Building a better digital economy through data analytics

Digital marketing is cheaper for financial institutions than managing physical transactions, and traditional banks have been downsizing their U.S. branches for more than a decade. The COVID-19 pandemic has halted many personal services and accelerated the digital economy, which allows consumers to access money, goods and services through websites or mobile applications.

It is not surprising that banks are increasingly using modern technology; The most important thing is to improve customer and service. But financial institutions are still not getting the maximum benefit from their digital investments as users come into contact with increased expenses to address concerns about their assets.

A 2020 survey by management consulting firm Capital Performance Group found that online business activity, including marketing and other interactions, increased by 30% from the end of 2019 to December 2020, and the mobile market increased by 80% took it out. %. But it also shows that some organizations have doubled the number of call centers and continue to grow. A 2021 survey by Cornerstone Advisors shows that consumers are calling their banks more often because they can't find answers online or their financial institutions don't have the virtual support they need to solve their problems.

These statistics highlight the unfortunate reality that the products and services many digital companies have been offering for years, from payments to personal identification information like Zelle, still fail to meet customers' needs, primarily due to usage. They don't need it enough. To make matters worse, Zelle shifts a lot of money due to controversial issues, especially regarding fraud, unauthorized trading, and unintended consequences.

As a consultant to many of the largest companies in the United States, I have seen firsthand the challenges of digital transformation. Banks have valuable information that can help them develop strategies to increase customer retention and retention, but this information is not being used. Support teams often use small sample sizes when writing and monitoring quality assurance tickets; At one bank I work with, the sample size is 1% or less. Therefore, additional findings may lead to erroneous conclusions.

To truly solve these problems, banks need to develop detailed and comprehensive customer data analysis on a larger scale, including all calls. They must then use observed behavioral patterns to design and develop digital resources that meet customer needs. In this article, I will tell you the steps I take to help banks do this.

Creating a strategic Team

The first step to increasing customers' use of digital banking and improving customer satisfaction is to create an internal team of financial analysts and product experts who will access data collected from all methods and products. This group is important because there is so much information that one person can collect, examine, and draw conclusions from. The team must actively work with management to ensure that their findings can be implemented in the client's business.

I recommend breaking into small study groups for each item. In my work with companies, I tell each team how many customer calls they can eliminate using a particular feature or capability, so they can do what's important to their efforts accordingly. To calculate the potential for this reduction, I used a custom customer journey analytics platform we developed for user traffic and content friction. (Tealeaf and Google Analytics are similar tools you can use for the same purpose.)

Select datasets and set your goals
The second step is to analyze and gain insight into the various data sources accessed. along. all platforms and features. In a traditional bank, resources can be divided into two categories and various subcategories. For example, the team I created at the bank uses:

Products and business
1. Shopping as payment and savings
2. Credit card
3. Finance
4. Auto Finance
5. Financial Management
6. Contact Point


Contact Center Phone
1. Call Center Interactive Voice Response (IVR) Communication
2. Face to face meetings in sales offices
3. ATM interaction
4. Desktop Application
5. Mobile Application
6. Outbound Notifications/Alerts
Most of my analytics work is examining call center statistics, which I think is the area that needs the most attention. This is where most support requests come from: Banks often do not accept email requests from customers because they are too expensive, even if they are made abroad. Chat has replaced email, but it is only a small portion of all customer interactions at top banks, less than 5% of the organizations I consulted. In addition, call representatives have now become expensive for banks due to the high demand. Therefore, I will focus on this point while explaining the process.

Once the data is analyzed and made accessible, financial institutions can begin to develop measurable key objectives that can form the basis of problem-solving strategies and help define the project. Here are the goals we set for the banks I consult with:

Improving customer experience - measured by Net Promoter Score (NPS), a customer satisfaction metric that measures how likely people are to recommend the company Key Metrics - Across all pipelines
1. Increase adoption and digital engagement
2. Reduce calls to call centers when they add little or no value
3. Lower costs, meetings in the bank's branches
4. Increase efficiency across service operations while reducing risk
As expected, data shows that the primary channel for customer engagement volume is digital. But surprisingly, at the bank where I work, digital customers are more likely to find support than digital customers and bank customers. Digital marketing companies bring more than twice as many calls and inquiries to call centers as the banking industry.

Know The Reason Why Customer Are Calling

When determining the data set and objectives, the team should consider the type of questions the bank data will ask that will help analyze the nature and status of customer demand. At the bank I consulted, we focused on information on the call center, especially about the customer's interaction with representatives on the phone, and asked the following questions:

1. How many customers are talking to live customers?
2. Who are these callers, that is, what are their profiles in terms of service interactions across channels, transactional transactions across products, and customer value at the business level?
3. What percentage of these callers are working digitally? 4. What banking activity, if any, is occurring before the call?
5. If there was a job at the bank before the call, what job was it done? 6. What do customers say? 7. Did one of them call more than once? If so, how often?
8. What time does he call?
9. How long is the call time for our customers who call frequently?
While most banks track the number of calls coming into their call centers, they generally do not go into further analysis. This is in line with what I see commonly in the financial services industry: Companies keep track of events but do a poor job of evaluating events in their environment; This can explain behavior and help them improve. Knowing that a customer spent 20 minutes resolving a dispute or opening a gift card before calling will give call center staff important details and guide their interactions.

The team I created at the bank where I was consulting was established to help the company record all calls through the system. For example, we may see that an anonymous customer was online minutes before the call and tried to close his account but was unsuccessful; it is an experience that initiates the search. We will give each person a report about their purpose and time. We can also identify secondary and secondary causes of the call by analyzing the events surrounding the main catalyst, allowing us to paint a complete picture.

We then calculated an important metric called call delivery, which we used to measure overall call numbers. It can also serve as a reference point for improving performance.

1. Call rate: total number of calls expressed as a percentage of the total customer base
2. Contact rate: number of customer calls, also expressed as a percentage of all customers
3. Call Communication: Call Cost Report Communication Cost
Our main goal is to minimize calls and contacts to the highest banking standards, this is usually around 20% based on my experience and third party companies. Rates like McKinsey and Finalta are 1% and 10% per month respectively. Second, we want to make the two measurements equal, which means we remove duplicate numbers. When we achieve this, we can say that we have achieved the first solution, which means that customers only need to call to solve their problems, which is an important indicator for customer management.

Other measures we have identified and hope to reduce include:

1. Search time
2. Call forwarding
3. Update
4. Dissatisfaction
Call centers often use reasons for calls from customer relationship managers or callers, but systems are rarely used to drive improvements in these indicators. Access to the flow in the system can eliminate some calls altogether and, in others, help callers find the right call agent based on their previous activity, communication, customer profile, and level of service required.

Financial companies often use user experience management software to evaluate people after interaction to create a Net Promoter Score. At the bank where I work, we set (and achieved) the goal of increasing our call center to 55%.

Studying Call Patterns To Create Profile To Use

The next step involves analyzing call patterns using various machine learning data points to find out what drives customers to call.

1. Financial transactions such as point-of-sale (POS) charges, payments, and payouts.
2. Non-financial transactions such as address changes, card declines and disputes.
3. Related activities include phone calls, IVR communications, desktop games, mobile games, branch access, ATM transactions, and exit notifications.
4. Customer profile/segmentation such as profession, high-value marks (indicating high engagement), and power users.
5. Consumer products and non-working products, that is, the number of products each customer owns and whether they use them, are related to the company's fee-for-service. (Note: We focus on active customers to measure contact and call rates.)


We've collected some valuable and accessible information from these categories. For better understanding:

1. Aisle confinement: Tendency for customers to stay in one place for a short period of time, usually 15 to 20 minutes
2. Self-service: Customer self-service via digital, ATM or IVR
3. Channel Models or Preferences: Leads used by customers
After analyzing the above factors, we focus on lower impact by measuring and determining the representative phone call. Excluding searches longer than 60 seconds, we checked the search pattern and type, calculated the duration, determined how the search was performed, and identified the drivers we could find. We also collect information so we can see how customers spend their money with the company.

Using advanced analysis and measuring the duration of the call, we divided the population into segments:

1. Callers: These customers face problems that cannot be resolved independently, require human assistance for multiple interactions, and make multiple calls in a short period of time. (We define this as two or more calls in a 24-hour period.)br 2. Random callers: These random callers make fewer calls than groups, but there are regular matches and similar questions. Explosive callers made 3 or more calls over a six-month period with an average call duration of more than 48 hours and a difference in average call durations of over 100%.
3. Casual seekers: This segment consists of customers seeking information, characterized by frequent searches and various types of inquiries.

Discover Caller Needs as Per Group

Once the standard call review is completed, the data can be further analyzed by the team. In the bank where I work, this work is presented in detail and has the ability to focus. For example, analysis of one of the events showed that 5% of the population consists of a group of people searching, 30% is dispersed, and the rest consists of people searching. Most of the people in the group are digital, refugees see less digital work, some of them are daily or frequent callers.

This chart shows a 12-month view of live callers. Three years of consolidation around this project:

The cluster population, although small, has a significant impact on NPS and complaints. Based on these findings, we offer several insights and suggest strategies to achieve key objectives, starting from clustering. Here is a summary of the identified issues that caused the number of calls to increase:

1. Password/Username Verification
2. Fraud/Unauthorized Transactions
3. Zelle Payment
4. Outgoing notification not delivered
5. Advertising Fees / Insufficient Funds

Based on our deep understanding of customer support, our first step to reduce customer calls is to specifically improve the online experience and limit customers to their favorite codes moving forward. . I also work with the account team to gauge pain points and better develop investigative plans to resolve issues, and we use a pool of licensed telephone representatives to address this group. These agents may avoid fees or penalties, while other agents refer callers to a more senior agent.

A bank entered the wrong email address for most of its customer group, leading to a wider investigation to reveal that more than 10% of many customers had the wrong email address. A campaign was launched via push notifications and secure emails to inform customers and change their addresses.

Different callers are referred more to non-routine services such as payments, deposits and cash withdrawals, and less to non-routine services such as payments, deposits and cash withdrawals. Zelle Money Transfer. Their favorite channel is phone and almost 10% are labeled as heavy or normal channels. The group's strategy is to provide targeted messaging and guidance on digital usage and engagement.

Most of the callers ask questions about new features, products and branch hours. Some of this information is added to the online experience to reduce searches. During the peak period of the epidemic, customers need to make an appointment to visit the branch. The call center of a bank I consulted received thousands of calls for installation; So I worked with the digital team to help customers book appointments online, cancel more calls, and save money.

By dividing people into three segments, we have been able to create a model that helps the banks I work with better understand behavior, patterns and usage factors in many aspects, especially digital and telephony. These strategies are key to reducing rising call center costs even while the customer base remains the same. It also aims to improve customer experience and service levels while reducing wait times, changes, upgrades and callbacks. Three years later, one bank's results were particularly impressive.

1. Digital jobs increased by 10%
2. Contact rate decreased by 12%
3. Call rate drops by 15%
4. NPS increased by 5%

The decrease in contact value and call value also reduced many important factors. When we analyzed calls that do not add value to the bank, we found that many calls at the agency can be made self-service by customers, saving money and reducing the likelihood of human error due to rejections, product changes, accounts. Opening and closing and other customer service questions.

Banking and other financial services do not have enough information for analysis, and reducing calls and charges is just one example of this. You can make efforts to promote positive change: you will find other important things or use different methods of collaboration. The challenge is to translate your message into insights that improve customer experience, increase digital engagement, and save money. A good understanding of what works and what doesn't can lead to gradual changes that will make the body work better for everyone.