Smart use of data moves the needle in financial marketing. And in a highly competitive market with few net new prospects, financial institutions need to push the value of customer data to the max.
One of the traditional mainstays of financial marketing, segmentation, may soon be eclipsed by a new approach — artificial intelligence-driven micro-segmentation.
First let’s acknowledge that there is a very attractive element to segmentation — it works. Bank and credit union marketers target consumers based on family composition, life stage and zip code to find high-probability groups that are more likely to be in the market for mortgage, personal loan or investment products. This is a great first step towards targeted marketing. Research says that targeted email campaigns are 760% more effective than non-targeted.
Your institution’s database holds clusters of people that share distinctive commonalities. Traditional marketing segmentation is the art of thinking in terms of large groups. These groups receive more targeted marketing campaigns, where the campaign profiles and customer groups match up. By decreasing the size of those groups, looking for distinctive commonalities, we get closer and closer to the mythical “Segment of One.
In micro-segmentation we are taking the traditional segmentation approach a step further. Micro-segmentation is a more advanced type of segmentation which is able to create groups containing small numbers of customers to form extremely precise and targetable segments.
Example of Using AI for Micro-Segmentation
A case study on using AI for micro-segmentation comes from BECU, Tukwila, Wash., which wanted improved email responses to highly fragmented user segments to drive additional mortgage, credit card and auto loan applications.
The credit union used the machine learning software from Amplero to automatically create and understand micro-segments and determine the best messages — subject line, image, content and offers — to drive increased applications, which was the goal of the campaign.
As shown in the image, the segments are automatically generated and presented in an “optimization tree.” This visually displays how the program divides the total population in groups and then splits them out further and further, creating very small (but for the marketer understandable) segments and optimizing them with the aim to get additional lift on the institution’s key performance indicators.
This drove a 10% lift for credit card, auto loan, and mortgage applications over the results of BECU’s previous campaigns.
Traditional segmentations lean heavier on a limited number of large groups using a combination of relatively static profile data relating to geography and demographics. For current clients this is sometimes extended with an RFM analysis. RFM — Recency, Frequency and Monetary value — is a marketing technique that analyses customer and sales data to find your most valuable customers, based on their past purchasing habits. It is used to establish groups based on customer value.
“Even with a good segmentation model, you might be missing out on the people that don’t fit into traditional segments but are interested in your message and product.”
— Jordie van Rijn, emailmonday
The question is: How homogeneous are these segments? It seems like there is an inherent bias into what would make a good segmentation model — until proven otherwise. And even then you might be missing out on the people that don’t fit into traditional segments but are interested in your message and product.
As David Chin, President and CMO of Lexer, said in an article I wrote about use of customer data platforms: “Pricing and discounting can be used as a lever to drive sales growth and to influence buying behavior. Using data to calculate the mean discount rates per customer and analyzing customer value and profitability is hugely impactful. You can segment customers by mean discount rate and use this for targeting during a current sale. [A company can] direct their customers to the products and categories that match their discount threshold.”
In finance and banking it often isn’t discounts we are offering, but other types of offers and content and micro-segmentation is a great step forward. For instance, a challenge banks and credit unions face is to effectively educate older customers on the benefits of mobile banking, while at the same time capturing tech-savvy clients with their mobile banking capabilities as a salespoint and offering competitive products.
You want to pinpoint the appropriate proposition for a segment and target them — and the look-a-likes — using the right message.
- How Financial Marketers Can Move From Data Analytics Angst to Action
- Expert Tips to Boost Bank & Credit Union Email Marketing Results
- Financial Marketers Must Link Data-Driven Segmentation With Addressable Media
Using All the Customer Data You’ve Got
With customer profiles, the more data you can work with, the better. It enables you to extract more from this “digital body language” that your audience is showing — the floods of behavioral data flowing from your digital channels. At the same time we see that financial institutions are using more “zero-party data”; self-reported data that customers give about their preferences, psychographics and needs.
The diagram shows the basic flow of micro-segmentation. The system proceeds through a number of automated steps: ingesting the data from different sources, analyzing it, generating the segments and the offer variants and then running the campaigns. A feedback loop with campaign outcomes is essential so the system can learn what works and optimize the segments and campaigns accordingly.
You can imagine that the more refined the messaging (and therefore the smaller the segments) the more effective micro-segmentation is. Think about it — your database is made up of individuals, each with their own motivations, interests, profile, habits, etc.
In micro-segmentation we are taking the traditional segmentation approach up a notch by layering-in more data into the model and assembling new segments out of fresh combinations of data.
The smaller size of the segments and subsegments allow marketers to be even more focused about their messaging. One factor to consider is that the automated learning does better the bigger the input numbers are.
How Do We Serve 1,000s of One-Person Segments?
While the idea of treating individuals as a “segment of one” sounds irresistible, the complexity of layering hundreds or thousands of data points to find cohorts might sound a bit tricky. But it is the key to uncovering important micro-segments.
” It was never possible before to tend to these smaller segments, but with the advances of martech, and especially AI, many of the previous bottlenecks are fading.”
— Jordie van Rijn, emailmonday
The solution is to let AI software do the heavy lifting of complex data analysis for you, helping to discover micro-connections in real-time. Ideally the AI is integrated into your digital marketing platform to make the data actionable. It was never possible before to tend to these smaller segments, but with the advances of martech, and especially AI, many of the previous bottlenecks are fading.
Think of it as having a machine learning program solve a giant Rubik’s Cube of data for a specific customer, answering the question: “What is the best messages at this time?” (By the way, robots are actually very good at solving Rubik’s Cubes and hold the world record.)
Email is a great channel in which to apply segmentation and testing. A marketer can influence the segmentation model and variants in the message. I use a simple, yet effective formula for email segmentation impact:
Segmentation Model x Execution = Combined Impact (In this, the Execution = Relevance + Content + Design + Timing.)
In the case of micro-segmentation, let AI uncover the segments and generate the most effective message variants.
AI Success Requires Working Towards the Right Goal
Micro-segmentation has proven itself and works in all channels. Once automated in campaigns, it’s easy to extend its use from email to SMS/text or direct mail.
Algorithms work best when built for specific tasks. A machine learning algorithm can optimize messaging and generate clusters of recipients only if it has a set goal or goals to optimize for.
Goals can include: “Increase engagement,” “Optimize customer value” and “Maximize number of purchases” versus “Maximize total revenues.” These goals will ensure the AI drives to the right (and best) results for the financial marketing campaigns.