How to use AI to find cross-selling opportunities

 

Suppose you go and buy green tea on Amazon - Amazon might recommend a mug and a kettle to go with it. If you buy a laptop at MediaMarkt, the shop may recommend a laptop stand and a carrying case. If you have ever been asked "Do you want fries with that?", then you are already familiar with the idea of cross-selling. Crossselling is simply when you sell an item alongside another item that a customer was already planning to buy.

The complexity of cross-selling

However, it is not as simple as it seems at first glance, especially when you are dealing with many products. How do you know which is the best product to recommend, and to whom should you recommend it? Should you recommend a certain product or service, or would it be a turn-off? Analysing those large amounts of data to optimise cross-selling opportunities is what a AI platform like Trendskout can help.

Calculating the added value of cross-selling

If you need further convincing of the added value of cross-selling, consider this finding from McKinsey:

"35 per cent of what consumers buy on Amazon and 75 per cent of what they watch on Netflix comes from product recommendations based on such algorithms."

These are incredible percentages, and they can be a make-or-break difference for smaller e-commerce shops. There is almost nothing that has more impact on your (e-commerce) shop than optimising cross-sales.

No-Code AI to find cross-selling opportunities

Let's use Trendskout to predict cross-selling opportunities based on data from insurance companies. In a fictional dataset, there are customers with health insurance policies, and we want to know whether a new customer would be interested in car insurance.

The information in the dataset includes demographic info (such as customer's age, gender and region), vehicle info (such as vehicle age and damage), and policy info (such as customer's premium, purchase channel and duration).

Our KPI, or the value we want to predict, is called "Response." This column has the value "1" if a customer is interested in insurance and the value "0" if the customer is not interested in insurance.

By linking this dataset to Trendskout, we can evaluate how each customer characteristic affects the likelihood of them being interested in car insurance.

We will find that the length of time the customer has been with the company is extremely important in predicting how likely it is that the cross-sale will be a success. The longer a customer has been with the company, the more likely he is to be interested in car insurance.

In order of decreasing importance, we see that customer age, annual premium and sales channel are also predictive. In particular, older customers (who are likely to be more affluent and risk-averse) and customers with lower premiums are more interested in car insurance.

Cross-sale scaling with Trendskout

Everyone agrees: crossal is an absolute asset for many companies. To keep everything manageable, artificial intelligence is an absolute asset. The problem that many small, medium-sized and sometimes even large companies are facing is a lack of knowledge. You can look for a data scientist. (Just look on the job platforms how many companies are begging for a data scientist). Or you can opt for a plug and play platform.

By listening carefully to the market, we developed a flexible and efficient solution. With the Trendskout platform you connect your data (from your ERP, CRM, an Excel,...) you release numerous AI and ML models and you send the output wherever you want. Without the need for an internal data scientist. Book a personalised demo now and discover the possibilities for your company.

How to discover hidden sales opportunities in your data with Sales AI

Subscribe to our monthly newsletter