Almost every guide to treasury management software is published by a company that sells treasury management software. They are not dishonest. They are just answering a question you have not asked yet, which is which one, rather than the one you should ask first, which is whether.
We have no horse in that race. EX FI is a payments provider, not a treasury management system vendor, and we do not sell one. So here is the version nobody with a licence to sell will write: what these systems cost, when the spend is genuinely justified, and the three cheaper things worth exhausting first.
What treasury management software actually does
A treasury management system sits between your banks and your books. It aggregates balances across every bank and entity, forecasts liquidity, executes and reconciles payments, and manages exposures like FX and interest rate risk, replacing the portal logins and spreadsheets most finance teams still run on.
The cleanest way to hold the distinction: an ERP records what happened to your money and closes the books; a TMS manages what your cash is doing right now. They are complementary, not alternatives, and a TMS that cannot read your ERP cleanly is worth much less than the brochure suggests.
What it costs, with real numbers
Vendors quote on request, which makes this the hardest part of the evaluation to research and the easiest place to be surprised. The published bands look like this.
Lightweight cash management tools aimed at smaller businesses start around 10,000 to 20,000 USD a year, while mid-market and enterprise platforms typically run 20,000 to over 100,000 annually, with implementation, bank connectivity setup and ERP integration priced separately as one-time costs.
Broken down by generation, legacy platforms run roughly 130,000 to 350,000 a year, modern visibility-first platforms 50,000 to 120,000, and newer agentic platforms 35,000 to 90,000. Vendor-specific benchmarking fills in the rest: for one major platform, an upper mid-market deployment of one to three entities and five to twenty banks lands at 80,000 to 230,000 USD in total annual subscription, with year-one implementation services of 80,000 to 180,000 over 14 to 22 weeks, bank connectivity at 500 to 3,000 per connection, and renewal uplift of 5 to 8 percent.
Read those two sentences together and the important pattern appears: implementation frequently costs more than the first year of licence. Any quote that does not include a three-year total cost of ownership, with integrations, training and expansion, is not a quote. It is an opening position.
The threshold test
Vendors publish adoption triggers, and the list is a fair one. A TMS starts to make sense at three or more legal entities, five or more banking relationships, a finance team of five or more people, multi-currency operations, and governance pressure from a sponsor, auditor or an approaching listing.
Read that as a set of conditions under which the spend becomes defensible, not as proof that you need one. The honest version of the same list has a middle rung. The usual progression is Excel, then the treasury module inside the ERP you already own, then a cloud-native TMS connected to both the ERP and the banks, and plenty of groups jump straight from the first to the third because nobody told them the second existed.
If you hit three or four triggers and the ERP module genuinely cannot cope, buy. If you hit one or two, keep reading.
Three cheaper fixes to exhaust first
The most useful thing published on this topic comes from a bank rather than a software company, and it is a warning: a TMS provides value in specific circumstances and will not solve problems rooted in organisational structure, responsibilities, processes or controls. Three of the fixable causes come up again and again.
Rationalise the accounts before you automate them
Where banking relationships are spread across too many accounts and too many banks, consolidating them can bring cash together without any new system at all. It is worth sitting with that, because it inverts the usual logic. You are being quoted six figures to aggregate a mess. Reducing the mess is cheaper than aggregating it, and it makes any system you do buy later smaller and faster to implement, since connectivity is priced per bank connection.
This is the one place EX FI is relevant, so we will be precise about it. A multi-currency payment account with dedicated IBANs across 16 or more currencies, reaching SWIFT, SEPA, FPS, BACS and CHAPS, with free internal transfers between your own entities and Xero reconciliation replaces several relationships with one. That reduces the number of connections a future TMS has to aggregate. It is not treasury management software, it does not forecast your liquidity, and we are not going to pretend otherwise.
While you are looking at the account layer, it is worth knowing whether the provider holding that cash is a bank or an authorised payment institution, because that changes what protects the balance you are about to centralise.
Use the ERP module you already pay for
ERP cash management and accounting modules are routinely untapped. They are weaker than a dedicated TMS on multi-bank connectivity and on FX and hedging tooling, which is exactly why the third rung exists. But if your problem is consolidated visibility across entities that all sit in one ERP, the module may close most of the gap for the cost of configuring it.
Fix the process, not the tooling
If the real issue is that responsibilities are unclear, that there is no transition plan for acquired entities and their accounts, or that controls are missing, software will make the same mistakes faster. Approval workflow is a good example: a system that executes payments without controlled approvals is not an improvement on the spreadsheet.
If you are buying, ask these five questions
- •What is the three-year total cost of ownership, including implementation, integrations, training and expansion?
- •Which of my banks are natively supported, and how long does adding a new one take?
- •Is the ERP connector native, or middleware that needs a statement of work?
- •Is pricing per bank, per entity, per user, or a platform fee, and what happens when each of those grows?
- •What is the renewal uplift, and how is bank connectivity repriced at renewal?
Get the answers in writing before you compare vendors, because the list price is rarely the number that matters. And whichever way the decision goes, the account layer underneath is worth tidying first: fewer relationships, cleaner rails, and a multi currency business account that does not need a system to make sense of it.
FAQ
How much does treasury management software cost? Lightweight cash management tools start around 10,000 to 20,000 USD a year. Mid-market and enterprise platforms typically run from 20,000 to over 100,000 annually, and large multi-entity deployments go well beyond that. Implementation is charged separately and often exceeds the first-year licence.
When should a company move off spreadsheets? The common triggers are three or more legal entities, five or more bank relationships, a finance team of five or more, multi-currency operations, and governance pressure from a sponsor or auditor. Hitting one trigger is not a reason to buy; hitting several usually is.
What is the difference between a TMS and an ERP treasury module? The ERP is the system of record and closes the books. The TMS is the operating layer that manages cash in the present. ERP modules are generally weaker on multi-bank connectivity and on FX and risk tooling, which is what pushes larger groups to a dedicated system.
Can a treasury management system fix bad processes? No. Where the underlying problem is unclear responsibilities, missing controls, an untapped ERP module, or cash scattered across too many accounts and banks, a TMS automates the problem rather than removing it.
Does EX FI sell treasury management software? No. EX FI provides multi-currency payment accounts, rails and intercompany transfers. Those can reduce how many banking relationships a treasury system has to aggregate, which lowers connectivity cost, but they do not forecast liquidity or replace a TMS.