The scientific way of using customer data to boost sales

At Retention Science, data is king. Customer data serves as the cornerstone of all our campaigns, which is why we spend a great deal of resources collecting and analyzing as many metrics and data points as we can. To give you a better idea of what we do with all that information, below is an explanation of the data we use to power our science and get you results:

Historical Data

We begin by uploading your sales history from the last six months to four years – the longer the better. We examine this data and build profiles based on your customers’ previous behavior. We look at data points like consumer spending amounts, time of purchase, discount usage and product selection, and then we use these queries as the foundation of each customer profile.

Recent Customer Data

We supplement the historical sales data with an incremental feed of the same data fields updated every 4-24 hours. These fields form the basis of our campaign recommendations, and they allow you to get recent data about your customers. This also enables you to keep up with their evolving market interests and needs, so you can tune your campaigns going forward. For example, our marketing engine can determine if a customer is a bargain or low-cost shopper or perhaps motivated more by quick delivery.


Behavioral Data

We also observe how your customers interact with and behave on your website. Our models track various metrics including length of time on your site, page and product views, frequency of visits, and items added to cart. This behavioral data is the strongest component of our probability-to-purchase models, because it predicts with high accuracy a person’s engagement level. Once we identify highly engaged individuals, we can make recommendations about the best way to treat them.

We work with our clients and use a variety of tools such as JavaScript Tag Integration, Tealium, and Google Tag Manager to track website traffic. We also collect data from email campaigns by integrating with our client’s email marketing vendor.

Demographic Data

We supplement the proprietary customer information our client gives us with public demographic data. These data sources add further detail to each customer profile and give Retention Science more precise knowledge on customer preferences. For example, based on a customer’s zip code, we can look up average household income, commute style, and type of home.

Custom Client Data

Every business has different signals and market data points. Retention Science surveys your market and industry through social media and other networks to gain more data about your customers. This enables us to flesh out additional information, such as their hobbies and interests, and give you a deeper understanding of the inner working of your customers.

For example, many businesses have customer loyalty programs. We will look at data points such as if a person redeems their rewards or do the regularly interact with loyalty program. Another example is a telecom operator, and for them, the number of minutes used every month is an important data point.

How all the data comes together

After Retention Science has gathered and aggregated all of this data, it’s run through a predictive marketing platform to deliver 1-to-1 multi-channel campaigns with customized incentives and recommendations, delivered at the right time.


As a result of this, our clients have seen up to a 125% increase in on-site activity and a 78% rise in total revenue. Through using data science to predict customer behavior, Retention Science’s award-winning platform has proven to successfully reduce churn and extend the customer lifecycle.