For bank marketers, housing the data is just the first piece of the puzzle.
By Mark GibsonEver wonder how Amazon knows exactly what you need to purchase next—and when? Or how Trip Advisor emails you spot-on recommendations you cannot help but click on?
Many consumers and businesses appreciate and prefer companies that tactfully use the data consumers have given them to improve their lives or their businesses. It’s time for bankers to do the same. Your customers and prospects expect this.
As a result, a top priority for many banks is getting their arms around the myriad of customer data that exists throughout the organization, or the lack thereof. There is much talk of a ‘360-degree view of the customer’ or ‘personalized marketing’ based on a true picture of each customer and how they do business with you.
However, any bank executive or marketer who has investigated how to operationalize this vision quickly learns that the implementation path is complicated, confusing and fraught with difficulty.
This article attempts to identify relevant pieces and parts of this solution, explain each of them, and how they can fit together to approach the nirvana promised by the service providers selling the individual tools.
The pieces of the puzzle
American Bankers Association’s recent data analytics survey of bank marketers highlights that there is no standard approach to housing and managing customer data. While a data warehouse or marketing customer information file were most common, more than 30 percent of respondents used some other approach. The results–Data warehouse: 41 percent; MCIF: 28 percent; core system: 8 percent; CRM: 5 percent; data lake: 4 percent; other: 15 percent.
But as any marketer quickly learns, housing the data is just the first piece of the puzzle, and it takes more pieces to create targeted campaigns, let alone personalized ones.
Let’s take a look at each of those pieces and how they fit together into a complete data management picture, empowering marketers to run intelligent campaigns 24/7/365.
MCIF: The marketing customer information file has been around for a long time and is primarily used by the marketing department to household account and customer data. Rolling this data up at the household level allows marketers to see what products are owned by a household, and what products may be needed. This insight can be useful when developing customer onboarding, cross-sell, or retention programs at the household or customer level. It can also prevent an institution from implementing blanket programs that damage the customer experience.
For instance, implementing an inactivity fee on a savings account without understanding that the same customer has a large balance checking account and a commercial relationship might not be the best decision. But without an effective MCIF, that type of action happens all the time.
The MCIF is basically an aggregation of customer and household product data and limited utilization data. It is used to support basic cross-sell and up-sell campaigns. It can be housed in a server on-site at the bank or with the MCIF provider. A customer/household file is updated daily, weekly or monthly.
A part-time or full-time data analyst maintains it, performs analyses, reports and targeting lists for the marketing team and/or lead lists for the sales team.
A common use case for the MCIF would be to execute a customer cross-sell campaign, where the marketer identifies households that own a checking account but no savings account, then designs an e-mail or direct mail campaign to cross-sell savings accounts.
Data warehouse: One disadvantage of the MCIF is that its householding, product ownership and profitability data is good, but it is not 100 percent complete. It is not usually tied to real-time transaction and channel interaction data. Often its MCIF data does not sync with finance or operations reports. That is because the MCIF is not a comprehensive data repository. It gathers the information it needs to perform important marketing tasks, but not other data analytics tasks. That’s where a more comprehensive customer data warehouse comes in.
At its most basic, a data warehouse receives customer data from all sources, cleans it, standardizes it, stores it and makes it available to anyone in the institution who needs it. While this sounds straightforward (and critical!), aggregating, normalizing and organizing multi-source data is a large task that takes at least 12 months of constant team effort to achieve. However, once it is completed, the institution has a powerful tool.
What kind of data goes into the warehouse? In short, all of it. Core data including PII, transaction and channel usage data. Customer service data. Product purchase and balance data. Online and mobile banking transaction data. Wire transfer data. ACH data. Remote deposit capture data. Investment product data. Credit and debit card transaction data.
And remember, common data definitions and parameters are essential for all this data. And that probably does not exist today. Different vendors define the same data element differently. That needs to be fixed before it can be placed in the warehouse.
So, it’s much more complicated than an MCIF, but once completed, it serves the entire organization and provides insight and action in many more ways. Let’s think about a few marketing use cases that are enabled by a data warehouse.
- Which of the bank’s business customers use a credit card as the primary accounts payable tool, and maintain credit balances larger than $50,000?
- Which of the bank’s consumer customers with deposit balances over $100,000 have experienced over-limit declines from either their debit card or mobile deposit?
- Which of the bank’s customers with $1 million in investable assets owns a free checking account but easily qualifies for the best package checking product?
- Which of the bank’s customers are likely to qualify for wealth management? And/or own businesses and can be cross-sold business services?
Each of these opportunities, as well as hundreds of others, are enabled by pulling together all of a customer’s data in one place. However, you can imagine it’s a pretty large team effort. It requires a cross-functional team to select a vendor and manage the multi-month (or year) implementation
It requires a full-time administrator as well as at least one full-time data analyst. And it’s generally not as easy to use as an MCIF, so more specialized training is required.
The data warehouse is often located either in the Information technology or finance area. The analysts are either housed in a central business Intelligence group or can be decentralized and have remote access to the warehouse once they have become certified users.
Another significant advantage of the data warehouse is that, because data and definitions are standardized, as long as an analyst uses the data appropriately, reports are consistent across the institution, whether they are generated by business intelligence, marketing, or finance.
Customer data platform: Even if your institution has a data warehouse and marketing has access to it, that doesn’t automatically translate into automated marketing programs, let alone personalized marketing programs. You still need to run a query, run a report, then use the data in a report to build a campaign. And often higher priority finance or other operational needs can take priority to marketing requests.
A customer data platform can be thought of as a marketing data mart that replicates a smaller set of key customer/household data from the warehouse to enable marketing to quickly and nimbly use it to power campaigns and lead lists. The CDP builds customer profiles with complete details about each customer. It then provides those details to systems such as marketing automation or a CRM. It is typically updated daily from the warehouse so the information is timely. And it allows marketing to perform all these tasks without submitting a request to business intelligence or the data warehouse team and waiting in a queue.
Back to one of the use cases. Let’s say we want to run a CX/retention campaign for those wealthy consumer customers who have been declined for over-limit mobile deposit attempts.
We define the audience characteristics (who was negatively impacted), and use the CDP to ‘build an audience.’ Then an appropriate action or treatment is defined, a delivery method is specified (e-mail, contact center call list) and the campaign is launched.
If a marketing automation platform is used, the audience is automatically pulled from the CDP on a pre-defined basis (daily or weekly), and the delivery action automatically happens. Then, results reports are automatically generated.
Think of a CDP as a smaller version of the data warehouse that quickly provides relevant customer data that marketing uses to effectively target customer campaigns. The CDP is typically evaluated and implemented by a team of marketing, IT, and the data warehouse administrator, and it typically sits in marketing.
An important caveat here is that fully automated campaigns require a marketing Automation Platform in addition to a CDP. And each campaign needs to be designed up front before it can be automated. This is important because many marketers assume that once they buy a MAP, they will instantly have a bunch of marketing programs running. No, no, no. You need a dedicated resource to identify and prioritize which campaigns to automate (i.e., consumer checking onboarding, savings cross-sell, retention of checking clients with $100K+ balance, etc.), then each of these needs to be built in the system.
Data requirements need to be created, data elements need to be identified and sourced from the CDP or data warehouse, elements need to be from the CDP, reports and dashboards need to be designed and marketing or sales treatments (e-mail, lead list) need to be built in the MAP. Only then can the campaign can be automated.
So, one way of thinking about MAP is that it is a supercharged e-mail marketing system that has direct access to critical data, and many automated campaigns are stored in it, along with performance monitoring, management and reporting capabilities.
Data management platform: A data management platform collects, organizes and activates first-, second-, and third-party audience data from various online, offline,and mobile sources, including from your CDP. However, it stores this data in an anonymized manner, which is very different than the CDP. It then uses that data to build detailed customer profiles that drive targeted advertising initiatives for prospect acquisition. Recency is important here, since old behavior is not as predictive.
As a result, many DMPs only store a few months’ worth of data. However, because you are appending third-party data from places like Acxiom or Experian, these databases can be large.
Customer relationship management: We saved the best for last! CRM is probably the most talked about and least understood tool. Many executives use CRM and data warehouse synonymously, and there are similarities but they are very different. A CRM is a ‘relational database’ which establishes relationships between customer data, employee data, and various transaction and sales activities. So, it doesn’t usually ‘house’ the customer, employee, or transaction data. It pulls that information from the appropriate system, organizes it according to pre-defined rules, and delivers pre-defined opportunities to the banker.
A CRM can perform a variety of important functions for a financial institution. For instance:
1. From a customer service and CX standpoint, a CRM can maintain a record of a service encounter, allowing a branch employee to immediately see that the customer had an issue the previous day and spoke to the contact center and that the issue was resolved. The CRM can assign a ‘job number’ to each inquiry, as well as an ‘owner,’ and can report on how these issues are resolved (or not).
2. From a relationship management standpoint, the CRM can house notes from the most recent relationship call with a client, identify opportunities or action items, and set alerts for the banker to follow up at the appropriate time. And all of this can be tracked and reported upon.
3. Finally, from a sales standpoint, the CRM can help from both a cross-sell and a new business standpoint.
Regarding cross-sell, various trigger events can be established, and alerts can be automatically sent to a banker for follow-up. A simple example would be a commercial loan customer who exceeds a line of credit limit. Another would be the deposit of a check over $1 million. Or a wire request over $250,000. Triggers can be set up in a MAP powered by a CDP, then delivered and informed by the CRM.
Thinking about new business opportunities, many organizations have begun using digital marketing to generate new prospects for bankers. The way this typically works is that a prospect performs certain pre-defined actions (visits a landing page, downloads a white paper, fills out a form) which ‘qualifies’ them to speak with a banker. This qualification work typically takes place within the MAP, then the qualified lead is sent to the CRM where the banker is alerted to take action. Again, all of this is tracked and reported so that Marketing ROI can be calculated.
Start small! It’s important to get on this bus now and begin your journey. While mature customer data management may seem daunting, it is absolutely essential both for effectively meeting your existing customers’ needs and expectations, as well as for growing efficiently.
But don’t think you need all these ‘boxes’ to be successful. And you don’t need all the subject matter expertise within your bank. Personalization might be nirvana, but there are many successes you can achieve along the way. The important thing is to get started and have a plan to enhance your capabilities over time.
Don’t wait for the entire system to be operational. Build a team of internal marketers and external partners. Start with one ‘box’ or tool (MCIF, MAP, etc.), plan small wins, measure them, communicate them internally and build your business case for the next step on your customer data roadmap.