< Articles & News

5 Ways Banks Use Big Data Analytics To Win Back Customer Confidence

September 28, 2016   |   3:23 PM
Big Data Analytics
Teradata Articles

by Dominic Vincent Ligot, Teradata

It’s a recurring theme. Log on or open your favourite newspaper and you’re hit with story after story of problems in the banking world. Fraud, the fallout from UK Brexit, a veritable roll call of financial crises and misdemeanours – factual confetti spiced with rumour and insinuation, colouring the public perception of what banks really do.


No wonder then, that despite the industry’s robust growth and the fact that banks have been an integral part of the fabric of society for hundreds of years, the public view the banking community with suspicion. Customers focus on issues like security breaches, lack of service expansion, and poor customer service, while the banking fraternity look to the heavens and downplay their concerns.

Winning hearts and minds

To win back customer confidence and maintain their place in the face of revolutionary digital disruption, individual banks (as well as the industry as a whole) need to take a long hard look at their traditional business models and operational practices. Some banks have already begun the digital transformation journey – adopting new technologies and tapping existing data resources to develop better products and services. Big Data and Analytics are the key but largely, their full potential still remains unrealised. Banks need to take some practical steps towards turning consumer-perception obstacles into data-driven business opportunities.

Payments data

Start with the most under-appreciated dataset. Payments reveal a great deal about each user – how much they’ve paid, what they paid for, who was paid, the banks involved, transaction time and location, and so on. In fact, a customer’s payment profile says much more about her, or him, than any social media metric or record. Payments data is highly accessible and can pinpoint lifestyles, detect which companies make up a supply chain, and plot spending trends by time or place. At the same time, although customer data is not as dynamic as payments data, in banking systems it can be attached to other profiles such as payments and credit history to enhance analytics and create successful “Next-Best-Offers”.

Fintech sensibilities

Should banks be worried about the Fintech boom? Not necessarily. Banks have both the resources and the ability to retain their position in a way that start-ups really don’t. They just need to adopt a bit of Fintech thinking. Banks can try some of these simple and practical things in the short term that could make a significant difference:

  1. Play with some data around a recommendation engine – It can be done as an experiment with a few people. Group customers by preference, products by customer, and transactions by pattern similarity. Everyone’s always looking for the elusive ‘Single Customer View’, but guess what? A ‘Partial Customer View’ linking two to three product portfolios is already enough to get started.
  2. Look closer at payment and behaviour data – Payments can help banks understand the sequence of events that leads to somebody leaving the bank. Payments can reveal hidden social networks within a bank’s portfolio. Customer-to-customer, customer-to-merchant, company-to-company, product-to-product – what could you do if you knew these relationships?
  3. Fraud and compliance – As mentioned before, banks are incredibly adept at regulatory compliance and fraud mitigation. But the industry needs to start getting better at text analytics and using web behaviour to detect high-risk patterns. Insights such as ‘who clicked on what before fraud happened’ can be very enlightening. These days, companies can match weblog data with branch data and check the difference between web and in-branch behaviour.
  4. Service experience – In the brick-and-mortar era it was ‘Location, Location, Location’. Now, in the digital era it’s ‘Customer, Customer, Customer’. Use event data to spot processes that are causing problems for your customers and fix them. Contact Centre logs are a hidden source of insight. It doesn’t take much to parse them for sentiment and recurring patterns. There could be new products hiding behind these complaint logs, if only banks were inclined to look.
  5. Improve the mobile experience – Many banks have mobile apps but they usually concentrate on facilitating payments, fund transfers, and account management. What if a local bank’s app could act like Mint and provide the user with cool ways to manage budgets, see financial profiles at a glance, and even offer helpful advice? You can parse those mobile servers for hidden patterns in data (location profiles, IP addresses, mobile browsing, etc) – the ‘fingerprints’ of customer satisfaction.

Okay, these five things won’t turnaround troubled relationships on their own but they could be the first, tentative, steps towards reconciliation.

And once the ‘relevance’ and ‘confidence’ fences have been mended and an enterprise-wide digital transformation strategy embedded, banks can get back to developing meaningful, long-term, data-driven customer relationships instead of settling for a diminishing series of ad hoc, one-night stands.