The fintech industry has a naming problem. Its products are described with the same handful of words - money movement, treasury, ledger, accounts, real time - until a cash forecast starts to sound like a payment rail and an API platform starts to sound like a bank. Put Modern Treasury, Trovata, Highnote, Increase and Column on one procurement slide and the logos appear to be rivals. Follow a dollar through the system, though, and the picture changes. They sit at different layers, take responsibility for different failures and often belong in the same architecture.
A simple first pass is useful. Modern Treasury is centered on payment operations: initiating, tracking, ledgering and reconciling money movement. Trovata is centered on the treasury view: gathering bank and ERP data, explaining cash positions and forecasting liquidity. Highnote is centered on embedded financial products, particularly card issuing, with program management and a unified ledger. Increase offers programmable accounts, payments and cards through banking APIs and direct bank partnerships. Column N.A. is the regulated institution underneath: a nationally chartered bank with its own core, balance sheet and connections to payment networks.
Start with the dollar, not the dashboard
Imagine a software marketplace that collects money from buyers, pays thousands of sellers, offers its best suppliers a virtual card and wants to know whether it can cover payroll thirteen weeks from now. One transaction may pass through a bank account, an ACH or real-time payment rail, an operational ledger, a reconciliation workflow, a card program and a forecasting model. Every vendor can truthfully say it helps manage the money. None necessarily owns the whole journey.
That map is a center-of-gravity test, not a prohibition against overlap. Modern Treasury now offers a payment service provider with accounts, compliance and fiat and stablecoin rails. Trovata has expanded beyond forecasting toward treasury management and payments. Highnote describes one platform for issuing, acquiring, credit and money movement. Increase supports card issuing as well as bank accounts and every major U.S. payment rail. Column can support cards, lending and payment origination directly. The edges are messy because these companies kept building.
The middle remains distinct. Trovata's forecasting system consumes normalized balances, transactions, ERP information and assumptions to produce a view of future liquidity. Its own help material calls forecasting an alternative to storing assumptions in Excel and stresses variance analysis. That is a decision aid. Modern Treasury's payment order, by contrast, is an instruction to move money. Its job includes validations, webhooks, returns and references that help an operator understand whether the instruction succeeded. One helps answer “Can we afford it?” The other helps make “Pay it” reliable.
Cards create their own operating world
Highnote is easiest to understand through the lifecycle of a card. A company must onboard an account holder, create a physical or virtual credential, decide whether an authorization should be approved, impose spend and velocity controls, fund the transaction, handle settlement and disputes, and maintain an accurate ledger. Highnote packages those jobs with program management and developer tools. It has since added acquiring and credit, but issuing remains a useful doorway into the product.
Increase can also issue cards, approve authorizations in real time and expose Visa network data. The distinction is therefore not “cards versus no cards.” It is the desired product boundary. A team seeking relatively bare financial primitives, accounts and many rails may lean toward Increase's API model. A team designing a managed embedded-finance program may value Highnote's program layer. The right diligence asks who supplies the sponsor relationship, who makes compliance decisions, which controls are configurable, where balances are authoritative and how exceptions appear at 2 a.m.
See and predict cash
Begin with Trovata when aggregation, position reporting, scenarios and forecast variance are the binding problem.
Move and reconcile
Begin with Modern Treasury when approvals, payment states, bank connectivity and reconciliation consume the workday.
Launch a card program
Evaluate Highnote and Increase around the exact issuing model, controls, rails and program responsibilities you need.
Work at the bank layer
Consider Column when direct partnership with a chartered, regulated bank is central to the product architecture.
A charter changes the answer
Column is the outlier because Column N.A. is a bank. It is nationally chartered, regulated by the Office of the Comptroller of the Currency and a member of the FDIC. The company built its own core and direct connections to systems including the Federal Reserve and The Clearing House. It can hold deposits, originate payments and act as the institution behind an account or card program. Its API may resemble those of infrastructure software companies, but its regulatory position does not.
Increase is explicit that it is not a bank. It works through direct-to-bank partnerships and gives customers APIs for accounts, ACH, wires, checks, RTP, FedNow and cards. That can place it quite close to Column in an architecture diagram, and a particular product may make either one a plausible starting point. Yet the contracting chain, regulatory oversight and division of compliance work differ. “How many layers sit between us and the charter?” is not a philosophical question. It determines escalation paths, product control and what happens when a risk team says no.
Ledgers deserve the same skepticism. Nearly every platform here refers to a ledger, but the word does not guarantee identical scope. A card ledger can track authorizations and settlement. An operational ledger can map customer balances to payment events. A bank core records the bank's books. A treasury application may normalize reported transactions without becoming the legal record of funds. During evaluation, replace “Do you have a ledger?” with “Which balance does this ledger make authoritative, and against what is it reconciled?”
Build the shortlist around failure
Feature checklists reward breadth. Failure maps reward clarity. Take the ugliest transaction your business can produce: a returned ACH debit after a card-funded payout, perhaps, or a wire that arrives without useful remittance data. Trace the event through every account, webhook, ledger entry, approval and report. Then ask where an operator sees it, who can repair it, whether the repair is auditable and how quickly the cash forecast reflects the change.
Procurement should test the calm days too. Ask each vendor to model an ordinary month-end close using your account structure and naming conventions. Give it a sample of the files, API events and approval policies that exist today. A platform can have an elegant sandbox and still create manual work if its transaction model does not match the way finance closes the books. The revealing metric is not merely time to first API call. It is time until an accountant, support agent and engineer can look at the same event and agree on what happened.
This exercise also exposes when multiple products make sense. A company might hold customer funds with a bank partner, use payment APIs to access rails, run orchestration and reconciliation in another system, operate cards through a specialized program platform and aggregate the resulting accounts in Trovata. Composition adds integrations and vendors, but it can give each team a tool shaped for its actual job. Consolidation reduces handoffs, but may force a product to be merely adequate outside its core.
Pricing should be translated into the same flow. One platform may charge for accounts, another for transfers, another for cards or active users, and another for implementation. A low unit price can disappear beneath exception handling, duplicate ledgers or a reconciliation team that grows with volume. Conversely, a broad platform can be wasteful if the company only needs one narrow primitive. Estimate cost per successful, reconciled outcome, including internal labor. It is a less tidy number than an API fee, and a more useful one.
The decision is not a beauty contest. It is an allocation of operational and regulatory responsibility. Ask what your own engineers must build, what finance must reconcile, what compliance must approve, what the vendor guarantees and which bank ultimately stands behind the account. The company names may still sound like variations on the same promise. The work underneath them does not.
Are Modern Treasury and Trovata competitors?
They overlap around cash visibility and treasury workflows, but their centers differ. Modern Treasury focuses on moving, tracking and reconciling payments; Trovata focuses on cash data, positioning, reporting and forecasting.
What is the difference between Highnote and Increase?
Both support card programs. Highnote centers on embedded-finance program infrastructure across issuing, acquiring and credit, while Increase exposes a broad set of banking and payment primitives through APIs and bank partnerships.
Is Column a banking-as-a-service platform?
Column calls itself an infrastructure bank. Column N.A. is a nationally chartered, OCC-regulated bank and Member FDIC, rather than a middleware-only BaaS provider.
Can a company combine these products?
Yes. The layers can complement one another. The cost is more integration and vendor management, so every handoff and system of record should be explicit.
What should buyers compare first?
Start with where funds are held, which rails are required, who owns compliance, which ledger is authoritative, what must be forecast and which team handles failures.