Building Next Year’s ATM and ITM Budget: What Banks and Credit Unions Often Underestimate

The ATM and ITM line in next year’s budget almost always costs more than the number on the planning spreadsheet suggests. It looks like a fixed, predictable expense, a maintenance contract plus cash and a little cushion, which is exactly why it gets the least scrutiny of any line a finance team reviews. In practice it is one of the least predictable categories an institution carries, because the costs that actually move the total are the ones that do not show up until they hit. 

For CFOs, COOs, and operations leaders building the plan for next year, the ATM and ITM budget is worth a closer look than it usually gets. The components below are the ones most institutions either leave out entirely or estimate too low, and understanding them is the first step toward a number you can actually hold to. 

The capital you are quietly carrying 

A single ATM costs roughly $40,000 to purchase, and an interactive teller machine (ITM) runs another $20,000 to $30,000 on top of that. Those machines are not a one-time purchase; they sit on a replacement cycle of seven to ten years, which means that for any fleet of meaningful size, you are always somewhere on the curve toward the next capital outlay. If a portion of your fleet is approaching end of life, next year’s budget is not just maintenance, it is the beginning of a replacement conversation that ties up capital you could be deploying into lending, branches, or other growth initiatives. 

This is the cost most plans treat as a future problem rather than a present one. Spreading the replacement across the planning horizon, rather than absorbing it as a lump sum in the year a machine fails, is the difference between a managed budget and a fire drill. 

The operating costs that look small until you add them up 

Beyond the hardware, the recurring spend is where budgets drift. Annual maintenance averages around $6,000 per machine. Compliance and regulatory upgrades, which arrive on someone else’s schedule rather than yours, routinely run $10,000 or more when a Windows version reaches end of support or PCI requirements change. On top of those, every fleet absorbs a category of spend that is genuinely difficult to forecast: vandalism, Windows upgrades, PCI compliance work, and out-of-scope service charges that fall outside whatever your current contract happens to cover. 

These are not edge cases. They are the normal cost of owning ATMs, and the reason a line item that was supposed to be flat ends the year over plan. A useful exercise during budget season is to pull the last two or three years of actual ATM spend and compare it against what was budgeted. The gap between the two is usually sitting in this category. 

The line item that never appears on the budget at all 

There is one cost that no spreadsheet captures, because it is paid in staff time rather than dollars. Keeping ATMs and ITMs running in-house means managing a stack of separate relationships: equipment, software, service, armored courier, remote management, processor, network connectivity, and more. Each of those is a different vendor, a different contract, and a different point of failure, and when something breaks, the work of managing vendors, escalating issues, and managing compliance deadlines lands on people whose time is worth far more spent elsewhere. 

That is the real cost of the multi-vendor model. It is not only the invoices; it is the operations staff, the IT staff, and sometimes the executives who spend hours coordinating between providers when accountability is spread across all of them and owned by none. Those are hours your team is not spending on lending, advice, and growing relationships instead of managing service providers. 

Downtime is a cost even when it is not an invoice 

An ATM that is down is not generating fee income, and an ITM that is down is sending members and customers back into a teller line you may have been working to shrink. Downtime rarely appears as a budget line, yet it shows up in member satisfaction scores, in branch staffing pressure, and in the slow erosion of the self-service experience you invested in. When you evaluate next year’s plan, uptime is worth treating as a financial metric, not just an operational one. 

If branch transformation is on the roadmap 

For institutions weighing an ITM rollout or a broader branch transformation, the budget conversation expands in both directions. The investment is real: an ITM costs more than an ATM, and core integration, the project that connects the machine to the system running all of your accounts, is roughly a $200,000 effort that takes six to nine months. Core integration is what turns a standard ATM into a true ITM, unlocking remote video teller service, check cashing, denomination selection, and access to all of a member’s or customer’s accounts on a self-service basis. 

The return is equally real and worth modeling alongside the cost. Institutions that make this shift can move up to 90% of teller transactions to self-service, which is what makes a smaller branch footprint possible. The proof points from NextBranch customers are concrete: Hawaii State Federal Credit Union saw 96% of surveyed members across all generational segments say they would use the ITMs again, recorded a five-point increase in net promoter score after implementation, and cut new-branch-employee training from two weeks to four days. The budget question is not simply “what does an ITM cost,” it is “what does it cost relative to the teller capacity, real estate, and member experience it changes.” 

Questions worth asking before the budget is finalized 

A few questions tend to surface the costs that hide in this category: 

• How much did we actually spend on ATM and ITM operations over the last two years, and how did that compare to what we budgeted? 

• How many separate vendors are involved in keeping our fleet running, and how much staff time goes into coordinating them? 

• Where is each machine on its replacement cycle, and what capital does that imply over the next three years? 

• What did we spend last year on items that were technically out of scope, and how predictable is that spend going forward? 

• If branch transformation is a goal, have we modeled the ITM and core integration investment against the staffing and footprint savings it enables? 

Making the number predictable: outsourcing and managed services 

The reason ATM outsourcing has become a serious budget consideration for community banks and credit unions is that it converts most of the categories above into a single, predictable figure. Under an outsourced model, a provider owns the equipment and bundles operations into one fixed monthly fee per machine, which removes the upfront capital outlay and replaces the unpredictable spend with a known number you can build the plan around. 

A well-structured ATM outsourcing arrangement includes vendor management and expert managed services, so the work of coordinating equipment, software, service, armored courier, processor, and connectivity moves off your team and onto the provider. It maintains the latest technology, so you are not budgeting separately for the next compliance-driven upgrade. And it carries no out-of-scope service charges, which is precisely the category that pushes in-house budgets over plan. 

The comparison below frames the difference in budget terms.