A "primary" banking relationship sounds like ownership, but it’s really not. McKinsey defines a primary relationship as an institution holding 40% or more of an SMB’s deposits — which means even the bank that holds that title is, by definition, watching most of the money from the outside. The rest of the picture is already fragmented: the top 25 US banks hold 84% of primary SMB relationships, while community and regional institutions saw their share of those relationships fall from 28% in 2018 to 16% in 2022. And the deposits are not sitting still — 41% of SMBs told McKinsey they were likely to switch primary banks within 12 months, with another 32% considering it. The question is not whether an institution runs open banking or traditional banking. It is whether it can see the movement before it happens.
However, there is a distinction between the two when it comes to how financial institutions work. Under traditional banking models, a relationship manager finds out an SMB is in trouble when the SMB tells them. Under open banking models, the banker sees the trouble coming weeks earlier. Same customer, same products, same pricing sheet — different outcome, because the underlying data model is different.
This article compares the two models and the way a financial institution experiences the difference: what each one lets you see, what each one lets you do, and what the same SMB looks like when it runs through each.
The difference in 50 words
Traditional banking uses only the data held inside the bank’s own accounts, refreshed in batches and read after the fact. Open banking, or, more accurately, open data, uses consented data from across the customer’s full financial stack, delivered via secure APIs in near real time — turning the bank’s view of an SMB from a partial record into a working picture.
That is the definitional answer. If you need the fundamentals first, start with What Is Open Banking? — this article assumes them and goes straight to the comparison.
Plenty of comparisons stop at "APIs versus statements." That is true and not very useful. The differences that show up in a bank’s P&L are these five.
|
Dimension |
Traditional model |
Open data model |
|
Data scope |
Only accounts held at the bank |
Every account and business app the SMB consents to share |
|
Timing |
Batch, retrospective, often next-day |
Near real time via secure APIs |
|
Consent |
Implicit at account opening; static |
Explicit, granular, revocable by the SMB |
|
Unit of value |
The transaction record |
The insight derived across sources |
|
Banker posture |
Reactive — respond when the SMB calls |
Proactive — act when the signal fires |
A traditional view shows the bank what happened inside its own walls. That is typically only 10–20% of an average SMB’s wallet — a sample no credit team would accept from a bureau, used every day to make relationship and risk calls. Open data widens the aperture to the accounts, accounting ledgers, payment processors, and payroll systems the business actually runs on.
Traditional data tells the bank what an SMB did last month. That is fine for reporting, but it does not help with intervention. A cash shortfall trending three weeks out is only actionable if the bank sees it three weeks out. Latency is not a technical detail here; it is the difference between a retention call and an exit interview.
Under the traditional model, the bank has the data because it holds the account. Under open data, the SMB grants access deliberately and can withdraw it. That sounds like a constraint, and it is — but it is also the strongest trust signal a financial institution can earn. An SMB who has actively connected six external accounts to their FI’s dashboard has told you something no NPS survey will.
Aggregated data is not the product. Normalizing feeds from hundreds of sources into something comparable is the engineering half of the job — covered in Data Aggregation in Banking — and the insight layer on top of it is the half customers see. Traditional banking sells records. Open banking sells answers.
This is the one that executives feel. In the traditional model, the relationship manager is a service function that responds to inbound. In the open data model, the same person becomes a coverage function that acts on signals. The headcount does not change. However, what the headcount is capable of does. And it can improve the bottom line.
The abstract comparison lands better as a sequence of events. Here is how three routine moments in the life of an SMB customer play out under each model.
|
|
Traditional |
Open banking |
|
Trigger |
SMB submits an application |
Deposit and receivables data show a 30-day shortfall forming |
|
Bank’s view |
Statements, tax returns, a credit pull |
Live revenue, expenses, and payment flows across providers |
|
Timeline |
Two weeks to a decision |
Pre-qualified offer already sitting in the dashboard |
|
Likely outcome |
SMB has already taken an online lender’s offer |
Bank books the loan it would otherwise never have seen |
The traditional path is not slower because the credit team is slow. It is slower because the process cannot start until the SMB starts it, and by then the SMB is shopping.
An SMB quietly opens an operating account elsewhere and begins moving deposits. Under the traditional model, the FI notices when the balance drops far enough to trip a threshold report — typically after the relationship has already moved. Under the open data model, the shift in inbound payment flows is visible while the customer is still deciding, and the outreach is a conversation rather than a save attempt. This is not a fringe scenario. Around 15% of SMBs changed banks in a recent 12-month period and 35% said they were open to it, which means a material share of the book is in motion at any given time — invisible under a traditional data model until the balance has already gone.
A retail SMB’s card volume triples over a quarter. Traditionally, that business lands in a segment-based campaign and receives a merchant services offer along with 4,000 other accounts, most of whom do not need one. With consented commerce data, the offer goes to the handful of businesses whose numbers justify it, in the week the change happened. The economics of that shift — and the risks that come with it — are covered in Open Banking Benefits for Banks. Roughly 70% of SMBs say they want end-to-end cashflow management from their primary bank, and while fewer than half currently get commercial enablement tools such as payroll and accounting there, more than half say they would prefer to get them from a single provider. The demand exists; the data model is what determines whether the bank can act on it.
Lead with the FI’s side of the ledger, because that is what funds the program:
The downstream SMB benefit is real too — a business that can see its full cash position makes better decisions.
An honest comparison has to say what open data does not replace. Core processing, the ledger of record, settlement, and the FI’s own compliance and controls all stay exactly where they are. Open banking is an additive data layer, not a core replacement, which is precisely why it can ship in months rather than years.
The framing that keeps projects on track
Open data is not a replacement for the core. It is a visibility and engagement layer that sits on top of it. FIs that treat it as a re-platforming project take a year longer and deliver less.
The second thing traditional banking still has: the relationship manager. No fintech competing for that SMB has a person who can pick up the phone with authority. Open banking data plus a relationship manager is a combination the challengers cannot assemble. Data alone is not the advantage — data routed to someone who can act on it is. The data backs the combination. 47% of SMBs name dedicated relationship manager support among their top criteria for choosing a primary bank, while 32% cite a better digital experience as a reason they would switch. Banks are the only competitor positioned to deliver both — but only if the RM is given something to act on.
Most institutions describe themselves as open-banking-enabled. Fewer are operating that way. Four questions separate the two:
Four yeses mean you are operation an open data model. Four negative responses means a traditional model with an API in front of it. Most FIs land somewhere in the middle, and the honest answer to where usually points directly at the next investment.
The sequence matters more than most teams expect, because each stage produces the inputs the next one needs.
|
Phase |
Focus |
Why this order |
|
0–6 months |
SMB-facing cashflow insight |
Earns the consent footprint and behavioral data everything else depends on |
|
6–12 months |
Banker-facing intelligence layer |
Turns data already flowing into RM workflows, cross-sell, and retention plays |
|
12–18 months |
Lending signals and pre-qualified offers |
Underwriting needs the historical depth the first year created |
|
18+ months |
Predictive triggers and proactive outreach |
Requires data volume, model maturity, and RM trust built earlier |
Lead with the use case the SMB wants. The use cases the bank wants compound from there. The full sequencing logic, the three failure patterns that derail rollouts, and the four KPI categories executive teams hold the program against are laid out in The Banker’s Playbook.
9Spokes gives banks the open banking operating model without the multi-year build: aggregation across 800+ financial and business sources, an insight layer that turns those feeds into cash position, forecasts, and lending signals, and banker-facing intelligence that puts the same data into relationship manager workflows. It is white-labeled and embedded in the FI’s existing digital channels, so the SMB experiences it as their bank — because it is.
Ready to move from comparison to execution?
The Banker’s Playbook covers what to build first, the four-step implementation framework, the pitfalls that stall rollouts, and the KPIs that keep the program funded into year two.
The pillar guide covers open banking end to end for banks serving SMBs — definitions, use cases, challenges, and the business case. Read Open Banking & Open Data for Banks: How to Win and Grow SMB Accounts →