You don’t need to be a professional chef to know the food at a restaurant is good… or not. Similarly, marketers don’t need to become experts in machine learning and artificial intelligence just to use data analytics effectively.
That’s according to Christopher Penn, Co-Founder and Chief Innovator at Trust Insights, a leading expert on AI in the banking industry. In fact, he says the most sophisticated form of machine learning is “between your ears.” Penn maintains that marketers just become more skilled at asking “really good questions” if they want to become more effective data-driven strategists.
“You need to become the ‘Chief Questions Officer’ in your organization,” he says. That means asking questions like, “Is there a correlation between website engagement and credit card sales?” After asking such questions, marketers need to then push the data specialists they work with — internal or external — to extract and analyze their bank or credit union’s information to get the answer.
In 2019, humanity is expected to produce 40 zettabytes of information. (A zettabyte is a “1” with 21 zeros after it.) To give that some perspective, Penn says that if you began binge-watching Netflix now, it would take you 55 million years to consume just 1 zettabyte of content. His point is that financial marketers have insane amounts of data to digest daily.
Even the marketing technology designed to cope with that data presents challenges. Today’s Marketing Technology Landscape encompasses at least 7,000 different marketing software tools.
Penn suggests the way for financial marketers to avoid drowning in this analytical maelstrom is to tackle it within a framework of people, process and platform.
Read More:
- Data Analytics Pitfalls Banks and Credit Unions Must Avoid
- The Art and Science of Data Analytics in Financial Services
Key Digital Imperatives for Gen Z and Millennials
As your financial institution prepares for the impending Great Wealth Transfer, ensuring that your digital strategy appeals to younger generations is essential.
Read More about Key Digital Imperatives for Gen Z and Millennials
Messaging Mastery: Tailoring Your Brand for Every Audience in Digital Advertising
This webinar explores the challenges of crafting tailored brand messages and the critical relationship between audience understanding and effective multichannel digital advertising and strategy.
Read More about Messaging Mastery: Tailoring Your Brand for Every Audience in Digital Advertising
The Three Experts Marketing Needs on Staff Now
“One of our sins as modern marketers is that we are a little over-reliant on platform and under-reliant on people and process,” says Penn. There are three types of people financial institution marketers need to assist them in managing and exploiting oceans of data.
Developers — All financial institutions have data scattered throughout the enterprise in different forms and systems. “Developers have the technical skills and tools to connect the data throughout the institution and pull it into one place for processing,” says Penn.
Data Scientists — These are the people who refine and process the data — clean it, standardize it, eliminate missing data, correct anomalies and so forth — so that it can be used to provide insights and drive marketing actions.
Marketing Technologists — Employees with a blend of marketing and IT skills who can implement marketing technology platforms and tools — from simple dashboards like Google Data Studio to enterprise resource planning (ERP) systems — to create business outputs.
Of course, banks and credit unions with one or two people handling marketing may not be able to afford such resources, but they can tap outside expertise through vendors or build up their own skills with educational materials, several of which are listed toward the end of the article.
Four Marketing Process Requirements to Adopt
As noted, marketers have not typically been known for strong process management. Now, however, they need a way to govern the technology that has become an integral part of their discipline.
There is no need to create a framework from scratch. Penn suggests financial marketers use a proven framework used by IT for decades — ISO 38500-2015. Here are several key elements of this process governance framework.
1. Align with business strategy — It seems like such a basic, but Penn says he is surprised how many marketers do not connect their marketing analytics efforts to business goals, says Penn.
2. Document martech — A bank or credit union’s marketing technology strategy should be documented, covering such points as who is responsible for checking Google Analytics and implementing the findings.
3. Create a marketing balance sheet — Penn considers this step critical because it reflects the Marketing’s “capital.” “Data itself is an asset,” he says, and some of the processes marketing develops qualify as intellectual property. “If you have a cool process for putting together stories on Instagram, that belongs on your balance sheet.”
4. Establish clear responsibility for security — Another critical point because data analytics can create massive vulnerabilities under regulatory requirements like the European Union’s General Data Protection Regulation. Banks and credit union marketers, for example, must install safeguards to be sure that personally identifiable information isn’t fed into a new lead generation tool.
Read More:
- 15 Applications for AI and Machine Learning in Financial Marketing
- How to Build a Data Strategy That Yields Big Insights and Maximizes ROI
Moving Up From The Bottom
Five levels define the hierarchy of analytics that financial marketers should use to guide their use of data. Penn says these include : Descriptive, Diagnostic, Predictive, Prescriptive and Proactive.
The descriptive level is the starting point — where nine in ten companies are stuck, according to Penn. At its most basic it means ascertaining what data an institution has and where it resides. This quantitative stage tells a financial institution “what happened,” through such tools as Google Analytics, Google Search Console, Hootsuite or even an Excel spreadsheet. While important, this is “only the beginning of the analytics journey,” says Penn.
The diagnostic level tells marketers “why,” for example, a consumer chose to apply for a consumer loan — or not. Using tools Google Surveys or Google Optimize or similar tools can do this. More sophisticated analytics tools, using attribution analysis, can determine, for example, which channels contribute the most to online conversions.
The next step in the analytics sequence is more powerful. This is predictive, or “What’s going to happen next?” With this capability banks and credit unions can build content calendars far enough in advance to take advantage of opportunities suggested by analyzing past campaign results.
The prescriptive and proactive stages of analytics are quite advanced and require very sophisticated technology such as neural networks, according to Penn.
Where Financial Marketers Can Build Their Knowledge About Data Analytics
- “AI For Marketers: An Introduction and Primer, Second Edition,” a book by Christopher Penn
- “Design Basics for Creative Results,” a book by Bryan Peterson
- “The Visual Display of Quantitative Information,” a book by Edward Tufte
- Statistics for Applications by MIT OpenCourseWare (free)
- Google Analytics Academy (free)
- Hubspot Academy (free)
What Smaller Financial Institutions Should (And Shouldn’t) Do
During the webinar, Penn was asked by a credit union marketer: “We have 3,000 members and about 200 employees. How deeply does an institution like ours need to go with analytics?”
His response: “You need to go as far as you can until you stop seeing results.” Concentrate first on “What happened” and then, “Why did those things happen,” Penn also says that going by “gut feel” isn’t wrong for financial marketers to rely on if using predictive analytics tools is not feasible.
That said, Penn points out that the term “data sciences” emphasizes the need to be scientific about analyzing data. Don’t make an assumption, he says, such as “more website engagement increases sales.” Instead, build a statistical model and figure out whether there actually is such a relationship.
Finally, Penn observes that a big part of every marketer’s job is “data storytelling” — being able to explain to people within the institution what the data says. “What happens too often with analytics is that marketing people back up the data truck and pour it out. That’s not a good idea if the recipient is not an analyst.”