Ask a predictive marketer: how is predictive marketing different from recommendation engines?

Nov, 11 2013

Knowing the right products recommendations for the visitors on your website is a critical piece of the marketing puzzle. If you can use predictive marketing to personalize the browsing experience by showing the right products on the landing page, on the product page, or on the cart page – these products are bound to help you get engagement from your customers and become more profitable, because you are presenting goods they’re interested in, front and center.

We've got you covered

Predictive marketing by AgilOne can help you with your next sell, cross sell and upsell recommendations, served up via AgilOne’s convenient plug and play API solution.

It’s important to remember that although critical, this is merely one piece of the marketing intelligence puzzle. Your data is very valuable, and AgilOne has the right tools to help you extract much more mileage from it than simply finding the right recommendations. AgilOne can give you a 360 degree view of your customers – based on information our program collects, cleans, dedupes and link from all customer touch points to a single customer ID.

True marketing intelligence should enable you to really understand your customers, predict their actions, customize your messaging, and touch them with the right offers, at the right time!  For this to happen, you need more than just recommendations. You also need clustering (segmentation algorithms) and propensity models (prediction algorithms) to target the right customers with your recommendations and offers.

Propensity models

Propensity models make predictions about customers. There are three types of propensity models; the propensity to buy model, the propensity to churn model and the propensity to unsubscribe model.

The propensity to buy model tells you which customers are ready to make their purchase, thus allowing the marketer to effectively target the correct customers. Moreover, once you know which customers are ready to purchase and which ones still need a little more incentive, you can decided to incentivize the “maybes” by offering them a discount or special offer in the email campaign. Those likely to buy won't need high discounts (you can stop cannibalizing your margin) while the customers who are not likely to buy may need a more aggressive offer, thereby bringing you incremental revenue.

The propensity to churn model tells you which active customers are at risk to drop off, so you know which high value, at risk customers to put on your watch list and create a campaign to retain them.

The propensity to unsubscribe model tells you which customers not to touch: if there are high value customers you are at risk of losing to unsubscribe, you need to find other ways to reaching out to them that are not by email. Social media or direct mail could be a few options.

With the same token, our propensity to engage model seeks out the customers that are waiting for your offers and products.

Clustering algorithms

Our clustering algorithms also find what different personas and customer profiles you have.

The product based clustering algorithm discovers what different groupings of products people have bought from. Brand based clustering tells you what brands people like. Now you know who to pitch what products and what brands to. When a brand releases new products – you’ll be likely to know exactly who to target with those new brands.

Behavioral clustering informs you of how people behave while purchasing: do they use the web site or the call center? Are they discount addicts? How frequently do they buy? How much do they spend? How much time will go by before they purchase again? This algorithm helps set the right tone while contacting the customer. For instance, customers who regularly buy with low sized orders might react well to offers like 'Double points if you spend more than $100'.

All of AgilOne’s models are automatically refreshed daily and each customer's propensities as well as cluster information is updated, so as they purchase more, browse different products you can stay close to them; and act when they start to pull away!

For a complete overview of all our predictive analytics models, including recommendations, click here to read: The Definitive Guide to Predictive Analytics Models for Marketing.

Want to learn more about AgilOne’s predictive marketing platform? We’d love to tell you more.

AgilOne - Predictive Marketing