ATM Route Location: Assess Demand and Test Sites

What to take away

ATM route location selection is a forecasting problem dressed up as a sales problem. You are trying to predict how many people will stand at a machine and take money out, using evidence gathered before the machine exists. This guide sets out what predicts that number, what only appears to, and how to test a site cheaply rather than argue about it.

What you are actually predicting

Not visitors. Not sales. Withdrawals.

A venue can be full of people who all have cards, all use them, and never touch the machine. Another can be half as busy and produce twice the count because the trade is cash-friendly, the tickets are small, and the nearest alternative is a walk away in the rain.

So the question at every site is narrower than "is this place busy". It is: how many people here, in a month, will want cash badly enough to pay for it. The rest of this guide is ways of getting at that.

The demand signature of a good site

Look for these together rather than individually. Any one of them alone is weak evidence.

Against those, treat the following as noise: the size of the building, the number of cars in the lot, the owner's enthusiasm, and any comparison to a machine somewhere else that you have not seen the records for.

Position inside the venue

This is where placements are quietly won and lost.

Two constraints override preference. The machine must be reachable and usable, and the 2010 ADA Standards for Accessible Design include provisions addressing automatic teller machines. Read the 2010 ADA Standards and ask a qualified adviser how they apply to the specific site; do not assume the venue has considered it.

The machine also has to be serviceable: a position you cannot open safely with cash in your hands is one you will resent.

What the venue tells you, and what you can check

Owners are generous with numbers and vague about which month. Ask narrow questions instead of broad ones.

Ask how often people ask where the nearest machine is, and who they ask. Ask what the venue pays on card transactions under a small amount; they usually know because it annoys them.

Ask which nights differ and why, and ask if a machine was here before, and if so what happened to it. That is the single most useful question on the list.

Then check what you can yourself. Visit at the hours that matter rather than at a convenient hour. Look at where the nearest working machines actually are and whether they are available at those hours. Sit in the venue for a while on a night the owner called typical.

Promising a host anything

At some point the venue will ask what they will earn. This is the moment operators talk themselves into trouble.

Give a method, not a promise: their share depends on withdrawals, so show them how it will be calculated. Say plainly that neither of you knows the count yet.

If you make an earnings claim, you own it. Claims to a business must be truthful and supportable, and the FTC's advertising guidance for small businesses sets out how that expectation works.

A host who was promised a figure and got less will terminate, and will tell the next operator who calls.

Test rather than argue

Sites that look marginal should be tested, not debated. Install with the smallest sensible cash load, visit more often than you plan to long term, and give it a fixed window of at least ninety days.

Record the withdrawal count per week, the value dispensed, the faults, and how long each visit took including travel. Set the cash load from that machine's own dispense rate as you go, using the arithmetic in the inventory and sourcing guide rather than a route average.

Ninety days is the floor because a month is too short to separate a slow patch from a slow site, and because a new machine gets a curiosity bump in the first fortnight that flatters it.

What each new site adds to the rest of the business

A placement is never just a placement. It adds a stop to every round, a permanent amount of cash parked away from your account, a counterparty who can end the arrangement, and a set of records you now have to keep.

The banking side is worth understanding before you accumulate sites rather than after. Banks assess an independent ATM operator through risk-based due diligence covering the whole relationship, not machine by machine.

The FinCEN statement to banks about independent ATM operators describes what that review is meant to weigh. An operator whose records are tidy and whose sites are documented is a much easier account to keep.

Deciding to remove a machine

Write the exit condition before the install. Something like: below a stated withdrawal count for three consecutive months, with no downtime explanation and no change at the venue, the machine comes out.

Write it early because you will not want to do it later. There will be a relationship, a host who is nice to you, and a sunk cost.

A machine in a weak site is not neutral: it consumes cash you could have parked somewhere better, it earns a visit on every round, and it holds hardware that a waiting site needs. The overview of how a route makes money sets out why the idle capital is the part that hurts.

Removal is also a contract event, so the agreement should already say who removes the machine, who repairs the wall, and how much notice each side owes.

Common questions

Is a busy retail location always better than a quiet bar?

No, and the assumption costs people money. Retail customers overwhelmingly have card acceptance in front of them. A venue where cash does a job it cannot do elsewhere often produces more withdrawals from far fewer people.

Should I take a site because a competitor wants it?

Only if the evidence supports it on its own. Sites bid up in a hurry tend to carry the commission terms that were agreed in that hurry.

How many sites should I look at before choosing one?

Enough that you have declined some. An operator who has never walked away from a venue has no standard, and the first weak placement is usually the one that teaches them.

What if a good site says no?

Ask what would change the answer and when to come back. Venues change hands, machines get removed, and the operator who left a good impression is the one who gets the call.