From idea to first machine: the ATM route launch sequence

What to take away

Starting a route goes wrong in a predictable order. People buy machines, then look for venues, then discover the banking problem, then discover the float. Reversing that order costs nothing and removes most of the risk. Here is the sequence, with the test that must pass before you move on.

Stage 1: decide which business you are actually in

Owning machines and loading them yourself is one business. Placing machines and paying a cash-in-transit partner to load them is a second. Servicing machines a merchant already owns is a third. They have different capital needs, different risks and different daily rhythms.

Pick one to begin with. The overview of how the money moves sets out what each arrangement does to your working capital.

Typical time for this stage: one or two days.

Exit test: you can describe your model in three sentences without using the word "flexible".

Stage 2: answer the money questions before you shop

Typical startup ranges, before you buy anything:

FinCEN's guidance on when certain owner-operators fall inside the money services business definitions, and its separate statement asking banks to assess ATM operators individually, are the background reading. They are not the answer for your state.

Set up the entity and the books at the same time. The IRS material for new business owners covers structure, identification numbers and what to keep, and the SBA's guide to launching and running a business covers permits, insurance and financing.

Licensing is a state question. The regulator is usually the state banking department, the department of financial institutions, or the department of financial regulation. New York, Florida and Texas each license money transmitters.

Put the question in writing: does owning machines and loading them with my own cash, at venues I do not own, count as money transmission in this state?

A usable written answer reads like this: "The described activity does not fall within the state's definition of money transmission. The operator funds the machines from its own account and does not transmit funds for a third party. No license is required for the activity as described."

An answer vaguer than that is not an answer.

Typical time for this stage: four to eight weeks, with banking the slow part.

Exit test: a processor has said yes in writing, a bank has opened the account, and you have a dated written answer on licensing from someone qualified to give one.

Stage 3: find one venue, not ten

Look for places where cash is genuinely useful and card acceptance is genuinely awkward or expensive: late hours, small tickets, an entry fee, a cash-preferring trade, or a site where the nearest bank machine is a real walk away. Talk to the person who can sign, not the person behind the counter.

Do not negotiate the host commission on the first visit. Ask what the venue currently pays in card fees on small transactions and how often customers ask where the nearest machine is. The answers tell you what the placement is worth. The method for building that pipeline is in the material on winning and keeping hosts.

Host commission is a fee split, not a guess. A common arrangement is a flat monthly payment, typically $50 to $150, or a per-withdrawal share, typically $0.25 to $0.75 from a surcharge of about $2.50 to $3.50. Volume decides the number.

Typical time for this stage: two to six weeks to sign one venue.

Exit test: one venue wants a machine and you can explain, from what they told you, why you expect it to be used.

Stage 4: paperwork before installation

Confirm the placement position with the venue in the same conversation: where the machine sits, how the approach works, and whether anything about the position would obstruct a customer. Do not let the nearest wall socket decide it, and ask your adviser how accessibility obligations apply to that site.

Typical time for this stage: a week or two once the venue has agreed.

Exit test: both parties have signed, and neither of you had to say "we will sort that out later" about anything.

Stage 5: install, then measure for ninety days

Load the machine, and from the first visit onward record four things every time:

The licensing and compliance overview sets out the records that must also exist for reasons beyond your own curiosity.

Ninety days is the minimum honest window because venues have weekly and monthly rhythms, and the first two weeks are unrepresentative in both directions.

Exit test: three months of counts, and a trend that is flat or rising.

Stage 6: decide about machine two

This is a fresh decision. The first machine proved the venue. It did not prove your route.

Ask whether the second placement shares a trip with the first, what it does to total cash you must hold, and whether machine one's fault pattern changes your hardware purchase.

Sourcing decisions and spare-part planning are covered in the stock and equipment material. Strains that appear as the count climbs are set out in the guide to growing a route.

Typical time for this stage: a week of analysis after the ninety days.

Exit test: it raises stops per trip or transactions per dollar of cash held, or it is a hobby.

Common questions

How long does the whole sequence take?

Longer than the hardware suggests and shorter than the banking suggests. Stage 2 is usually the slow one, and it is the one worth being slow.

Should I buy an existing route instead?

It is a reasonable path, but verify the transaction counts yourself from processor records rather than from a spreadsheet the seller prepared, and read the placement agreements to see whether they transfer at all.

What if a bank turns me down?

Ask what specifically concerned them, fix that, and try institutions that already serve cash-intensive businesses. A refusal is a data point about that bank's appetite, not a verdict on the model.