5 Common Customer Segmentation Mistakes to Avoid
February 4, 2009 By Joe Boland, Assistant Editor, Target Marketing
|
The best way for marketers to make the most of their customer relationships is to understand their customers' needs, wants, values. That’s precisely why customer segmentation is so vital—it provides valuable information about customers so marketers can furnish stronger, more targeted offers.
The trouble is customer segmentation comes with its challenges, and marketers often trip up during the process. Here, industry professionals discuss common mistakes marketers should avoid when performing customer segmentation.
1. Not accounting for duplicate names and addresses. Recency, frequency, monetary segmentation is a standard, but often, operating systems have a lot of tolerance for duplicate names and addresses, points out Mary Ann Kleinfelter, vice president of marketing at L-com, a B-to-B multichannel marketer of connectivity products in North Andover, Mass. “Therefore,” she says, “a customer might appear twice—once with a recent order and once with an earlier order at slightly different names and/or addresses.” This can lead to a customer segmented with an early order receiving mail that says “We miss you,” when that person has recently ordered.
2. Segmenting arbitrarily instead of strategically. “Oftentimes, marketers are not strategic in deciding how to segment,” says Gary Hennerberg, founder of Hennerberg Group, a Colleyville, Texas-based direct marketing consultancy. By that, Hennerberg means, for example, some marketers segment based on frequency without looking at dollar value, which can provide misleading results. For instance, higher value customers may be contacted more frequently, resulting in more frequent purchases; that’s why their frequency is higher.
Hennerberg suggests dividing customers into three groups—your lower third, middle third and higher third in terms of monetary value—and segment from there. That doesn’t necessarily mean a third of your customers, he cautions, but possibly a third of sales volume or some other measure that's important to your company's goals.
3. Emphasizing demographics or attitudinal responses. Customer demographics and attitudes are useful to marketers, but the best indicator of future behavior is past behavior. Too frequently, marketers rely on demographics and attitudes during segmentation, says Randy Erdahl, co-founder and president of the Eden Prairie, Minn.-based database marketing analytics firm Decision Intelligence, when they should focus on behavioral attributes in conjunction with demographics. For example, Erdahl offers, marketers should zero in on metrics such as how many times customer buy, what types of things they buy, when they buy and what channels they use to glean more information as to how customers react to your marketing messages.
Michelle FauntLeRoy, assistant vice president of segmentation, data product management at data management provider Nielsen Claritas, agrees. “A lot of people rely on household level data,” she says. “While it gives the highest performance in terms of being able to identify what different households are like, there’s a lot of holes in the data. … The smarter way to think about it, or the more productive way, is to think, ‘What kind of performance am I getting?’”
4. Getting too complex. “A lot of companies start out by trying to create a custom or really complex segmentation model or scheme and can’t implement it,” says FauntLeRoy. Marketers should begin with a few key metrics that are critical to their customers and build from there.
5. Lacking consistency in defining segments. Going through the process, metrics are not always consistently defined, which can provide inconsistent results. For example, says Kleinfelter, one-time buyers often get mixed up by firms. There are two types of one-time buyers: recent ones who have not had a chance to buy again and one-time buyers who bought a while ago and have not purchases since. These are two very different segments, and they require distinct marketing approaches.
Erdahl agrees that consistency is key, adding that some marketers have too many metrics they’re looking at and don’t stay focused on which ones they track and apply to programs. As a result, their efforts lack traction because the organization is running in too many directions.
No matter how marketers segment their customers, all the experts agree that it’s critical to make sure the data is clean and the results are analyzed from both a short- and long-term perspective to achieve positive progress.
The trouble is customer segmentation comes with its challenges, and marketers often trip up during the process. Here, industry professionals discuss common mistakes marketers should avoid when performing customer segmentation.
1. Not accounting for duplicate names and addresses. Recency, frequency, monetary segmentation is a standard, but often, operating systems have a lot of tolerance for duplicate names and addresses, points out Mary Ann Kleinfelter, vice president of marketing at L-com, a B-to-B multichannel marketer of connectivity products in North Andover, Mass. “Therefore,” she says, “a customer might appear twice—once with a recent order and once with an earlier order at slightly different names and/or addresses.” This can lead to a customer segmented with an early order receiving mail that says “We miss you,” when that person has recently ordered.
2. Segmenting arbitrarily instead of strategically. “Oftentimes, marketers are not strategic in deciding how to segment,” says Gary Hennerberg, founder of Hennerberg Group, a Colleyville, Texas-based direct marketing consultancy. By that, Hennerberg means, for example, some marketers segment based on frequency without looking at dollar value, which can provide misleading results. For instance, higher value customers may be contacted more frequently, resulting in more frequent purchases; that’s why their frequency is higher.
Hennerberg suggests dividing customers into three groups—your lower third, middle third and higher third in terms of monetary value—and segment from there. That doesn’t necessarily mean a third of your customers, he cautions, but possibly a third of sales volume or some other measure that's important to your company's goals.
3. Emphasizing demographics or attitudinal responses. Customer demographics and attitudes are useful to marketers, but the best indicator of future behavior is past behavior. Too frequently, marketers rely on demographics and attitudes during segmentation, says Randy Erdahl, co-founder and president of the Eden Prairie, Minn.-based database marketing analytics firm Decision Intelligence, when they should focus on behavioral attributes in conjunction with demographics. For example, Erdahl offers, marketers should zero in on metrics such as how many times customer buy, what types of things they buy, when they buy and what channels they use to glean more information as to how customers react to your marketing messages.
Michelle FauntLeRoy, assistant vice president of segmentation, data product management at data management provider Nielsen Claritas, agrees. “A lot of people rely on household level data,” she says. “While it gives the highest performance in terms of being able to identify what different households are like, there’s a lot of holes in the data. … The smarter way to think about it, or the more productive way, is to think, ‘What kind of performance am I getting?’”
4. Getting too complex. “A lot of companies start out by trying to create a custom or really complex segmentation model or scheme and can’t implement it,” says FauntLeRoy. Marketers should begin with a few key metrics that are critical to their customers and build from there.
5. Lacking consistency in defining segments. Going through the process, metrics are not always consistently defined, which can provide inconsistent results. For example, says Kleinfelter, one-time buyers often get mixed up by firms. There are two types of one-time buyers: recent ones who have not had a chance to buy again and one-time buyers who bought a while ago and have not purchases since. These are two very different segments, and they require distinct marketing approaches.
Erdahl agrees that consistency is key, adding that some marketers have too many metrics they’re looking at and don’t stay focused on which ones they track and apply to programs. As a result, their efforts lack traction because the organization is running in too many directions.
No matter how marketers segment their customers, all the experts agree that it’s critical to make sure the data is clean and the results are analyzed from both a short- and long-term perspective to achieve positive progress.
Companies Mentioned:
 |
|
|
The Business of Database Marketing