Why SME funding still feels difficult — and how technology is changing it
A business owner can be told that lenders are keen to support UK SMEs and still spend weeks completing forms, supplying the same information more than once and waiting for a clear decision. That apparent contradiction is not simply a customer-service problem. It reflects the economics of SME lending and the technology and processes through which many established lenders still deliver it.
The direct answer
SME lending is difficult to make simple because many assessment and servicing costs arise per borrower, not per pound lent. Established lenders also have complex systems and controls that are expensive to replace. Newer entrants can start with modern data and workflow, making the process faster and more convenient when technology is combined with experienced credit judgement.
The important point is not that traditional banks lack capital, expertise or good people. Many have all three. The difficulty is that smaller business facilities can be costly to originate, assess and monitor, while long-established infrastructure can make even sensible process improvements expensive and risky to implement.
Newer lenders have an opportunity because they can start with modern data, connected systems and a customer journey designed around today's SME. The strongest of them will not use technology to avoid judgement. They will use it to remove the administration that prevents skilled people from exercising judgement quickly.
The hidden economics behind SME lending
A Bank of England analysis published in June 2026 estimated that SME lending produced the lowest average return on equity of the main lending categories it examined for large UK banks. The Bank was careful to describe the work as an estimate based on public disclosures and supervisory data, rather than a definitive profitability measure for every bank.
The more useful finding was why returns appeared lower. The analysis concluded that expected impairments and operating costs were the main drivers of the gap between SME and larger corporate lending. Capital requirements played only a relatively small role.
This matters because many costs are incurred per borrower rather than per pound lent. A lender still needs to identify the customer, complete financial-crime checks, understand affordability, review security, document the facility, monitor performance and service the relationship. A £250,000 facility does not require one fortieth of the work needed for a £10 million facility.
The result is uncomfortable but logical: the smaller and more individual the request, the more difficult it can be for a large institution to spend substantial human time understanding it economically.
Legacy systems are not just old software
When people talk about legacy banking technology, it can sound as though a lender simply needs to replace an outdated screen with a better website. The real problem is far more complicated.
A lending journey may rely on separate systems for customer onboarding, identity checks, credit assessment, documentation, security registration, payments, account management, arrears and regulatory reporting. Data may pass between them through older interfaces, spreadsheets, manual uploads or rekeying. Different departments can hold different versions of the same information.
These systems are often supporting large numbers of existing customers and legally binding facilities. Replacing one component can affect historic records, payment processes, reporting obligations, fraud controls, operational resilience and the way employees carry out their roles. A change that looks simple from the outside can require coordinated work across technology, credit, compliance, legal, operations and risk.
The Bank of England's July 2026 Financial Stability Report referred, in a broader resilience context, to banks' dependence on legacy technology, complex change-management processes and large numbers of external applications and software. That does not mean every established lender has the same problem. It does explain why large financial institutions cannot safely rebuild critical processes as though they were launching a new app.
Legacy processes can be as restrictive as legacy technology
Technology is only part of the issue. Processes accumulate over time as well.
A manual review, approval step or document may have been introduced after a loss, a regulatory finding, a fraud event or an operational failure. Years later, the original reason may be less visible, but the control remains embedded in policy. Removing it requires evidence that the risk is being managed another way.
This can create multiple handoffs between relationship management, credit, operations, legal and compliance. Each handoff introduces delay, creates another opportunity for information to be interpreted differently and can make it difficult for the borrower to know who owns the next action.
The natural response for a large lender is standardisation. Standard products, scorecards, minimum information requirements and centralised decisioning make volume manageable. They also mean that a viable business with an unusual model, a recent change or a non-standard requirement can struggle to fit the process.
What newer funders can do differently
A newer lender does not need to unwind decades of infrastructure before improving the experience. It can design the lending journey around modern sources of information and a single workflow from the outset.
- Companies House data can populate basic company information and reduce input errors.
- Open Banking can provide consented transaction data rather than relying only on downloaded statements.
- Accounting-platform integrations can provide current management information and debtor data.
- Digital identity, fraud and sanctions checks can run earlier in the journey.
- Documents can be requested, stored and reviewed through one secure portal.
- Workflow tools can show who owns each action and what remains outstanding.
- Electronic signatures and automated document generation can reduce avoidable completion delays.
- Ongoing data can support earlier-warning monitoring after the facility is live.
The advantage is not merely speed. Connected data can reduce repeated questions, improve consistency and provide the underwriter with a more current view of the business. It is the same argument set out in speed, transparency and certainty.
Open Banking shows how quickly the data infrastructure is maturing
Open Banking Limited reported 16.5 million user connections by December 2025, up 36 per cent over the year, alongside 24 billion API calls. The connection figure is not the same as 16.5 million unique people or businesses, because connections are counted by bank brand and are not fully deduplicated. Even with that caveat, the scale shows that consent-based financial data sharing is no longer a niche experiment.
The Government has also said that real-time current-account information can improve credit-risk assessment and may help some small businesses that previously struggled to obtain finance. The value is straightforward: a lender can see how cash actually moves through the business, subject to the customer's consent and the quality of the data, rather than relying entirely on static historic documents. We look at this in more detail in open banking, AI and SME funding.
Better data should strengthen judgement, not replace it
There is a danger that discussion of lending technology becomes a false choice between slow human underwriting and instant automated approval. Good SME lending needs both data and judgement.
Automation is well suited to collecting documents, checking completeness, retrieving public information, identifying inconsistencies, calculating ratios and highlighting areas that need attention. It can help a straightforward case move quickly and give an experienced underwriter more time to investigate a complicated one.
It is less suited to understanding every nuance of a management team, a temporary trading event, a customer concentration or a business model that has limited historic comparables. Models can also reproduce poor assumptions, miss context and create false confidence if their limitations are not understood.
The aim should not be to automate every decision. It should be to automate the parts of lending that should never have required so much human effort in the first place.
Convenience is now part of the funding product
SMEs increasingly judge a funding provider on more than price and headline availability. They also notice how much time the process consumes, whether questions are relevant, whether progress is visible and whether they can speak to someone who understands the decision.
A convenient process should not mean a careless one. It should mean that the borrower explains the requirement once, gives informed consent for appropriate data access, sees what is outstanding and receives a clear answer. A decline delivered quickly and honestly can be more useful than an uncertain process that drifts for weeks. The reasoning behind each request is worth understanding — see why lenders ask the questions they ask.
Newer does not automatically mean better
A modern interface can hide a narrow credit policy. Some providers move faster because they lend against a tightly defined data set, require strong security or charge more for convenience. A rapid automated offer is not automatically the most suitable or sustainable facility.
Established banks can offer cheaper capital, broad product capability, experienced credit teams and long-term relationship support. Newer entrants can offer speed, focus and flexibility. The strongest outcome is likely to come from combining modern infrastructure with disciplined underwriting and people who understand how SMEs actually trade.
What should an SME expect from a modern funding process?
- A clear explanation of what information is required and why.
- No unnecessary repetition of data already supplied or retrieved with consent.
- A realistic indication of eligibility before a full application consumes significant time.
- Visibility over progress, outstanding actions and the person responsible.
- A decision that considers the purpose and repayment source, not only a generic score.
- Access to a human when the information does not tell the whole story.
- Clear pricing, conditions, security requirements and next steps.
The Juno Funding view
The persistent friction in SME finance is not evidence that established lenders do not care about smaller businesses. It is evidence that the economics and infrastructure of SME lending are difficult.
Technology can change those economics by reducing data collection, rekeying, duplicated checks and manual monitoring. It can make a smaller facility more economical to deliver and a complex business easier to understand. But technology alone is not a lending strategy.
The next generation of SME funders will not win simply because their systems are newer. They will win by using technology to make funding quicker, simpler and more convenient while retaining the judgement, challenge and discipline needed to lend responsibly. That is the direction we set out in the future of business funding.
In summary
- Bank of England analysis suggests SME lending generates lower average returns for large UK banks than the other lending categories it examined.
- Higher expected impairments and per-borrower operating costs are more important drivers than capital requirements.
- Legacy systems and accumulated processes make change costly, risky and slow for established institutions.
- Newer funders can use connected data, automation and digital workflows to remove avoidable friction.
- The best model combines modern technology with experienced human credit judgement.
FAQ
Why is SME borrowing often slower than personal borrowing?
A consumer loan can often be assessed against highly standardised data and a tightly defined product. SMEs vary far more in sector, structure, ownership, cash cycle and purpose, so the lender may need to understand both the business and the proposed use of funds.
Do legacy systems cause every delay?
No. Delays can also arise from incomplete information, complex ownership, security requirements, fraud checks, credit concerns or unclear responsibility between teams. Legacy systems can make these issues harder to resolve because data and tasks are spread across multiple platforms.
How can Open Banking improve an SME funding application?
With the customer's consent, Open Banking can provide current transaction information directly from the bank account. This can reduce reliance on downloaded statements, help verify cash flow and allow the lender to assess recent trading more quickly. It does not guarantee approval.
Does automated underwriting mean there is no human credit decision?
Not necessarily. Many good lending models automate data collection, checks and straightforward calculations while escalating complex or unusual cases to experienced people. The balance depends on the lender and product.
Are newer funders always faster or more flexible?
No. Some have narrow policies or higher pricing, and a digital journey can still produce a rigid decision. The relevant question is whether the provider combines a suitable product, reliable data, fair pricing and access to judgement when the case requires it.
Sources and further reading
These links were checked on 3 August 2026. Date-sensitive statements should be rechecked before relying on them.
- Bank of England — What drives differences in commercial banks' product level returns? (19 June 2026)
- Bank of England — Financial Stability Report, July 2026
- Open Banking Limited — Open Banking in 2025 (29 January 2026)
- Department for Business and Trade / GOV.UK — Small business access to finance
This article is for general information only and does not constitute financial, legal, tax or investment advice. Funding availability and terms depend on the lender, product, business circumstances and satisfactory assessment.