Customers who last purchased 13-60+ months ago are big-time fun for folks who mine databases. If you have direct mail, catalog, online marketing, or e-mail marketing programs, consider use of the following attributes.
Attribute #1 = Returns: Customers who return at least sixty percent of the merchandise they purchased are likely to return merchandise in the future. If you're looking to trim marketing expense, here's a place to do so without negatively impacting profitability.
Attribute #2 = Fulfillment: Customers who failed to receive items in their last order are often less likely to order in the future than are customers who receive what they wanted to purchase.
Attribute #3 = Geography: Hillstrom's Zip Code Forensics indicate that about a quarter of the United States feature customers who spend twice as much as the national average. Use zip code modeling to your advantage!
Attribute #4 = Channel: Customers who order from catalogs over the telephone "age slowly", meaning these customers can be profitably targeted for a long time. Customers who order online age faster, meaning that after "x" months they become unresponsive. Customers who order via stores age very fast, becoming unresponsive soon after a purchase.
Attribute #5 = Items: Customers ordering multiple items tend to be more responsive in the future than customers ordering just a single item.
Attribute #6 = Price Point: Customers ordering expensive items tend to be more responsive in the future than customers ordering less-expensive items.
Attribute #7 = Social Media: Increasingly, we observe that customers who respond due to social media are "buying because of the buzz". Segment and learn!
Attribute #8 = Merchandise: Identify products that have long repurchase cycles. Every company has a diverse assortment of merchandise, with some products leading to repurchase cycles that are more beneficial to the brand.
Attribute #9 = Promos / Free Shipping: Closely monitor if these customers buy again, and observe whether they will ever buy merchandise at full price. You are in business to generate profit, not to give goodies away to customers.
Attribute #10 = Gross Margin: Always segment customers who buy products that have healthy gross margins. Classify these customers differently than those who buy products with razor-thin margins.
Attribute #11 = Online Visitation: Customers who last purchased three years ago, but visited your website in the past week represent an opportunity!
Attribute #12 = Referral Activity: Among customers who visited your website recently, infer causality from the referring URL.
Attribute #13 = Lifetime Activity: Lapsed customers fall into two groups --- those with a rich purchase history, and those without. Guess which customer is more likely to purchase again?
Attribute #14 = Multiple Merchandise Divisions: Customers who embrace a diversity of merchandise tend to be more valuable than customers who focus on just one product line.
Attribute #15 = Day Of Week: Monitor the day of week when a customer last purchased. Online and telephone customers are frequently more valuable if the last purchase happened on a Monday or Tuesday. Retail customers are sometimes more valuable when the last purchase happened late in the week.
Attribute #16 = Employee ID: Many companies know that a small subset of employees do a nice job of treating customers well. Conversely, some employees anger customers. Identify the employees who polarize customer repurchase activity, and use it as a segmentation variable.
Attribute #17 = Expedited Shipping: Identify customers who, in non-holiday periods, pay to receive merchandise fast! This can be a positive or negative, depending upon the circumstances of the request.
Attribute #18 = Open To Buy: If you have proprietary credit, then you know that customers lapse when they consume the majority of their credit limit. Pay close attention to customers who are within ten percent of their credit limit.
Attribute #19 = Hits: If you are able to track hits against competitor housefiles, you know this variable identifies customers who are responsive.
Attribute #20 = Tender Type: Visa, Master Card, Amex, you name it, each tender type used impacts future activity.
Attribute #21 = Seasonality: Pay close attention to holiday shoppers, who are frequently unresponsive in non-holiday periods. Conversely, customers who last ordered during the same season we're in experience an increase in responsiveness.
Attribute #22 = Visit Depth: Among customers who recently visited your website, segment those who got deep into the site, vs. those who only visited a landing page or the homepage.
Attribute #23 = Titles: If you are a catalog brand, you understand the importance of having multiple titles to serve the customer. Customers who respond to multiple titles behave differently than customers who respond to only one title.
Attribute #24 = Outlet Site: If you house your discount channel under a separate URL, you know that the customers who visit different URLs behave differently than customers who visit the main URL.
Attribute #25 = Sales Tax: Monitor whether customers who are required to pay sales tax have reduced repurchase activity. In most cases, sales tax does not have a big negative impact, but it is worth monitoring.
Attribute #26 = Product Cycle: The customer who buys a blu-ray DVD player is fundamentally different than the customer who purchases a 480p DVD player.
Attribute #27 = Want vs. Need: The customer who purchases product out of need is different than the customer who buys a fashion-based item.
Attribute #28 = Demographics: Demographics play a much smaller role in data mining than most folks realize. We're trained to believe that the 40-44 year old female head of household with annual household income of $100,000+ matters --- sure, she matters, but not because she is 40-44 years old with income of more than $100,000. Some demographic variables do matter --- incidence of kids for merchandisers of kids product, for example.
Attribute #29 = Change of Address: For some businesses, this is a big deal --- sellers of home products know this matters.
Attribute #30 = Dwelling Type: It's not hard to obtain data that indicates the type of home the customer lives in. A customer living in a high-rise condo is different than a customer living in a single family home.
Attribute #31 = Stores: If you have a retail business, you know that a customer spending $167 at each of three stores is more valuable than a customer spending $500 at just one store. You also know that certain stores drive more loyalty than others.
Attribute #32 = E-Mail Subscriber: If you have two lapsed customers, one who subscribes to e-mail marketing and one who does not, you know that the former is more interested in your brand than the latter.
Attribute #33 = RSS Subscriber: In particular, RSS subscribers of various product lines warrant considerable attention!
Attribute #34 = Blog Subscriber: Many catalogers and online marketers host blogs. Customers subscribing to blogs are not necessarily more profitable, but are certainly more engaged.
Attribute #35 = Source of Acquisition: Pay attention to customers acquired via different online sources, list rentals, and co-ops. Always remember that Google is a channel, separate from all other online activities.
Attribute #36 = Amazon: If you have an Amazon store, you're probably aware that customers ordering from Amazon are less likely to order in the future than are customers ordering via your own e-commerce website. As long as we're going down this path, segment customers from affiliates as well, as these customers buy for different reasons.
Attribute #37 = Brands vs. Proprietary Product: When you sell branded merchandise and merchandise you sources yourself, you wind up with customers who have different interests. This can be a telling variable, all companies have different relationships here.
Attribute #38 = Replenishment Merchandise: Product that is replenished on a period basis, or product supportive of continuity programs yield customers with different subsequent behavior.
Attribute #39 = Auction Behavior: For those of you who have auction sites, you know that customers visiting these sites are fundamentally different than are customers looking to buy merchandise at a fixed price.
Attribute #40 = Upsell / Cross-sell: Brands that eek out an extra $25 of unanticipated merchandise from a customer can tag these buyers, as their future behavior is likely to be different.
Attribute #41 = User Generated Reviews: Customers who also aid in the process of selling merchandise are frequently more engaged, and can generate indirect profitability via influencing other consumers. Pay attention to these buyers!
Attribute #42 = Blogoshpere: Tag customers who visit the website via various blogs or social media websites. These customers have a different relationship with us than do customers we outbound market to.
Attribute #43 = Buy Online, Pickup In Store: For those of us managing multiple physical channels, we know that connections between physical channels are important.
Your turn --- leave a comment, stating attributes that are important to your business!
This is maybe a year old, +/-, and is worth a year of consideration if you sell apparel: https://www.bloomberg.com/graphics/2018-death-o...
Look at the first four rows of our life table (values of 0/1/2/3). These are the first 12-15 weeks after a customer buys for the firs...
If you don't like geeky math, please skip this post, because I am about to show you how the sausage is made! I have eight variables in...
You probably run Life Tables for your customer file, right? Right? They've been around forever ( click here for a reference f...