Why Abel?

Abel is a lending technology platform, purpose-built to profitably extending credit to SMEs. It was born out of the challenges each of us had experienced in risk and technology teams at a multitude of established lenders and fintech start-ups, and the result of £20m worth of work we achieved together at FIBR, a high-growth challenger lender.

Here we spell out the specific challenges we faced and addressed as SME lenders, why the technology we were using fell short of our ambitions, and how we combined data science, domain expertise and high-quality engineering to make good on them.

SME lending is a unique and complex market

Lending to small- and medium-sized enterprises (‘SMEs’) is a hard business. Compared to consumer and corporate customers, SMEs present a completely different complexity profile that, if underestimated, can cause significant profitability problems, and make lending franchises unviable; just ask the graveyard of would-be lenders and fintech companies who’ve tried to enter and serve it.

Five key distinctions make the SME market particularly complex for credit institutions to serve:

Diversity

‍Globally, approximately 90% of all businesses are classed as SMEs – in the UK it’s 99.9% according to the FSB. As such, SME’s represent a huge breadth of diversity, ranging wildly in terms of size, industry, structure, profile and funding requirements.

Uncertainty

With significantly less information and typically shorter trading history’s, it can be hard to get an accurate read on a SME’s prior performance and future growth prospects. The advantage traditionally went to their incumbent bank provider and Relationship Manager, but these folks are now often stretched too thinly across geographies and sectors to keep abreast of the outlook ‘manually’.

Volatility

‍Information travels faster, change happens faster, scale happens faster – all contribute to larger, more frequent swings in economic activity and sentiment. Small businesses are generally more exposed to economic shocks than larger firms, but often exhibit different vulnerabilities and responses than larger firms (link).

Economics

A function of the diversity of the SME landscape, loan sizes can range anywhere from hundreds to millions of pounds. Often a lender needs to commit to a single mode of service - seeking to provide a highly automated 'low touch' service for smaller loan values and businesses, or accommodating high loan values with a high touch service, using manual underwriting and white-glove service to handle the additional risk and organisational complexity. In both cases, margins are challenged by price compression at the lower end and high cost-to-serve on the larger end of the financing spectrum. Larger banks often face the stiffer challenge of accommodating a variety of service models, and economics, across a broad range of products and customer segments, further frustrating efforts to improve and sustain healthy margins.

Confidence

‍Borrow confidence is low. Driven by both their and origination staff perception of likelihood of success on application. Bank of England reporting states 6 of 10 would be borrowers choose to inject personal funds instead, with half of those applying expecting rejection. This self-selection means that the process and perception of the process is fuelling adverse selection, which drives blunt conservatism, leading to a vicious circle.

Altogether, these factors create a complex environment to lend into, making it hard for incumbent lenders to create sustainably viable businesses and presenting significant barriers to entry for new lenders to innovate and improve the credit supply.

We’ve faced this very problem ourselves

We were the risk and technology team behind FIBR - a challenger lender launched just before the pandemic of 2020.

When we first built FIBR, we pieced together a stack of ‘leading’ components (logos that we all know and sometimes ‘love’), but we always fell short of delivering the results we expected - tripped up by one or more of the complexities of the SME credit market outlined above.

Speaking to colleagues in other lenders, we realised this was a pattern repeated across the industry. With returns uncertain, SME lending leaders were being asked to transplant strategies and technology applied in other parts of credit (e.g. corporate and consumer), and cobble them together to try and serve the SME market.

As a result, we were all falling foul of the same challenges, presenting the same symptoms:

• High exceptions requiring manual intervention by Underwriting and LoanOps teams
• Increasing impairments when implementing high or straight-through automation
• An inflexibility to change constrained us from anticipating and avoiding risks, or quickly capitalising on new opportunities as they arose.
• Overly conservative risk policies, in response to impairment and margin pressure, further hampered growth and profitability.

Together, we were falling short of our duty of care: presenting cumbersome experiences, higher pricing and lower supply, particularly to businesses who needed it most and when they needed it most.

We realised you need specific technology to serve the SME market

The technology we had got us so far, but we needed to make significant and continuous improvements across the entire credit lifecycle to combat the SME complexities we faced. We explored, experimented with, and made progress across the key questions that made the biggest difference to our bottom line, including:

• How different subsectors respond to the economy, impact our customers, and help us better handle volatility
• How we could better evaluate customer growth plans, and whether affordability would hold up
• What data we needed - and didn’t need - and how we could processes it for maximum efficiency and impact
• How we could improve borrower confidence by adjusting what offers we put out and what options we provided
• How could we get to know a small business as quickly as their bank RM, within 30 seconds and online?
• How to triage data and risk parameters to unlock quick decisions and be more targeted with underwriting intervention
• How we organised our platform and teamwork for instant and precise change amidst uncertainty and volatility

This investment of over £20m in technology, risk and commercial expertise culminated in a new, intelligent technology layer that helped squeeze more performance out of our stack.

We now make this technology available to you through Abel.

Introducing Abel

Abel provides the APIs and software to build, control and continuously improve your SME lending business. Our platform combines advanced risk and data science with SME domain expertise to maximise lender profitability and borrower experience.

Our APIs power the full credit lifecycle and seamlessly integrate with your ecosystem, delivering new insights and automation across your channels, partners, and technology estate.

Intelligence = data science + domain expertise

We believe that intelligence isn’t purely a technical or ‘more data’ phenomenon but that it requires deep domain expertise to truly commercialise. Our platform combines smart data science and engineering with deep risk and domain knowledge to power SME-specific features across the entire credit lifecycle: from origination and decisioning, through onboarding, servicing, and monitoring.

• Forward-looking affordability modelling, predicting future affordability levels to improve eligibility and consumer duty (see The Tragedy of SME Growth)
• Sub-sector scorecards, comprising of granular P(d) models and economic outlook
• Upfront soft offers, generated in 60 seconds on just 5 data points, with 85%+ firm offer conversion rates
• Multi-product credit triage, maximising eligibility, affordability, and lender income
• Criteria-driven underwriter referrals prioritising resource where it’s most profitable to place it.
• Ordering data within the decision to provide the quickest time to 'no' and 'yes'
• Transaction categorisation engines built specifically for SME accounts and transaction behaviours
• Early-warning indicators, driven by open banking data, economic outlook indices and our proprietary Behavioural PD score, support proactive loan management
• Fully parameterised risk criteria allowing for a clear and rapid change capability

Proven to perform

Abel has been battle-tested during the extreme lending conditions of the pandemic, delivering strong commercial growth amidst volatile economic conditions and in competition with the ‘free’ government funding.

• Approval rates improved by 25-30% whilst P(d) halved to 2.5%
• Acquisition costs held at <2%, 10% of applicants were zero-touch helping us sustain >10% margin
• Our lending rate grew over 600% in 12 months
• We could provide customers with multi-product soft offers in 60 seconds, with dispersal times in under 2 hours.
• Brokers would use our API and pricing engine to establish the ‘market rate’ before evaluating other funding sources, giving us a first-look advantage.

Under the hood

To enable these innovations, the Abel platform was architected to provide an intelligence layer on top of our existing technology components, helping us to surface insights from a wide range of data sources, including traditional lending data sets (bureau, financial statements), new financial sets (Open Banking, Subsector economic indices), operational data (including cost-to-serve and RWA calculations), and experimental datasets (TripAdvisor sentiment, management competency sentiment, Marketplace data).

We combine those inputs into a single data platform, providing a holistic view of a customer and our ability to service them profitably. Importantly, this single view allowed both technology and the business to collaborate from a shared view of risk and reward, accelerating decisions, aligning strategy, and improving cohesiveness in execution.

Our models are capable of triaging data – using a cost/benefit analysis to determine the marginal improvement a data point has on p(D) confidence to reduce unnecessary data costs and customer churn.

We continuously explore and evaluate new statistical and machine-learning methods to enhance models, whilst ensuring we take a measured approach that maximises explain-ability and alignment with risk and compliance controls.

Future ready

We designed Abel for both continuous improvement and step-change to help remain agile amidst persistent market volatility. This meant we could adjust risk parameters instantly, integrate new models or data sets in minutes, and enact wholesale changes to complex, multi-pathway journeys within single development sprints.

Today, Abel’s architecture allows for quick and effective change across new channels, partners, datasets, products and risk factors. Decision parameters can be adjusted instantly in no code, whilst model and dataset changes can be made within hours.

No more “rip-and-replace”

No company wants the cost and risk of ripping out mission critical systems and the significant workflow and re-training lag that comes with it. Instead of the traditional ‘rip-and-replace’ approach we integrate tightly with your existing stack and workflow, providing a layer of smart orchestration and intelligence to help you get more performance out of your stack.

Our API-native system can be integrated with almost any popular lending tooling and data sets, with pre-built configurations for popular services such as OnFido, Signicat, Mambu, Slack and AWS cloud management systems. Our headless platform enables you to build bespoke customer and employee workflows without technology constraint.

So, if you’re new lender looking agile go to market, or incumbent lender that wants to get more from your technology and channel estate - the Abel platform can support your ambition.

Find out more

Our mission is to build a deep and durable market for SME credit. We’re passionate about this goal and want to be generous with our experience and learning if it brings more credit to the market. Get in touch with our CEO, Paul or CRO, David to learn more about Abel and join our mission to bring credit to the small Businesses of the U.K.