Two-times DataIQ Talent Awards 2016 award-winner The Royal Bank of Scotland (RBS) has transformed its frontline customer service as a result of adopting a big data environment and data wrangling solution.
Through developing an innovative and world-first tool for customer agents, it is now gaining customer insights from big data that previously went unspotted. As a result, Net Promoter Scores are up, complaints and repeat complaints are down, with a positive impact on costs and revenues.
Founded in 1727, RBS is an international banking and financial services company with over 30 million customers and 700 retail branches. Its 16 million UK customers are supported by 20,000 frontline staff across multiple channels, from traditional bank branches to online webchat. RBS has committed to becoming the number one for customer service, trust and advocacy by 2020.
Every month, some 250,000 RBS customers use its webchat service, enquiring about every aspect of its product portfolio and being provided with information by a team of 600-plus agents. The sheer volume of these interactions meant that <1% were being read, using a manual process focused on data quality. On average, only 200 web chats per month were being reviewed. This gave RBS a basic understanding of customer pain-points, but didn’t provide insight into how it could better serve its customers. The team was unable to make intelligent decisions about customer service based upon these limited and therefore potentially-skewed samples.
The existing storage system at the bank was not designed for the management and analysis of unstructured data at scale, meaning that potentially valuable insights and buying indicators were being missed. As well as potentially failing to meet customer service expectations, RBS was also missing out on root causes of complaints, leading to repeat complaints and higher costs.
RBS also wanted to listen to customer feedback proactively as part of its service goals, including getting 25% of customers to complete a satisfaction survey. Capturing this information would allow the bank to focus its staff training more appropriately.
RBS has an innovation unit based in Silicon Valley which tracks technology developments for their potential to be of benefit to the business. One key area of focus was data preparation tools which would support a migration of webchat data into a new Hadoop environment which the bank had invested in to manage unstructured data outside of its existing traditional database.
Having evaluated a number of potential vendors, RBS selected Trifacta. “We first met Trifacta through our Innovations Team on the ground in Silicon Valley and were impressed from the outset with their ability to quickly derive value from diverse, unstructured data sets,” said Christian Nelissen, head of data and analytics at RBS.
“When we took a closer look and evaluated them against a range of large and small competitors, they stood out for their commitment to simplifying complex processes, something which is now really helping us to deliver great solutions for our customers. As such, we’re excited about the benefits we’ll see from the latest advances to Trifacta’s user interface and transformation workflow. We are now on our way to building a world-class data capability that will help us better understand and better serve our customers,” he said.
The enterprise-level data wrangling solution provides RBS with the ability to run data discovery across multiple Hadoop data formats, with native integration into Hadoop and scalability as the project grows. Trifacta also offers integration with the bank’s existing tools, while its front-end usability has been a significant advantage in winning rapid adoption among non-technical users who can profile and explore data sets for themselves without technical user involvement.
For technical users, the new environment and tools have transformed productivity and the value of their activities by allowing webchats to be classified by topic as well as sentiment. By identifying how to respond to customer needs better, for example through identifying product interests or pre-emptive complaint resolution, millions of pounds in incremental revenue is being delivered, while also achieving substantial cost savings.
“In order to have a visual interaction with highly unstructured data, you’d have to be a person with 30 years’ of experience in data transformation technology. Trifacta data wrangling gives you that kind of a view on the problems out of the box. It’s not only that it has increased productivity, it’s that we are now addressing things that we wouldn’t even do before because it would just take too long. I can’t put a value on how much it has improved how I work!” said Spyros Marketos, data scientist at RBS.
The outputs from anaysing webchat data are delivered via an interactive, intuitive data visualisation toolkit giving a dashboard of key performance metrics. This data democratisation process also means frontline agents can identify where coaching is most required, leading to improvements in customer service. With a better understanding of topics that customers want addressed, RBS can more effectively train its chat agents and create more effective service and product propositions based upon real customer feedback.
“The dashboard is transforming the way I run my business. It is improving the customer-centric approach in our chats and it is showing in the output. We are now able to coach our staff much more effectively and it is based on the themes that are coming from the application,” says Akshay Vats, head of webchat operations India for RBS.
As it deploys Trifacta at scale, the company expects to analyse 100% of webchats, a huge surge from its previous 0.1% usage rate. As more and more of its customers turn to web chat for customer service, this improvement will only become more significant. Already, complaints have fallen by 15% while personalised feedback is being provided to agents, rather than a generalised Top 10 of the most-complained about issues. Repeat complaints post-feedback have fallen to zero.
RBS is continuing to develop its new big data environment with an eye on some further, major innovations. Using big data to improve webchats which are run by AI agents is one example, in line with the stated goal of RBS “to be there for our customers in the moments that matter.”