As the financial sector continues to mature in its approach to data analytics, leaders are shifting their focus from insights to impact. 

Tomorrow’s winners will be those who successfully make the case to invest in harnessing data sources like speech and text. These sources contain the richest information to identify innovations and efficiencies that will deliver business outcomes. 

The data advantage for financial institutions

Spending on data solutions grew 10 per cent in 2021, with banking topping the list of industries making the largest investments. Using data to deliver more personalized experiences is a priority in financial services today, not only to differentiate against competitors but to drive further efficiency.

Take contact centers, where a number of manual processes currently push cost into operations:

• Identifying the root causes of customer contact by having agents manually tag reasons for calls — or by capturing feedback from the dwindling minority of customers who respond to post-call surveys

• Trying to optimize service recovery with only a partial, subjective view of why customers are calling in the first place

• Labor-intensive quality assurance processes, where managers listen to a sample of call recordings (which may or may not be representative) and then manually populate scorecards to measure and coach team performance

And yet, whenever an agent engages with a customer, that dialogue has the potential to yield dozens of micro pieces of intelligence about underlying issues and their relative importance. 

Consider the potential cost saving for organizations that can leverage those data points, in an automated way, to improve processes. The exponential increase in data availability that we are seeing represents a treasure trove for financial institutions — the challenge is harnessing it correctly. 

Companies need a plan to convert this data opportunity

For a start, the datasets we are talking about are far too large for humans to process. Retail banks alone typically have thousands of data points for every customer: online payments, inbound/outbound calls, branch interactions, ATM withdrawals — not to mention profile information captured for identification and verification. 

Critically, the majority of data increasingly comes in unstructured formats.

That means we are not talking about straightforward numerical data readily indexed in row-and-column databases. Instead, they are frequently text-heavy and content-rich: emails, complaints, social media, agent or relationship manager notes — and, in the case of contact centers, chat logs and speech. Analysts estimate as much as 80-90 per cent of all data is now unstructured.

Consequently, traditional methodologies for interrogating data are inadequate. While all financial institutions have invested in data lakes, how many have successfully moved beyond curating or harmonizing data, and have front-end systems that truly enable them to make sense of all the customer experience intel they collect? Complications in integration and the need for investment further hinder the process. 

Successfully challenging this paradigm starts with embracing a three-pronged strategy for delivering data-powered financial outcomes:

1. Identify the business case for prioritizing unstructured data

First-movers have started realizing the potential of unstructured experience data to deliver bottom-line impact. Unfortunately, too many still treat customer experience as a compliance function — capturing simple survey scores to report on satisfaction with agents. But, increasingly, leaders demonstrate the strategic vision to leverage customer data to solve business challenges instead.

Real life case studies help showcase how unstructured data can reduce operational costs. Shortly after the pandemic struck, GM Financial’s inbound chat requests ballooned from 10 per cent to 60 per cent of total contacts. By applying automated theme detection and emotion recognition across all conversations, the company was able to determine the themes it needed to equip its chatbot to handle, enabling the bot to manage 50 to 60 per cent of all inbound requests without costly human intervention. 

Similarly, identifying the relevant business case for your organization will be critical for unlocking investment.

2. Onboard the right technology

Next, financial institutions should look for solutions that not only render speech into text, but also offer industry-tuned topic models that can detect themes across all text-based data sources. Such capabilities empower organizations to categorize root causes automatically across every call. Additionally, solutions exist that can analyze key drivers of customer experience, and predictively quantify their impact on operational outcomes, like first call resolution and handling time. Such insights can dramatically reduce managerial effort when capacity planning or optimizing for service recovery.

Sentiment analysis, too, can identify coachable gaps in agent competencies. Rather than having humans listen to calls and populate scorecards, best-in-class solutions can intelligently score interactions for you. What do your top-performing agents do when they are able to engage with an angry customer and somehow neutralize those negative sentiments, taking the call to a positive emotional place? The right technology can help isolate those behaviors, enabling you to coach and replicate them across your frontlines.

3. Plan to convert the data opportunity

Technology is a powerful enabler but not a magic bullet. 

Data-rich financial institutions often find that customers actually suggest fixes that might reduce friction. Each of those suggested actions has the potential to become an improvement initiative. Without the ability to take action, however, organizations will not see change. 

In one example, a major UK bank used text analytics to identify the root causes of complaints — but it was the bank’s ability to mobilize projects to tackle these root causes that enabled it to drive down complaint volumes by 100,000 in one function alone, avoiding $8.3m in operational costs. 

This illustrates why text is more powerful than surveys — after all, a customer’s complaint will reveal the issues that are top-of-mind for the customer, not just the survey designer. But institutions need to be thoughtful about establishing governance models that will enable them to prioritize such cross-functional improvement initiatives for investment. 

Beyond having mechanisms in place to enable project prioritization, early adopters are increasingly looking towards automation too. Key use cases here include being able to take theme and sentiment data, and trigger next best actions in systems such as marketing automation and customer support tools. 

Legacy IT solutions, complicated by decades of M&A and consolidation activity, can hamper progress — but pre-packaged integrations help. The pandemic laid bare how important it is for financial institutions to be able to evolve their operating models at speed. So, aspirational leaders should look for off-the-shelf connectors between experience management solutions and other tools, to enable live-time actions at scale.

Positioned for data-driven success

While the data revolution brings undeniable challenges for financial institutions, the opportunities are immense. As ever, technology provides part of the answer, but it is also incumbent on organizations to think deeply about the case for investment — as well as the mechanisms they need to drive customer-focused actions. The institutions that can do so will be well placed to convert the latent potential of unstructured data into business success.