Psychographic Payment Data: A True Differentiator for BNPL
Merchants trying to cater to today's consumers' WIN (Want It Now) mentality have their work cut out for them. The average merchant website sees four out of every five visitors abandoning their cart prior to any purchase. Of those four, only one is likely to come back at a later time to complete their purchase, while the other three go on to purchase their item or continue to browse somewhere else. The blame may go to unintegrated, uncompelling visuals and customer experience on the website or lack of effective drip campaigns to ensure attracting already decided buyers. Yet, while these factors certainly contribute, the real deterrent to transaction closure is the customer's suboptimal—if not broken-- Point of Sale experience (PoSX). All the journey mapping that's been done, the countless hours of innovation, the ingenious marketing and lead generation, and even the best-in-class recommendation engine on the website,… would ultimately fall at Payments' sword.
As the lockdown of 2020 cemented customers' hyper-personalised expectations of their shopping experience, BNPL (Buy Now, Pay Later) crept back in fashion. Yes, good old BNPL, which has had many reincarnations as instalment billing in developed markets, and as micro-lending in emerging markets for decades, is now helping retailers remediate that oft-forgotten PoSX. This time, with the playing field being digital, customer reach is instantaneously global, which means retailers enjoy all the help they get in wrapping their arms around customer conversion in-store and online. At the same time, they focus mainly on attracting visitors to their websites. The competition being steep, merchants are even willing to pay more in transaction fees, as conversion rates and the purchase amounts that go with each transaction more than compensate for such high fees. Banks, who used to offer financing to buyers directly, are just as happy partnering with BNPL systems and technology providers, as it also means larger market capture for them without them having to lift a finger to upgrade their usually behemoth-sized IT systems to afford the dynamicity that today's e-commerce requires. Yet, a few things remain broken:
Pavlov and his dogs are back, unnoticed
Let's consider the customer's buying journey, for example. They go on the website, spend some time looking at items, customise their items (if allowed), put them in their cart, and – depending on how much they want their customised item (reward) – they are willing to endure particular idiosyncrasies of the payment experience (stimuli). Yes, indeed, it's the classical conditioning we all learned but have forgotten from Psych 101 in college. Simply put, consumers have trained themselves to expect annoying identity and income verification and credit decisioning during their buying experience, which they've learned are disjoint and choppy from their shopping experience. Furthermore, they must optimise their response to figure out how to get their reward faster with the least hassle (variable schedule of reward). It is all this "figuring out" and learning of the BNPL's systems' idiosyncratic credit decisioning that indicate what value the buyer places on their desired item to keep them hanging on instead of abandoning their cart. That "value" is what's captured in psychographics.
Psychographics? What's that?
The consumer's sentiment, propensities, attitudes, relationships, and decisioning criteria make up their psychographics about a certain subject, be it an item, a situation, or other people. Psychographics can quantify the value the consumer places on the things, experiences and relationships that surround them. For the BNPLs who are experiencing the standard 80% cart abandonment rate mentioned earlier, learning how to incorporate these psychographic levers in their credit decisioning could be their survival differentiator. For the retailer who is choosing which BNPL to partner with, scrutinising how they address cart abandonment via the lens of how they calculate what the buyer values should separate players who have thought things more thoroughly than those who are purely relying on digital scale. As part of "deciphering" the buyer, psychographics are an essential contributor to "compelling" the buyer to hang on in the hopes to see messaging, offers and terms that are keeled in their favour, as depicted in Figure 1 below. The retailer and their BNPL partner must crack this code together to remove the choppiness of the customer experience across their systems.
Figure 1. The role of psychographics in value-driven customer experience.
Credit decisioning on an event-driven intelligent infrastructure
While it is significant in and of itself to figure out how to capture buyers' psychographics in their value attributions and decision making, today's competitive landscape also requires BNPLs to do credit decisioning on the fly. This means pushing credit decisioning more and more towards the user interface (UI) and not anywhere else. Consequently, this also means faster draw of information about the buyer from anywhere in the BNPL-retailer data ecosystem onto the UI. In the past (and actually even today), judgmental rules behind a few qualifying questions on the UI gave the impression that dynamic decisioning was happening. This does not scale to today's standards amongst WIN mentality consumers and the kind of hyper-personalisation they demand from retailers and lenders. What does scale is a data infrastructure that maps information in contextual chunks to allow faster retrieval. This is why knowledge graphs, as some would say, have crossed the Rubicon. Knowledge graphs are made smart by the AI-compatible schema that can be installed in them. Ontologies and different types of deep learning via neural net can provide self-learning schemas to graph databases. Even decentralised data schemas like blockchain are best rendered on graphs. When you have every piece of information equally accessible on knowledge graphs, then triggers (events) and causalities are easily observable and mapped.
Analytics and scoring-wise, there are methodologies that are also more knowledge graph-friendly. Bayesian neural net, federated learning via genetic algorithms, and even swarm learning (where no synthesis of disparate data sets is needed) can accelerate real-time credit decisioning when applied to a knowledge graph data infrastructure.
Conclusion
Knowledge graphed data, where deep learning-based semantic layers provide context to payment experience, not only facilitates access to information needed for live credit decisions, it can also posit machine learning in the database itself, thus making self-learning credit scoring systems achievable. Additionally, when psychographics are leveraged in real-time, dynamic profiling of buyers allows hyper-personalisation and real-time credit decisioning right on the UI. Net-net, the BNPLs who can pull off this kind of innovation are the ones who will prevail, with clear advantages and value propositions to highlight to their future retail partners.
***
About the Author: Maria Singson, PhD, VP & General Manager, Data Science, Mastech InfoTrellis, brings 20+ years of experience as a technical and business leader of analytics-driven Centers of Excellence (CoE) for clients aiming to be strategic and culture-conscious in their digital transformation.
In her past roles as Leader of Innovation Analytics at Dun & Bradstreet, CEO of twoMS.co, and Chief Science Officer at Genpact, Maria established CoEs that helped companies realise value in their data and reimagine their risk decisioning and sales and marketing analytics.
She is also the founder of multiple startups in analytics and retail, where she leverages AI/ML to create economic opportunities for disadvantaged women and benefit disabled children.
Maria teaches AI strategy and metrics for organisations to gauge and forecast their AI adoption at Rutgers Business School for Executive Education in her spare time. Her human performance-centric approach to AI readiness and transformation is rooted in her PhD in Cognitive Science (UC Irvine) and bachelor's degree in Psychology (University of Southern California).
