An ABC Introduction to Data Mining for Dollars: Slicing and Dicing Your In-House List for Profit (Part 2 of 2)

In my last post, I introduced the RFM method, an effective direct response strategy to slice and dice your list for better conversion rates. The “R” represented recency—how long your customers have been with you. Today, I’m going to talk about the other components of frequency and monetary.

In my last post, I introduced the RFM method, an effective direct response strategy to slice and dice your list for better conversion rates.

The “R” represented recency—how long your customers have been with you.

Today, I’m going to talk about the other components of frequency and monetary:

Frequency
This segmentation tactic is another way to break down your house list: by how frequently customers have bought from you. So once you’ve divided your list based on recency, you look at it in terms of your customers’ purchase behavior. First, you identify your multi-buyers—customers who’ve purchased more than one product from you. You then split this list further, segmenting out two-time, three-time, four-time (and more) buyers. Those who have bought from you most often have proven their loyalty and obviously like the products and services they’ve been getting from you.

So if, for example, you’re considering launching a new product with a high price point, these would be your best prospects.

Monetary
Finally, you look at your list in terms of money. One way to do this is to divide your list by the amount of money each customer has spent with you. You might, for example, assign a benchmark dollar amount, such as $5,000, $10,000 or more. Customers at that level make up your “premium buyers.” This is the group that has the most favorable LTV for your company. These are your “VIPs.” Once you discover who your VIPs are, you can design products or offers specifically for them. Let’s say you have some kind of exclusive—and expensive—lifetime membership club. You would market this to multi-buyers who also fall into your “premium buyer” category.

If you offer payment options to your customers, another monetary way to divide your list is according to the payment options they have chosen: monthly, quarterly, yearly, etc. This will help you determine the initial purchase tolerance of each group of customers and which ones may respond best to future price points. As you can see, by looking at your customers’ purchasing habits—recency, frequency and monetary—you can identify the best customers for certain products. And by offering a product to customers who are likely to want it, you can improve your conversion rates.

By using the proven RFM method and other data-mining techniques, I’ve seen conversion rates double and triple. I’ve also seen inactive subscribers’ open rates surge from 0 percent to more than 30 percent.

However, many companies that send emails don’t have the capacity for data mining.
Unfortunately, some smaller businesses or start-up companies typically cut robust email features and analytics for cost savings. Oftentimes, these companies save money using online email service providers that can certainly get the job done, but don’t offer segmentation tools that allow for list analysis, where you can dissect your database into subgroups or “buckets.”

So when you’re searching for an email service provider, try to project what your segmentation needs may be going forward and if data mining is a strategy that you’ll want to deploy.

Hot Tip! When looking at email marketing companies, make sure you ask if there’s a list segmentation or data mining feature that can easily be done through their email platform. Find out the level of segmentation capacity (how far the segmentation of data can be drilled down to); if certain segmentation features are a standard feature or an upgrade; and what those costs may be on a monthly basis. Sometimes it may be an additional fee, but will certainly pay for itself over time.

Create a Bucket List

Whether you’re new to database marketing or a seasoned pro looking for some new idea to get your creative juices flowing, one of the most useful, and impactful, activities you can embark upon is to create what is called a “Bucket List.”

Whether you’re new to database marketing or a seasoned pro looking for some new idea to get your creative juices flowing, one of the most useful, and impactful, activities you can embark upon is to create what is called a “Bucket List.”

No, I’m not talking about a building a list of activities that you and a middle-aged companion wish to complete before you shed this mortal coil. I’m talking about taking a long, hard, and close look at your customers or prospects and getting to know them—really getting to know them—well enough to create broad classifications about who they are, what they do, what they like, and what affinities they share.

Remember, at the end of the day, database marketing is about sending out the right message to the right people at the right time—and, hopefully, achieving the desired response from the customer or prospect as a result. And without proper customer segmentation, this task simply cannot be done cost effectively, if at all.

Now, of course, there are many great customer segmentation models out there you can use. In a great article titled “Selecting a Customer Segmentation Approach” by Andrew Banasiewicz, Director of Analytic Services at Epsilon, four groups are identified: Predictive, Descriptive, Behavioral and Attitudinal.

Out of these four, the Predictive and Attitudinal models are arguably the most popular and widely used. Predictive is a model that uses value segments driven by customer purchase behaviors, extrapolating past behavior into future actions. An Attitudinal model, on the other hand, identifies affinity segments based on respondents’ expressed attitudes toward a company’s brand or products.

Now of course this list isn’t exhaustive and there are other models you can use. One popular alternative is Psychographic Profiling, which is used widely in the B-to-C space. In this model, consumers are assigned into groups according to their lifestyle, personality, attitude, interests and values.

Many B-to-B marketers, on the other hand, may prefer to use a segmentation model based on Firmographic variables, such as industry, number of locations, annual sales, job function and so on. Many software companies, not surprisingly, trend toward usage-based profiling, which includes variables, such as type of device used (desktop, tablet, mobile device), Operating System and so on.

One important fact that’s routinely overlooked is that successful customer segmentation requires taking a holistic approach. This includes aligning a firm’s segmentation goals to its marketing objectives and data acquisition investments. In other words, the data you have will determine not only which model you use, but also what marketing campaigns you’re able to run.

Now of course both data inputs and needs are in flux throughout the firm’s lifecycle. As Banasiewicz points out, for a firm in high-growth customer acquisition mode an Attitudinal model might work effectively for demand generation initiatives among qualified and segmented pools of prospects. Marketing campaigns in this scenario, we can assume, would speak to customer desires and affinities, with purpose of lead generation/nurture.

On the other hand, once the firm has acquired a large pool of customers, it’s not unrealistic to think that transitioning to a Behavioral model using inputs from past purchases will be more effective for running what are now CRM campaigns, focusing on driving lifetime value and repeat purchases.

Different groups not only have different attributes and attitudes, but consume different types of media. As such, they will respond to different types of offers, communicated in different ways and in different places. Where should a firm spend its marketing budget? Online display, email, direct mail, social media, print? … The choices are dizzying in today’s multichannel environment. Having a robust customer segmentation model can definitely help in the decision-making process.

Another important feature of customer segmentation is the realization that different customer groups can not only have wildly different demographic and psychographic identities, but very often will have strikingly varying lifetime values. To the surprise of some, a customer segment with a with younger average age will very often have a higher lifetime value than a group far senior to them, despite having far less disposable income to spend today. This may be based solely on the fact that the younger customers have, simply by being younger, many more years of being a loyal customer ahead of them. Taking this into account, many brands’ obsession with successfully penetrating the youth market should come as no surprise.

Now of course it’s easy to miss the forest for the trees, as customer segmentation is simply a means to an end, not an end in itself. Once you have broken your customers or prospects down into segments, the trickier (and for those who are not data geeks) more fun part of the equation involves devising incentive and reward strategies for each segment, and creating compelling marketing messages and collateral that can be used to get the message out across the various marketing channels. Knowing your customers, this part is a lot easier, which brings me full circle back to my point from the top: Create a bucket list.

Getting Started With Email Segmentation

Creating effective email connections that drive response and revenue requires segmentation. That sounds fine in concept, as many marketers know they need to do more segmentation in order to engage subscribers and break through the clutter. However, many marketers struggle with getting access to data and developing creative approaches that match the customer lifecycle. I urge you to not be intimidated. Demand greater data integration and access from your vendors. Start testing new content and creative approaches so that you can be automated and fully functioning immediately.

Creating effective email connections that drive response and revenue requires segmentation. That sounds fine in concept, as many marketers know they need to do more segmentation in order to engage subscribers and break through the clutter. However, many marketers struggle with getting access to data and developing creative approaches that match the customer lifecycle. I urge you to not be intimidated. Demand greater data integration and access from your vendors. Start testing new content and creative approaches so that you can be automated and fully functioning immediately.

Ease into segmentation to avoid overtaxing your precious resources. Use early tests to learn about subscriber interests and understand key success metrics. Doing so will build your confidence and help you make a case for automation, data integration and creative services — all of which are essential for advanced segmentation and better results.

There are two ways to get your arms around your segmentation opportunity, both with the goal of “right message, right person, right time.”

  1. Segment by customer profile and craft messages around customer demographics, firmagraphics and behavior.
  2. Segment by customer life stage and speak to customers who are in specific life stages.

Customer profile segmentation: With profile approaches, even simple segmentation can make a big difference. Separate your file into large segments that distinguish subscribers by a factor that has significance to your business. Clickers are a good place to start. Those subscribers who have clicked on something in the past month are more likely to be engaged than those who haven’t.

You can do less storytelling with clickers. For example, a retailer may simply alert clickers that a sale continues until Friday or put in specific sale prices for pants, sweaters and scarves. A business software marketer may send clickers three of their most popular whitepapers or an invitation to participate in a LinkedIn community. In both cases, clickers need less background info and more options to get them to act, whereas nonclickers may need more guidance and education prior to taking action.

Why burden clickers with info they don’t need and that gets in the way of their actions? At the same time, don’t skimp on critical storytelling information for nonclickers, as they clearly don’t have a strong connection with your offer yet.

Other starter segments worth testing include new subscribers versus long-time subscribers, buyers, geography (e.g., north versus south) and gender.

Draw segment lines around key drivers for your business — i.e., differentiators that give you a clear path to a custom message that will make an impact. For many B-to-B marketers, the most important driver of customization is job title. For B-to-C retailers, the key data point is most recent purchase. Don’t choose geography if location has no bearing on purchase behavior. Your business is unique, but good marketers understand the key customer attributes that lead to increased sales and satisfaction. Focus there for your segmentation and you’ll be rewarded with the biggest lift.

Life stage segmentation: To effectively segment by life stage, first abandon the notion that every email program has to be a long-term affair. Short-term email conversations can be even more powerful, particularly because they address a specific need at the time when that need is most acute. A four-message reminder series that disrupts the messaging flow around renewal time can be much more powerful than a generic newsletter which comes like clockwork every two weeks. Why not replace the newsletter with custom messaging for all customers who are up for renewal in a particular quarter? Similarly, create custom series of two messages to 20 messages that cluster around that particular life stage.

Remember to think through the dialog of the conversation if the message series is longer than three messages so that you can intensify, cease or adapt the message stream to accommodate response. For example, stop pitching the purchase midstream if someone has already upgraded from a free trial.

Not every email program has to be long term. The goal of “right message, right person, right time” can be achieved through segmentation that focuses on a specific life stage as well as customer profile.

What are your biggest challenges when it comes to nurturing engagement via segmentation strategies? Perhaps I can address them in a future blog or learn from others handling of them.