your attention, which would have gone on trying to rescue it.

Conclusion

Three hours of counting costs less than half a year of trying to rescue a weak point. Count not passers-by but those who have a coffee scenario, and assume a realistic conversion for the site type.

And keep a margin over the break-even threshold. A forecast that exactly matches the threshold means a loss in practice.

Frequently asked questions

How long does measuring traffic take

About three hours of pure time: three days of four fifteen-minute windows. That is enough to decide on a specific location.

What conversion should I use in the calculation

For fuel stations, 10% of fuel traffic with a basic zone; for grocery retail, 1-3%; for a business centre, 0.15-0.3 cups per employee per day. More optimistic figures need justification.

Can I do without manual counts

Sometimes the site itself provides data: visitor counters in shopping malls, the number of refuellings per day at a fuel station, the number of tenants and employees in a business centre. It is worth requesting them, but checking on the spot is still useful.

What to do if the forecast is close to the break-even threshold

Walk away or revisit the rental terms. A point with no margin over the threshold goes into the red at the first deviation from plan.

To calculate the threshold for a specific location: payback calculator. On formats: coffee machines for business, DrinkX station.