Chapter 7
Birds of a Feather Buy Together: Segmenting Your Customers
In This Chapter
Understanding demographic data
Grouping customers by age, income, lifestage, lifestyle, and geographic data
Using geographic data
Avoiding legal landmines regarding data
Your fundamental goal as a database marketer is to take advantage of information you have on your customers and use it to your benefit. You want to understand who is likely to buy and how to entice them to do so. You want to speak to them in a relevant way. You want to show them that you understand their particular needs and are able to meet them.
But your customers are not all the same. You will find substantial variation in virtually every customer trait that you track on your database. You will, however, be able to find pockets of customers who share similar traits. These pockets are frequently referred to as segments.
By grouping your customers into segments, you can focus separately on how best to serve each segment. Understanding the common traits of each segment will help you to define your offer, your message, your marketing channel, and even the timing of your message.
This chapter looks at some common types of customer groupings. The mechanics of coming up with useful segments usually requires some technical and statistical knowledge, which I don’t have room to address. I focus instead on the underlying data and how you might use different types of segmentation schemes.
Understanding Demographic Data
Demographics refers broadly to the statistical study of a population. What is the birth rate? How many people are married? What’s the average age of the population?
In marketing, understanding the demographic traits of a particular audience is crucial. For example, the whole art of purchasing television advertising revolves around understanding the demographic makeup of the audience. TV stations have a very good sense of who is watching and when. Providing that data is precisely the purpose of the famous Nielson ratings.
You may occasionally be asked to provide customer profiles to assist your advertising department with its media strategies. Remember, it’s not just about TV. Radio stations, magazines, and even billboards all have audiences who share particular traits. Your customer database can help in getting advertising in front of the right people.
The success of your database marketing campaigns also depends on your having a pretty good sense of whom you’re communicating with. Demographic data is central to all aspects of a marketing campaign. Whether you’re selecting a target audience or trying to customize your message, chances are good you’ll be using demographic data.
Common types of demographic data
The demographic data commonly found in marketing databases falls into a few general categories.
Financial data: Household income is probably the most widely used piece of financial data. In the financial services sector, sometimes net worth is more useful, although it’s typically much harder to get at. Home value is another common trait.
Gender: Certain products are gender specific. Gender is also used in selecting the so called head of household. This is the person who receives your marketing communication and is considered to be the primary decision maker. The world of marketing is not the world of Beaver Cleaver, however. The head of household is usually female.
Lifestage data: Age, marital status, and presence of children are all examples of lifestage data. The number and ages of children can also be important.
Lifestyle data: Lifestyle data revolves around consumption. Does the customer own or rent their residence? How many cars, boats, TVs, or motorcycles do they own?
Geographic data: This essentially amounts to the customer’s address. But there are some standard ways of grouping addresses together, which I discuss later in this chapter.
Where does demographic data come from?
You can get some demographic data directly from your customer. This is known as self-reported data and is generally considered to be the best and most accurate source. But in many cases, you can’t get it.
That’s when you turn to a third-party data provider. A number of companies will take your customer list and attach a wide array of demographic data to it, generally at the individual household level. Several of these companies have demographic information on virtually every household in the country.
At various times in my career I’ve done business with three of the largest marketing service providers. Acxiom, Epsilon, and Merkle can all provide customized segmentation schemes. The major credit bureaus, Equifax, Experian, and TransUnion also provide segmentation services.
These third-party databases are assembled from a variety of sources. The USPS, the U.S. Census Bureau, and other public records all provide valuable demographic information. This data is all public record data and so doesn’t generally raise privacy concerns.
Tread carefully with census data. Even though this data can be attached to individual households, it does not really reflect the individual household. It actually represents the average of a relatively small group of households that are geographically close together — a neighborhood, for example. The data for a given household is the same as for their next door neighbor.
These averages can compromise the quality of your targeting criteria, especially in densely populated areas. The residents of a large apartment building on the West Side of Manhattan may have, on average, two kids and a household income above $100K. But that doesn’t mean that everybody in the building meets those criteria. In general, it’s best to have data that relates to specific households if you’re using it to define target audiences.
Data providers also get information from their corporate partners. Product warranty information, credit-card transactions, Internet activity, and a host of other industry-specific data is combined to form a comprehensive picture of consumers. Chapter 3 discusses the notion of consumer privacy. Privacy comes into play here as well. For legal, ethical, and practical reasons, these vendors will not share the actual customer data that they have collected from their corporate partnerships. At least, they won’t share it at the individual customer level. Instead they use the data to analyze your customer base. They will produce a customized set of segments specific to your particular needs. In essence, they say things like, “I can’t tell you how I know, but I know that these customers are young affluent couples without children.”
Some of these third-party customer segmentation schemes are quite elaborate. They break your customer base up into hundreds of small segments. Each of these segments is quite uniform with respect to its demographic makeup. The best way to understand what segmentation is all about is to look at an example.
One segmentation scheme that's quite famous among marketers was developed by Claritas PRIZM, which is now owned by the Nielson company (of TV ratings fame). One of the endearing things about this particular scheme is that the company spent a lot of time naming these segments in a clever and informative way. You find segments with names like Money and Brains and White Picket Fences. My personal favorite has always been Shotguns and Pickups. You can check out these segments along with a brief description of each at www.claritas.com/MyBestSegments/Default.jsp#. Wikipedia lists them as well in its article "Claritas Prizm."
Third-party data does cost money. Over the years, though, it’s become cheaper and cheaper. It has also become more and more accurate and comprehensive. Good demographic data does contribute significantly to the effectiveness of your marketing campaigns. It’s worth your time to explore what’s available.
If You Have to Ask, You Can’t Afford It: Grouping Customers Using Income Data
Your marketing campaigns are intended to convince people to buy a product or service. Because that involves a financial transaction, understanding your customer’s financial situation is useful. Having some financial information, most commonly household income, helps you match your offers and messages with the means and needs of your customers.
Fish where the fish are: Understanding affordability
You have some products that are more expensive than others. It won’t do you much good to target middle-income families with information on a $200,000 Italian sports car. As you try to identify who might buy a particular product, it’s natural to ask a simple question: Who can afford it? If you have household income data in your marketing database, you can answer this question fairly easily.
A simple approach is just to look at who has bought the product in the past. Is there an income level below which very few customers bought?
Income data is normally reported in ranges rather than actual dollars. In other words, you won’t see a household with an income of $37,512. You’ll see that the household has an income in the $35K–40K range. This actually makes it quite convenient to look at how purchase behavior differs from one income band to the next.
If you find an income threshold below which purchases drop off, you can focus only on customers with incomes above the threshold. Marketers call this income qualifying an audience. It’s an extremely common approach to targeting, particularly when it comes to expensive or luxury items. You need to consider a couple things when establishing an income threshold.
Often the number of purchasers drops off a little at a time as you move down through the lower income bands. This means that your response rates drop as you include lower bands. You may eventually pass a break-even point where a few additional responses won’t pay for the additional mail costs. I talk more about evaluating campaign costs in Chapter 14.
For some low-cost channels, particularly e-mail, this break-even point isn’t an issue. But there is another kind of cost to consider: Mailing to people who can’t afford a particular product annoys them. You only get so many chances to talk to a customer before they tune you out. You’re far better off talking to these customers about a different product or approaching them with some sort of discount.
Speaking of discounts, there is a flip side to income qualification. You don’t want to be discounting your products unnecessarily. When marketing a discounted offer, you may want to keep high-income households out of the target audience. This is sort of a reverse income qualification. If they can afford full price, why start the conversation with a discount?
Champagne taste or beer budget: Price-sensitive customers
As I just said, when it comes to higher-income households, you may not want to lead off with a discount. But in some cases you may end up there. Some people just will not pay the initial asking price for anything … ever. For some people, being frugal is a matter of necessity. For others, it’s a matter of principle. For still others, it’s almost a matter of sport.
Just last week I saw a woman load up her purse with free cookies that were put out for kids on her way out of the grocery store. Judging from the luxury sedan that she climbed into, I was pretty sure she could have afforded to buy a box of cookies.
Whatever the reasons, there is a difference between a customer’s ability to afford full price and their willingness to pay it. Marketers often refer to price-sensitive customers as value oriented. These are customers who will wait for sales, discounts, or other deals before buying. The Black Friday people, who are in line at 3 a.m. on the morning after Thanksgiving, are classic value-oriented customers.
On the other end of the spectrum are customers who enjoy luxury goods and high service levels. They buy higher-end cars. They pay for first-class tickets. These consumers are variously referred to as service oriented or experience oriented.
When communicating with these experience-oriented customers, you don’t need to entice them with discounted offers. Often, you can actually entice them to pay extra for added services. Focusing your message on exclusivity, special treatment, or enhanced features of your higher-end products may actually resonate more with this group.
When trying to distinguish your value-oriented from your experience-oriented customers, household income certainly comes into play. The higher-end consumer does need to be able to afford the higher price tags. But past customer behavior is also important. Almost anyone will take a discount if one is offered. But looking for people who bought without a discount will help you to increase your profit margins.
Generation Gap: Grouping Customers by Age
You’ve heard of the Baby Boomers, Generation X, and the Greatest Generation. These are all examples of what sociologists call age cohorts or generational cohorts. Cohort is just a fancy word for group. These groups have been studied extensively by both academics and marketers. That research can be very useful to you in developing your marketing campaigns.
Understanding generational differences
The basic idea is that we are all influenced by the society we grew up in. Major events and cultural shifts help shape our attitudes. The Great Depression, World War II, the Cold War, and the Vietnam War are all examples of events that define generational cohorts.
People who grew up during the Great Depression tend to have attitudes toward frugality and self-reliance that reflect the scarcity of the times. Generation Xers’ attitudes toward authority are reflective of the tumultuous 60s and 70s. Their attitudes toward technology reflect the fact that they were around when the computer age took off.
There is a classic generation gap between my parents and me when it comes to banking. I don’t believe my parents have ever used an ATM machine. They go into a branch when they want to get cash or make a deposit. I, on the other hand, don’t believe I have been inside a bank branch in ten years.
Understanding this type of generational difference can help you communicate effectively with different audiences. For starters, it can help you to pick an effective marketing channel. Older and younger generations have very different attitudes toward and experience with technology. So called “new media” campaigns using text messaging or social media are probably not going to resonate (or even reach) older generations. But they are indispensable for younger ones.
You can tailor your marketing message toward different generations as well. Being relevant requires understanding the popular culture of each generation. If you own a music store, you aren’t going to help your business by pushing Lawrence Welk CDs to 20-somethings. And you aren’t going sell many gansta rap MP3s to baby boomers. Your communications need to match your inventory to the generations for whom it is relevant.
The generation gap and digital media
Generational differences are particularly critical when it comes to marketing through digital channels. When I was in junior high school, my family still had a party line phone that we shared with the neighbors. My first computer stored data on a cassette tape. Now a large and growing percentage of the population has never seen a cassette tape or a roto-dial phone. Not only have CDs replaced vinyl records, but CDs are now being replaced by smartphones as the primary medium for music.
One of the most important reasons to understand these differences is to gauge where consumers fall on the digital technology spectrum. In general, younger consumers have grown up in a highly advanced technological environment. When it comes to operating a smartphone, I’ll admit I’m totally outclassed by most ten-year-olds.
In order to communicate with younger consumers, using the so-called new media is critical. Having a robust web presence isn’t enough. Smartphones, tablets, and other mobile device are where young people are focused. Getting your marketing messages across depends on integrating with those devices. It also depends on making use of social media outlets like Facebook, Twitter, and Instagram.
In many ways, the process of buying ad placement in new media mirrors the way it’s done in traditional media like TV. You identify target audiences and place your advertisements on websites, search engines, and social media sites according to who’s viewing those sites. Communicating in the digital space is generally much less expensive than TV advertising. And there’s a whole lot of customer information available for these channels. This information enables you to target specific audiences very effectively.
There's a wealth of information describing in great detail the common traits of various generations available on the web. Countless studies have been done, and many of those publications are available for free or simply by registering on a website. I've found the Pew Research center to be a useful resource. Pew continues to research the characteristics associated with difference generational cohorts. You can find a wealth of articles on the subject at www.pewresearch.org/topics/generations-and-age/.
Grandkids? I Don’t Even Have Kids! Grouping Customers by Lifestage
Though useful, age groupings only tell you about broad similarities among your customers. Even within age groups, there is wide variation in people’s preferences and needs. Age is often combined with other variables to classify customers by their lifestage. In marketing, the notion of lifestage revolves around traditional rites of passage in our society. People grow up, go to college or straight to work, get married, have kids, send their kids to college, retire, and so on.
The simplest and most common lifestage groupings make use of age data combined with marital status and information about the presence and number of children in the household. Clearly, any particular age group will vary with respect to these other two factors — it will contain both single and married people as well as large families and childless couples.
Young newlyweds, families with young children, empty nesters, and grandparents are all examples of traditional lifestage segments. But you also have unmarried couples, single parents, and countless other variations that comprise a significant portion of your customer base.
The value of lifestage groupings to you as a marketer is largely, though not exclusively, related to the unique product needs of different segments. Young single mothers are probably not going to be your primary audience for a golf-club membership campaign, for example. But empty nesters might finally have the time to play golf regularly.
When you try to define a target audience for a particular product, look at the way previous purchases are distributed across lifestage segments. Don’t prejudge the groups that you think might be interested. Often you’ll find surprising niche groups that are buying a particular product more often than you would have expected. As I mention several times throughout this book, surprises lead to insights. It may not occur to you to market pacifiers and baby blankets to single women in their 30s without children. But after a look at your data, you may realize that this group of women gets invited to a lot of baby showers.
On the other hand, you may find lifestage segments that do not perform as well as you would have expected. Knowing that allows you to investigate why these segments are underpenetrated and take corrective action.
When developing or customizing a lifestage segmentation that fits your particular business, work with your advertising group. Lifestage segments are a fundamental part of how TV, radio, and other media are categorized. A particular cable station may be popular with young single women. A radio station may appeal mostly to retired men. Keeping your lifestage segments somewhat consistent with the way your advertising group buys its media lets you more easily integrate your campaigns with your company’s advertising.
What Would I Do With a Parka? Grouping Customers by Geography
Geographic data can be used in a surprising number of ways to enhance database marketing campaigns. As the title of this section suggests, some products are just not needed in certain regions. But you can also use geography to customize messages and determine appropriate campaign timing. Geography is also correlated with many demographic traits, which can be both helpful and potentially risky.
Geographic groupings
There are a variety of ways to group addresses together. The USPS uses zip codes to do just that. But the federal government also groups addresses together in a couple other ways. These groupings are related to government reporting and the census.
One common grouping is the Metropolitan Statistical Area, or MSA for short. An MSA is essentially a population center. They generally contain one or more fairly large cities. By fairly large, I mean at least 50,000 people. Pittsburgh and Minneapolis-St. Paul are examples of MSAs.
There is also a lower-level grouping called Micropolitan Statistical Areas that's used for government reporting. These areas are smaller than Metro areas — less than 50K but at least 10K in population. The delineations of both Metro and Micro areas is available on the U.S. Census website at www.census.gov/population/metro/.
Closely related to the Metropolitan Statistical Areas is the notion of a Designated Market Area or DMA. Be careful not to confuse this with the Direct Marketing Association. Unfortunately, the acronym DMA is used liberally to refer to both things.
Census results, economic conditions, and practically everything else that's tracked by the federal government is reported by MSA. This data is all publicly available. This means that you have at your disposal a fairly detailed demographic profile of each MSA. Because MSAs are closely aligned with TV and radio markets, these profiles are naturally of interest to marketers. The Census Bureau provides an array of data exploration tools at http://www.census.gov/main/www/access.html.
Census results are actually reported at a much lower-level grouping as well. They actually get reported at the census block level. You can think of this as a city block. It gets a little more subtle in rural areas. But these are quite small groups. There are more than 8 million census blocks in the U.S. This means that these groups typically only contain a few dozen households.
Because there is so much demographic data available on these geographic groups, many database marketers choose to have census group and MSA codes appended to their customer addresses. These codes also allow you to select audiences based on the market they live in. A variety of software packages do this. Melissadata provides geocoding services. If you’re using SAS software for analytics, it also offers geocoding capabilities. Or it may be simplest for you to have the vendor that’s doing your householding add geocodes to your file.
In the context of geography, a DMA is essentially a television market. These market definitions are managed (and trademarked) by Nielson. Because Nielson is the ultimate authority on TV viewership, television advertising purchases are broken up by DMA. Nielson will license data to you for a fee. To get started, visit www.nielsen.com/us/en/campaigns/dma-maps.html.
Knowing where to focus
Some products have regional appeal. This appeal can be related to a number of factors. Culture is a big one. A lot more grits are sold in the South than in the rest of the country.
That’s not to say that you can’t expand regional products to a wider audience. But the approach you take to marketing those products will definitely be different outside your core market. Expanding your market requires you to overcome an awareness gap. Everybody in the South knows what grits are. But in the Midwest, you need to answer the question “What’s a grit?” You need to be more informative about what grits are and how they should be served.
Regional appeal of a product can also be related to weather. Parkas aren’t going to sell in south Florida. After having lived in the South for many years, my wife and I recently spent a couple of years in Maine. Much to our surprise, none of the houses we looked at had central air conditioning. We quickly discovered why and are now happy to be Southerners again.
Geography itself can be a factor as well. Boats are not going to sell as well in the Great Plains as they are on the Gulf Coast, for example. Nor are surfboards going to be popular in the mountains of West Virginia.
It is important for you to let your data tell you where your products are selling. You may find unexpected geographic pockets where your products do quite well. This is especially true in cases where a product is associated with travel. Ski equipment is a good example. My wife has a closet full of ski equipment and she’s never lived anywhere near a mountain. And a product may have an unanticipated use. I associate moth balls with storing wool sweaters. But many people in warm climates use them to keep snakes out of their yards.
Sending your customers to the right place
Customizing your messages is another application of geographic data. It’s one thing to refer in your communication to “a location near you.” It’s quite another to actually tell your customer where that location is. The latter is far more helpful and consequently more effective.
Customizing is absolutely crucial if your business is adding new locations. Getting the word out about a new store is fundamentally about geographic targeting. By looking at your current customer base, you can come up with a profile of your typical customer. You then look for customers who meet that profile in the vicinity of the new store. In fact, the new location was probably chosen, at least in part, based on a similar analysis.
Geocoding software has gotten quite sophisticated over the years. It’s now quite easy to identify households that live within a given distance of a store location. This ability is apparent on many websites. Pizza delivery chains make good use of it in directing users to the store that serves their area. Many retailers make use of it to dynamically find which store may have a searched-for item in stock.
For businesses that don’t have the option of doing business purely online, it’s important to understand the geographic range of their customer base. Theme parks and restaurants are examples of businesses that require the customer to actually visit their location. To market these businesses effectively, you need to know how far people are willing to travel to do business with you.
Travel destinations typically group their customers according to how far away they live. A distinction is made between customers that are close enough to drive and those that need to fly. Understanding this distinction is important because of the significant cost difference between the two modes of travel. Customers in fly markets may need to have higher incomes in order to afford the trip. The drive-market customers may be close enough to visit frequently. The marketing messages that are most effective differ between the two groups.
Timing your communications
The lives of families with children revolve around school schedules. Winter break, spring break, prom, summer vacation, back-to-school time, and graduation come and go like clockwork. This calendar dictates the timing of vacations and purchases of everything from prom dresses to school supplies.
What does this have to do with geography, you ask? Quite a bit, as it turns out. I had a colleague a number of years ago who was quite knowledgeable about geography. We were working for a travel-related business that was quite dependent on vacation schedules. He pointed out to me that school calendars are not all the same. Depending on where people lived, school could begin anywhere from early August to mid-September.
This variation has a pattern — a geographic pattern. He essentially showed me that as you move from New England down the Eastern Seaboard, the first day of school gets earlier and earlier.
This actually turned out to be a very useful insight. By timing campaigns differently for different states, we were able to hit the sweet spot in our customers’ planning cycles and improve our response rates. We were also able to spread out the revenue. We started looking at spring break as five weeks long instead of one.
Many seasonal events vary by geography. Like school schedules, the arrival of winter varies geographically. And also like school schedules, winter’s arrival drives the demand for a specific set of products. By analyzing the timing of the demand for your products geographically, you can position your messages to be in market when they are most relevant.
Staying Out of Trouble: Some Legal Considerations
Customer groupings, or segments, are an indispensable tool in database marketing. But they sometimes require careful handling. Income and especially geographic data can prove problematic in some instances. The seriousness of the problems is heavily dependent on your industry. It’s important that you understand clearly which regulations apply to you.
Consult your attorney: Knowing your industry’s legal environment
The financial-services industry is heavily regulated with respect to data collection and usage. Several federal statutes related to credit reporting and fair lending apply not only to banks but to other lenders as well. It’s far beyond the scope of this book to address these regulations in any detail. It’s important for you to understand, however, that these regulations don’t just apply to banks.
Even if you don’t work for a bank, fair lending regulations apply to you whenever you extend an offer of credit. When you’re communicating to customers about payment plans, deferred payments, or co-branded credit cards, be sure to consult your attorney to ensure that your offers are compliant with federal regulations. The health care industry is also highly regulated. Consumer privacy with respect to medical records is jealously guarded by legislation and audit oversight. Practically any use of consumer data for marketing purposes in this industry should be run past your legal team.
There is also growing concern among consumers and legislators about the collection of geographic data from mobile devices like cellphones and tablets. The concerns stem partly from the fact that so many children are using these devices. As I’m writing this chapter, no legislation has been passed on this subject. But it has been proposed.
The regulatory environment related to database marketing in the U.S. is dynamic. You need to stay informed about and adapt to new legislation. In Chapter 19, I provide some resources that will help you do this. But you do need to get a legal opinion from your corporate counsel on these issues.
Watching where you tread: Potential data landmines
Chapter 6 talks about the notion of customer traits being related to one another. The technical term I used was correlation. Two traits are correlated if they tend to appear together. A simple example would be the presence of children and minivan ownership.
The customer groupings I talk about in this chapter are useful precisely to the extent that they correlate with customer purchase behavior. You target a specific lifestage segment with a message about your product because people in that segment tend to buy your product.
Some groups, particularly geographic groups, can be highly correlated with income and with race and ethnicity. By simply using an income level cutoff to target an audience, you may inadvertently be introducing a racially discriminatory selection criterion. Similarly, targeting specific geographies can lead to the same result. The smaller the geographic grouping, the more correlated it is with race. In other words, individual zip codes are more correlated with race than greater metropolitan areas are. Banks are very careful about this issue because of legislation surrounding fair lending practices.
It’s important to be familiar with the entire demographic profile of your audience to avoid introducing racial bias into your campaign. Even the appearance of racial discrimination, intended or not, can lead to legal problems.