SME Banking's real bottleneck isn't technology. It's the Org Chart
At a Qorus SME Banking roundtable in Dubai, bank executives agreed digitization solved the easy problem. What’s left is harder to automate.
SME banking has spent years treating digital transformation as the hard part: build the onboarding flow, automate the credit decision, and the rest follows. A roundtable discussion hosted by the Qorus SME Banking Community, bringing together bank executives and technology providers from the Middle East, Europe, Africa and South Asia, suggested something less comfortable. The technology problem is largely solved. What isn't solved is where, inside the bank, SME customers are supposed to live, and who gets to own the relationship once they do.
Digitization has already answered the retail-versus-corporate question
Like SMEs themselves — long described, half-jokingly, as banking's "missing middle" — SME banking has had its own back-and-forth between retail and corporate over where it structurally belongs. Raj Tatavarty, Head of Business Banking Products and Digital at Mashreq, argued that this ambiguity has effectively been resolved, not by strategy debates but by economics. "Digitization decisively put the SME segment into the retail sector," he said, because it collapsed the cost of acquiring and serving a customer who, in most markets, is a sole proprietor or freelancer rather than a layered corporate entity.
Özkan Erener, CEO of VeriPark, offered evidence of how far that compression has gone: a German corporate card provider cut onboarding from two weeks to fifteen minutes by digitizing it, and a large Turkish retail bank pushed its cost of acquiring and servicing a micro-business customer below $25, growing from 20,000 merchant relationships to 800,000. None of this needed new economics — only applying what retail banking had already solved for individuals.
The complex cases still need a human
The more complex end of the segment resists that logic. Bharat Subramanian, Head of Value Proposition, Portfolio, and Customer Experience for Business Banking at Emirates NBD, noted that mature AI-driven forecasting models can bring substantial value to SME cash flow management. Given the inherent volatility of SME cash flows compared with individual salary accounts, enhanced forecasting capabilities can help businesses and banks anticipate potential liquidity challenges and respond with timely, tailored solutions. Other participants noted that the first real trigger for specialist help is often not transaction size but friction with correspondent banks over documentation — a blocked remittance that needs someone who understands the business to explain the invoice behind it. Once a company starts dealing in complex FX/hedging or multi-partner ownership structures, Tatavarty acknowledged, “it cannot be that I just copy-paste as-is from the retail playbook into SME banking.” That split sounds simple in principle. In practice, it has produced an organizational problem banks have struggled with for years.
The "Super RM" keeps failing, and not for the reasons banks think
Tatavarty described trying, twice, to build a single relationship manager serving both a business owner's company and personal accounts — the kind of "universal RM" model banks periodically announce. Both times it looked good as a headline and failed in practice: splitting KPIs between a retail RM and a business RM caused open conflict over whose target a transfer should count toward, and merging the roles fixed nothing, since whoever filled the job carried the bias of wherever their career started. "If it's an SME RM, you would always focus on the SME side of things, where the individual gets relegated," he said. Nobody could agree whether the role should report to the retail head, the SME head, or the priority banking head — a fragmentation other participants kept raising from a different angle, as the unresolved question of how business banking should consolidate within the bank so customers get one coherent experience, not several disconnected ones. "It was a superb headline statement to say we'd started the Super RM — even universal RM, whatever you want to call it, it's basically a common RM," Tatavarty said. "It didn't work. Only at the headlines it stopped."
Daniele Coda, Chief Commercial Officer at Rakbank, described the same failure from the customer's side. Treating a business owner's personal and company finances as separate P&Ls, he argued, produces outcomes that make no commercial sense: a customer who has just used financing to bring a personal balance to zero can simultaneously be charged a higher rate on the business side, because the two sides of the bank don't share documentation or even talk to each other. "Banks need to think about the total relationship with a customer, both business and personal. That's what builds a real understanding of their preferences and opens the door to wallet expansion," he said. Zubair Ahmed, Qorus's Managing Director, framed the underlying cause bluntly: banks built segments to serve customers better, then "became slaves to those segments." He pointed to one bank in Qatar experimenting with a single owner accountable for a customer's value across every segment — still, in his words, "working in its infancy."
Why AI is being proposed as the answer — and why that's not obviously right
It was at this point that one participant raised a more provocative idea: what if the inconsistency wasn't a staffing problem to be fixed with better incentives, but a job better suited to AI than to any human role? "If you were building a native AI bank from scratch today, these discussions wouldn't even be happening," they argued, pointing to Meta's newly launched personal AI assistant as a preview of an adviser that could "know the customer better than anyone, without emotion, without incentives." Tatavarty's response was measured rather than dismissive — "far away at this point in time," but plausible eventually — and he reframed what AI actually fixes: not the decision about where a customer sits, but the consistency of applying that decision everywhere beneath it, since AI "doesn't have subjectivity" and "will exactly replicate down to each level the way it was envisaged strategically at the top level."
Daniele Coda offered a more structured version of the same caution, framing AI's role around four principles: proactive (using live transaction data to anticipate a business's needs and feed insight back to it), relevant (guiding RMs toward the next best action for each client), scalable (extending personalization beyond the small group who currently get it), and human-led (AI surfaces insights; judgment and accountability stay with people). Pamela Chan King Yeun, Head of Business Banking at Mauritius Commercial Bank, pushed back on the framing that started the exchange: a bank first has to decide, at the strategic level, what it wants to achieve and where SMEs sit within that. AI is an enabler of that decision, she argued, not a substitute for making it.
Not every constraint on SME growth sits inside a bank's org chart, either. Ahmad Abu Eideh, CEO of Jordan Ahli Bank, pointed out that in Jordan, SMEs account for roughly 98% of the economy — at that scale, what determines whether the segment grows is not only how any one bank structures itself, but how banks and governments work together to extend meaningful, coordinated support. For a segment that large, no single institution's reorganization substitutes for that wider cooperation.
That distinction is worth sitting with, because it reorganizes the whole discussion. At bottom, nearly everything debated in this session — whether SMEs sit in retail or corporate, whether one RM or two serves them, whether the next fix is a new title or a new algorithm — was a question of where and how the customer would be best served. Zubair Ahmed made this point explicitly near the end of the session: the structural choices the group spent most of its time on are themselves what decide whether that customer experience holds together. The strategic choice this leaves banks with is not digital versus human, or human versus AI. It is deciding, explicitly and at the top, what a good SME relationship looks like from the customer's side, and then using digital tools and AI for the part they are actually good at: executing that decision faithfully and cheaply at scale, instead of hoping a new org chart or a new RM title will do the job that only a clear strategic choice, and in some markets a wider partnership, can do.
Council Participants
.jpg)
Participating Institutions
SME Banking community
With Qorus memberships, you gain access to exclusive innovation best practices and tailored matchmaking opportunities with executives who share your challenges.