Savannah Price, CEO and Founder of Serene, discusses how financial data, particularly open banking data, can be useful for detecting and acting on early signs of customer vulnerability.
Behavioural analytics tools, with rich data, can now help financial institutions identify stress and life events through subtle shifts in spending behaviour.
The discussion covers the evolving ecosystem around vulnerability, customer trust, probabilistic modelling, integration challenges, and the future potential of vulnerability data across multiple sectors.
Find out more about Serene-> Here.
Key Take Aways
- Financial transaction data can reveal early indicators of mental health and stress-related vulnerability before crisis points emerge.
- Serene identifies vulnerability using erratic changes in spending frequency, volume, and category – particularly ahead of known life events.
- There is a feedback loop between financial stress and mental health, where one exacerbates the other.
- Specific financial behaviour patterns correlate with types of vulnerability, such as increasing grocery and utility bills for carers or decreasing income during life transitions.
- Micro-signals – minor shifts in transactional patterns—can act as early warning signs of potential issues.
- Open banking enables proactive vulnerability detection and offers significant potential for real-time support and tailored interventions.
- Serene’s platform integrates into lenders’ and banks’ existing journeys at origination, income and expenditure, and collections stages.
- Customers prefer to share open banking data with impartial third parties rather than directly with lenders or banks.
- Serene provides confidence intervals alongside predictions to account for data completeness and model certainty.
- The platform can assist in creating customised forbearance plans based on customer-specific need and risk trajectory.
- Underutilisation of support services is a known challenge; Serene helps direct appropriate customers to existing resources.
- Vulnerability modelling is expanding beyond financial services into utilities, telco, and government – especially in identifying scam susceptibility.
Innovation
- Use of open banking data to derive early, predictive “digital fingerprints” of customer stress and life changes.
- Behavioural analytics at micro-signal level to anticipate future events based on patterns of spend.
- Confidence-based modelling with adjustable risk thresholds tailored to firm-level operational preferences.
- Ecosystem model where vulnerability detection is decoupled from providers, enhancing customer trust and engagement.
- Probabilistic modelling driving engagement strategies rather than deterministic outcomes.
Key Statistics
- 85% of consumers are willing to provide open banking consent throughout the loan term if it leads to access to new products or timely support.
- 42% of consumers would prefer to share data with an impartial third party.
- Only 20% would share data directly with their lender or provider.
Key Discussion Points
- The origin of Serene was a personal need to support family members with mental health challenges using financial data as early indicators.
- Distinct transaction behaviours can signal specific types of vulnerability, such as mental health decline or caregiving responsibilities.
- Early detection based on changes in financial behaviour offers an opportunity to act before distress escalates.
- Traditional transactional flags (e.g., gambling spend) are lagging indicators—micro-patterns offer earlier intervention points.
- Open banking data must be comprehensive, with at least access to a primary current account to drive insights.
- Customer journeys need to respect autonomy—data-driven support should remain opt-in to avoid perceptions of surveillance.
- Personalisation is critical—vulnerability characteristics alone are insufficient without context and segmentation.
- Predictive modelling can inform “next best actions” based on severity, imminence, and available support within the firm.
- Companies face information overload; converting insights into practical, actionable flags is key.
- There is a growing need to incorporate cross-industry data sources (e.g., telephony, sentiment) to strengthen signal detection.
- Probabilistic thinking is required to replace binary decision-making within vulnerability and customer support strategies.
- Serene aspires to be a core infrastructure player in vulnerability analytics – complementing credit and affordability scoring.
#Serene
RO-AR insider newsletter
Receive notifications of new RO-AR content notifications: Also subscribe here - unsubscribe anytime
