Creeping Personalization 101

Gathering data is a crucial part of any successful web business. In the first place, data about visitor and customer behavior forms the rudimentary foundation of understanding return on investment.

But data can also be used to improve the customer’s experience. The data can be collected incrementally over time, constantly improving the experience of interacting with a web business. This process continually increases the value that you can provide to customers, and optimizes the profitability of your relationship with each customer. Because the process is gradual, it is sometimes called creeping personalization.

Creeping personalization is achieved through the gathering of both explicit and implicit data. Explicit data is the information that customers give through surveys, forms, and registration. It’s good data, but it isn’t always reliable. Besides, you don’t want to pester your customers with forms if you can avoid it.

Implicit data is the data that is gathered in the back end through the recorded actions of customers through your advertising and on your site. Using tools like cookies (make sure you have a privacy policy!), you can find out what type of customer someone is by the way he or she arrives at and navigates on your site.

When planning a data strategy, however, it doesn’t make sense to start with the technology. It’s best to start with a solid understanding of your targets, and how you want your online business to serve those individual segments based on the information that is gathered. Once that strategy is in place, you can address the technology needed to reach your goals.

First, you must determine customer target segments. Often it helps to use qualitative or quantitative research to develop these. Naturally, the data that you eventually gather will help you refine these targets, but you have to start somewhere.

Second, determine the characteristics that classify customers into each segment. On a car-buying site, for example, a qualified buyer might be someone who fits a certain demographic profile, shows interest in a vehicle’s price, and signs up for a test drive. A list of target segments with a corresponding row of these characteristics forms a grid that will organize your data for analysis and action.

Third, decide how you will gather the data you need. This is a cross-disciplinary task, requiring input from technology, creative, and marketing folks. Gathering information from customers is an art, because it is easy to overwhelm customers or scare them off with too many questions. Remember, a good strategy combines explicit and implicit data, and collects it over time.

There are a number of ways to crunch the data gathered from customer activity through advertising and on your site.

Clustering is a statistical process that groups people by their behavior. Collaborative filtering is a constant comparison of activities that allows sites to offer personalized merchandising and offers. No matter what the case, data will probably come from a variety of sources, and integration will be a major challenge.

When you understand targets and know how and what data will be gathered, you can start to craft the varying experiences that will be offered to different customer targets.

Perhaps qualified buyers will be offered a brochure and a video, while other visitors might get just an online newsletter. The decisions as to which targets qualify for certain offers must be made carefully.

Personalization requires enormous resources in money, expertise, and time. But it is not just a technology issue. Solid marketing and business principles should guide the investment and deployment of personalization in the frenzied competition to offer customer value.

Related reading