When thinking about customer behavior, it's easy to think about customers at the aggregate level because this is how data are presented to us. For example, we rarely analyze the path "Scot" took through website because there are just too many visitors. Instead, we look at the top level data, or a few different segments that might originate from a specific channel or campaign. While these numbers are helpful, marketers miss opportunities because the data blur the lines that make individuals individuals.
An individual is like a good meal made of fine ingredients. Most ingredients are used in multiple dishes, and just because you know the ingredient, doesn't mean you know what dish it is. To assume you do leads to big mistakes. What if the ingredient is milk? Could be cereal, ice cream, pancakes, cookies, fried chicken, cheese, or something else entirely. There is no way to know based on one piece of information (i.e. milk), yet this is how we often segment our customers.
We need to find ways to segment customers based on multiple criteria, and doing this right means consolidating data across multiple data sources. We obviously don't just interact with customers in one place; we do it on social sites, ecommerce platforms, search engines, etc. If we limit ourselves to one data source, we don't get the whole picture and we're stuck guessing whether we're eating ice cream or cheese. Consolidating data gives us more ingredients, and more information about our customers. We can know their history of items purchased, products "liked," pages viewed and emails opened. This gives us a powerful look into the demographics and behaviors that shape our customers and make them unique individuals.
Once we have the full picture, we can eliminate costly mistakes. For example, take a retailer that just uses a signup form to define email marketing segments. I mark that I'm a "male" and begin to get emails targeted at men. Good job - the retailer is on the right track. However, I then go and purchase three straight products designed for women, which is not inconceivable given that my mom, mother-in-law, sister and sister-in-law all have birthdays in March (rough month, indeed). Wouldn't it make sense to start sending me promotions that are geared towards women? Better yet, how about recognizing that I'm a guy, yet seem to buy female products? It's not a leap to assume I'm buying gifts, and then market gift offers to me.
Combining demographic data with purchase history expands the opportunities for me to understand my individual customers. It moves me out of mass marketing and into personalization. It moves me from being annoying to my customers to being relevant. It moves from guessing whether I'm about to eat ice cream or cauliflower casserole.