Guides
How to expand an ATM route without overreaching on cash, time and control
Expansion hits three walls in order: cash, time and control. What each wall costs, what fixes it, and how buying a route differs from building one.
What to take away
- Growth stops at three walls in a fixed ordercash, then time, then control. Each one needs a different fix, not longer hours.
- One placement typically parks $2,000 to $5,000 in cash at a quiet site, and $8,000 to $20,000 where traffic is heavy.
- One person can hold a typical range of 8 to 12 stops a day, or roughly 40 to 60 machines on a weekly visit cycle.
- Buying a route fixes density and time but doubles the float, so verify processor records before you sign.
The three walls, in the order they arrive
The three walls
The signal
- Cash
- Sites found, funding missing
- Time
- Unplanned trips rise while planned visits fall
- Control
- Service dates and differences live in your head
What fixes it
- Cash
- Matched-term funding, a cash partner, or slower growth
- Time
- Clustering, longer visit intervals, then help
- Control
- Written records and a scheduled review
The order is not a coincidence. Every placement parks money, so cash comes first. Time follows next.
[figure 1]
Three walls and their fixes
Cash
- Feels like
- Can't fund sites
- Actually is
- Load plus unsettled cash
- Fixed by
- Funding or cash partner
Time
- Feels like
- Days run long
- Actually is
- Stops exceed one person
- Fixed by
- Clustering or help
Control
- Feels like
- Things missed
- Actually is
- Route exceeds memory
- Fixed by
- Written process and records
What the cash wall actually costs
A live machine holds cash that never returns to you: the cassette load plus everything dispensed and awaiting settlement. A quiet site ties up a typical range of $2,000 to $5,000. A busy convenience store, $8,000 to $20,000.
Cassette size and settlement timing drive that range, and the ATM route inventory guide covers both.
A route can be profitable on every machine and still unable to fund the next one. Working capital is the constraint.
Three responses exist. Borrow on a facility matched to the placement's expected life. Use a cash-in-transit partner who supplies the load and takes a lower share. Or grow more slowly from the route's own earnings.
Banks review the relationship as the float grows, and that review turns on your actual risk. That is what the FinCEN statement on independent ATM owners and operators asks of banks.
When the time wall arrives
The trigger is unplanned trips climbing for two months while planned visits fall. That is the wall, not a scheduling problem.
In a dense district, one person can hold a typical range of 8 to 12 stops a day. On a weekly cycle that is roughly 40 to 60 machines.
Fix it in cost order:
- Cluster stops by geography so no site sits alone on a run.
- Stretch visit intervals where load size and traffic allow.
- Move low-value sites to a monthly cycle with monitoring.
- Add help last, because a second person brings payroll and classification questions.
[figure 2]
Time wall fixes in cost order
- Cluster new placements
- Lengthen visit intervals
- Drop low-return dedicated trips
- Add help last
Hiring per stop usually points to contractor status, and the IRS test turns on control over the work, not the label.
A full week of stops, including fault triage and parts, is worked through in the operations and workflow guide.
The control wall and the records that end it
Name the last service date and every open difference for ten machines, without checking your system. When that fails, the wall is here.
It is not a memory failure but arithmetic. The route's state has to live in records, and someone has to read them on a schedule.
Three records do most of the work: a service log, an exception list for differences over a set amount, and a monthly reconciliation.
Compliance needs revisiting here, because an arrangement that looked fine at three machines can look thin at thirty.
Three kinds of expansion, often confused
More machines in the same district. The cheapest growth, and the one to exhaust first. Travel per machine falls and your local reputation compounds.
A new district or state. Your referral network does not travel with you. Crossing a state line can trigger money transmission licensing, and states continue to adopt the Money Transmission Modernization Act, which changes bond and net worth tests. Ask the state regulator first.
A new venue category. Traffic, hours and security all differ. It also cuts concentration risk, which is a reason to do it on purpose.
Buying a route instead of building one
Acquisition solves density and time. It worsens cash, because you inherit every float.
Checklist before signing:
Buying a route
- Pull transaction counts from processor records, never from the seller's spreadsheet.
- Check each contract for transfer, rate and removal terms, using this placement agreement review.
- Confirm every machine meets your processor's requirements.
- Walk each site for position, lighting and accessibility.
Transfer and removal terms deserve as much attention as the rates. The 2010 ADA Standards include provisions for ATMs, so review the 2010 ADA Standards with advice for the sites you take on.
What has to scale with the machine count
Five items change with the machine count, and each has a trigger.
What has to scale
| Item | Trigger to act | One action |
|---|---|---|
| Spares | First emergency trip takes over an hour | Kit the van for the three most common faults |
| Insurance | Cash in transit nears your policy limit | Re-rate on the new float, not the machine count |
| Security | Any run becomes predictable | Vary service days and route order monthly |
| Host contact | A host reports a fault before monitoring does | Put the site on a quarterly call schedule |
| Pricing | Float grows faster than stop count | Renegotiate the commission or the fee |
Float and downtime both grow faster than stop count once a route spreads out, so track them separately.
Deciding to stop
A route one person services well, with hosts who refer, is a legitimate place to stay. Where costs sit against revenue per machine decides whether that still holds, which is what the break-even analysis for an ATM route sets out.
The test for the next placement is whether it improves the route or only enlarges it. Ground on finance and planning for either answer sits in the SBA business guide.
Common questions
How many machines can one person service?
A typical range is 40 to 60 machines on a weekly visit cycle in a dense district. A monthly cycle can carry 150 to 200. Measure your own stops per day.
How much cash does one machine tie up?
A typical range of $2,000 to $5,000 at a quiet site, and $8,000 to $20,000 where traffic is heavy.
Should I buy an existing route or build one?
Buy when density and time are the constraint and you can fund the inherited float. Build when cash is tight.
What is the first sign of growing too fast?
Hosts report faults before your monitoring does. Attention per machine has fallen below what the site needs.
