The bank channel is still an imperative for processors. How it is being served is changing.
Note: We write this from a UK and European vantage point, where partnerships like Lloyds–Stripe and Starling–Adyen are already live, but the same dynamics are now playing out in the US market.
Market context
Much of the commentary on US small business acquiring over the last five years has framed the bank channel as the loser. The working assumption is that bank-initiated acquiring is in structural decline, although the numbers tell a more interesting story. Fintech, ISV and processor-direct volumes represent around 31% of the US merchant acquiring market today, and forecasts from UBS and McKinsey put that share at 45% by 2029, which still leaves 55% of the market flowing through banks as a distribution channel. Given the acquiring revenue opportunity was $30bn in 2024, the bank channel remains structurally large and durable.
What is genuinely under pressure is the experience layer around the infrastructure that has served the channel for the last twenty years. The strategic question is therefore not whether the bank channel matters, but which processors are positioned to serve it well over the next decade.
The challenge for banks
Part of the reason the software-led narrative has taken hold is that very compelling SaaS businesses have been built within industry verticals. SMBs are heterogeneous, and a restaurant, a salon and a plumber have almost nothing in common operationally, which is how Toast, Mindbody and others have built the share that they have by tailoring deeply to one segment and embedding payments inside it. Banks, by contrast, serve SMBs across many merchant categories, and so tend to focus primarily on their shared financial needs. In terms of payment processing, their requirements from a processor are are homogeneous with authorisation, clearing, settlement, and reporting requirements broadly the same across institutions.
Where banks do require customisation is in the layer that sits between the processor's rails and the merchant: onboarding flows, branch and relationship-manager tooling, brand standards, the architecture of the online banking channel, and the way merchant services slot into the wider SMB relationship. It is the part of the value chain that sat outside the traditional processing mandate.
The consequences show up in penetration data. Banks typically have low single digit penetration of business banking customers - ie the percentage of their business banked customers who also take merchant acquiring from the bank. This can rise to 20-30% where the acquiring proposition is fully integrated to the business banking estate. Typically this is because onboarding journeys that should take minutes take days and branch teams have no tooling to sell at the point of need. Acquiring can end up feeling like an add-on rather than part of the banking relationship.
This is the gap processors are now moving in to. Stripe and Adyen, for example, have spent a decade building the experience and orchestration layer modern SMBs expect. They already power a large share of the ISV channel, which is more likely to be digitally native and experience oriented. And now they are turning their attention and product set to the bank channel, as illustrated by Stripe's partnership with Lloyds and Adyen's with Starling.
What modern partnership looks like
The next generation of bank and processor partnerships is starting to look meaningfully different from the outsourced model that defined the last twenty years.
Adyen’s partnership with Starling Bank brings a next-generation processor into the bank channel's mobile app and Stripe’s partnership with Lloyds is similar. Payment capabilities such as pay by link, tap to pay, invoicing and online checkout are being introduced around the edges of the bank's existing acquiring proposition, wrapped in a bank-branded experience layer. Lloyds has not moved away from their existing joint venture, triggering the back-book attrition that comes with a full re-platforming. The existing processor continues to run the core acquiring proposition with a new micro proposition runing on Stripe. The result is that the merchant sees one unified experience inside the bank channel and the bank is able to introduce capabilities it could not previously easily offer.
 × |  × | |
|---|---|---|
| Announced | June 2026 | May 2026 |
| Features announced | Lloyds Accept, built on Stripe Connect. Tap to Pay, payment links, terminal devices, sign up in minutes. | In-app acceptance. Tap to Pay, next-day settlement, payment links via invoicing later in 2026. |
| Structure | Processor behind the bank, embedded in the business account. Bank owns the brand and front end. | Same model, with the full lifecycle handled inside the Starling app and no redirect to Adyen. |
| Bank type | Largest UK incumbent, 1m+ business customers. | Digital challenger, SMB focused from day one, 450k+ business customers. |
| Legacy estate | Cardnet JV with Fiserv sits alongside. Stripe is additive and segmented. | None. A pure greenfield extension. |
| Processor angle | First UK traditional bank deal, a flagship bank-channel entry. | A route down market from enterprise, via banking partners. |
| Strategic driver | Defensive modernisation against Tide, Allica and the neobanks. | Offensive deepening of an already software-led proposition. |
| What success depends on | Back book skews toward slightly larger SMBs, so clean segmentation between Lloyds Accept and Cardnet is critical to route the right merchants to the right proposition. | Already has a large existing micro and sole-trader base, so it comes down to driving attach rates. |
Bank operating models, and how processors need to respond
The bank channel divides into to a few distinct operating models. Traditionally, regardless of operating model the acquiring experiences have not been truly white labelled and integrated within the broader business bank estate. Larger banks have been investing to bridge this experience gap in recent years.
Model 1 - the referral model: In the referral model, the bank acts as a distribution channel for the processor in return for a revenue share. The bank does not own the proposition. The proposition is usually co-branded at best, and typically operates outside of the bank digital estate through a simple link out. Roughly 2,000 US institutions sit in this model.
Model 2 - the hybrid / agent model: In the hybrid or agent model, the bank takes greater ownership of sales, servicing and pricing while the processor continues to provide the underlying platform and, in most cases, the acquiring licence.
Model 3 - the strategic model: In the full solution or strategic partner model, the bank owns the proposition and the merchant relationship and outsources the underlying processing. The bank tends to want a white labelled and integrated experience, and historically has had to invest significantly to build this on top of the processing capability. Truist already operates this model, TD and Huntington have moved towards it in recent years, and PNC and Wells Fargo look to be moving in the same direction as they exit their joint ventures.
Frontend experience capabilities now sit alongside considerations like functionality and price for banks.
The five questions that follow are the ones every processor should consider when speaking to banks.
1. What is the bank actually asking for when it says it wants to own the customer?
Bank teams often say they wish to "own the customer". But in practice that can mean different things. The diagnostic to run early in the cycle is whether the bank wants to own the MID or they are talking about wanting control over the merchant experience, which doesn't necesarily meaning owning the contract with the merchant.
2. What is missing today to run processor-behind-the-bank cleanly? For the banks that genuinely do want their own MIDs, serving them well requires the processor to support acquiring, settlement that flows through the bank rather than through the processor and in some cases on a gross basis, reconciliation and reporting that land in the bank’s own ledger rather than in the processor’s dashboard, and credential and network token portability so that the bank is not quietly locked in by the very thing it is buying, each of which is a meaningful piece of work in its own right.
3. Are the economics of acting as a processor attractive? This is the question that gets asked least often and matters most, because serving the bank channel as the infrastructure layer behind a bank-owned proposition is a different business from serving it as the acquirer of record and benefiting from the bank's distribution. Processors need to have modelled the trade-off between the additional transaction volume that comes with sitting behind more bank programmes and the margin they are willing to concede to win that position, with a clear view of what happens to the spread book as the mix shifts from acquirer-led to processor-led.
4. Does the go-to-market motion match how banks actually buy? Bank procurement cycles are long, and internal stakeholder management routinely runs to eighteen months or more, which means processors moving into the bank channel for the first time, or materially expanding the way they serve it, need an enterprise sales motion that can manage that cycle, with the discipline to qualify hard at the front of the funnel.
5. How is channel conflict handled? Every bank will ask this, because a bank handing its merchant portfolio onto a processor’s rails wants to understand what stops that processor from approaching the same merchants directly, particularly where the processor also runs its own SMB acquiring business. A bank needs to see a contractual answer, with explicit non-compete provisions covering the bank’s book and data ringfencing arrangements that the bank can audit on its own terms.
What this means for the next five years
For processors, the prize is the 55% of the US acquiring market that still flows through the bank channel. Processors that do not adapt will spend the next few years watching their referral partners lose merchants, while other processors are already moving into bank distribution.
The bank channel is being rebuilt, and that the processors who participate in that rebuild are the ones who will own the next phase of its economics.
Sources:
UBS, Payments, Processors and FinTech Industry Primer (The Question 6.0), May 2026, including Flagship Advisory Partners US SMB channel distribution data (August 2024) and McKinsey 2025 Merchant Acquiring Survey, sourced from “Decoding ISV maturity: A global playbook for payments growth” (January 2026). Pollinate analysis of US bank-channel programmes and operating models. PNC, Wells Fargo and Truist joint venture history per public reporting.
