10 Important Pieces of Data Banks Need To Collect On Commercial Customers

Predictive Data at Banks

While most banks understand the important data points when it comes to loans or deposits, most banks still could use help on collecting some of the basic information about their customers. The age of utilizing customer data to get predictive about risk, customer profitability and marketing is just beginning at banks so this is a new field for many. For example, a change in number of employees at your borrower is correlated to both credit risk and profitability. Knowing this data point, gives your bank an advantage over competitors as you increase sales effectiveness, increase risk-adjusted margins and be alerted to problems months in advance.


We have compiled a list of ten pieces of data that banks should consider collecting. We have found these data sets important to drive the most important decisions for long term commercial customer relationship management. Interestingly, this data is not only important for profitability, but also is the same data helpful in evaluating future credit risk and AML risk.


  • Email address – Of course, you need an email address for communication and marketing. However, these days most banks don’t realize that with an email address, more than 100 different pieces of information can be collected from various public databases. Having an address unlocks many other sources of information including helping in an online search.


  • Geography – Out of all the ten, banks do a great job at keeping the address current and in digital form. This comes in handy to geomap in order to not only understand concentrations and marketing, put to utilize the information to correlate to other pieces of information. While the classic example is to look at proximity to branch and compare that to profitability, other uses include adjusted credit to take into account changes in supply or demand in the area. Finally, banks sometimes use geography to set up alerts or to adjust credit for items like flood plains, proximity to earthquake faults or low lying coastal water areas.


  • Industry – Banks have been good at recording the NAICS code of the borrower, but have lagged actually using the data to drive credit stress, monitoring and portfolio management. Having a digital record (not just on paper) allows banks to set up monitoring and news about the industry to better manage the relationship.


  • Profitability/Products – The third most important piece of data is an aggregation of the products and services that the customer uses throughout the bank. Hopefully, this also includes profitability, but if not, at least it is a compilation of various touch points of the relationship with the bank. This should include not only major products like loans and deposits, but also minor ones such as electronic statements, overdraft protection and similar.


  • AML Ranking – Given regulatory risk these days, many banks have moved to some sort of measurement of downstream payment risk. Those industries such as payday lending, gaming, pawnshop, mortgages to non-US residents, may get a higher score. Understanding a customer’s source of funds and knowing how close the bank needs to monitor the situation can help with risk management as now the bank has the basis to quantify some of the risk, look at benchmarks and better manage the overall portfolio of customers.


  • Organizational structure – The ability to quickly discern where the borrower is domiciled (not just headquartered), when formed and what type of organizational structure the company is operating under allows another level of risk management and marketing. Recording how big a company’s board is can be utilized as a predictor of risk. Partnerships often carry more risk than sole owners. Legislation changes, political changes at the county level and ordinance restrictions all play a role in risk. Outside of risk, partnerships often have different banking needs than sole proprietorships so this data set can be helpful for marketing.


  • Related party – This is a record of other entities associated with the borrower. This includes parties, such as relatives of the borrower, affiliates, subsidiaries or parent companies at the bank (often called “householding of accounts) and outside the bank. This information helps with risk management (particularly credit), expected profitability and marketing.


  • Number of employees (and change) – One of the top predictors of profitability and risk is to be able to record the number of employees and the year-over-year change. This not only gives an indication of growth, but also determines what products the customer might be interested in.


  • Top stakeholders – This is a new area of risk management at many banks, but it pays to understand the top customers of your borrower. Maybe it is understanding some of these ten items on major tenants for commercial real estate, but it also could include looking at the largest suppliers for a manufacturer or largest contracts. Changes in top customers for example, have proven to be an early predictor of credit risk. Understanding the composition of revenue and expenses of borrowers is not only good relationship building, but will soon prove to be a good risk management practice.


  • Financial – This set is probably the most important set of information to record digitally and to drive decisions from. Level and changes in revenue, operating margins, net income, inventory levels (or occupancy for CRE) and salaries are statistically the most important items to collect and analyze. While there are many others, the mentioned items have tended to provide the overall best level of information.


None of this data has to be collected all at once and much of the information can be collected from various data sources. As implied, it is important to not only collect this information, but to be able to securely access the data in electronic form.


Banks are just beginning to understand how to get predictive with their data sets, but we do encourage banks to start collecting the above in order to be able to utilize and validate future models. In coming articles, we will discuss how best to utilize some of the above information to build predictive relationships that will reduce risk and increase sales. Until then, take a look at the above, add or subtract, but at least consider grabbing the data now so you can dramatically impact the bottom line later. 

Add new comment

Filtered HTML

  • Web page addresses and e-mail addresses turn into links automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <blockquote> <code> <ul> <ol> <li> <dl> <dt> <dd>
  • Lines and paragraphs break automatically.

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.