Product Recommendation: 3 Crucial Missions to Pursue

Product Recommendation: 3 Crucial Missions to Pursue

March 29, 2016

Disruptive is a fancy word in today’s ecosystem, but we should keep learning from the best practices of our ancestors. This is a valid statement for e-commerce as well, which still has to learn so much from brick and mortar. The most crucial best practice that we still miss in the e-commerce world is the “sales assistant” in physical stores. They know the old customers and they have the chance to get familiar with the newcomers. Similarly, they are always well educated and guided about the products so they have the chance to guide the visitors during their conversion journey. Once defined like this, a sales assistant would be quite handy in e-commerce, wouldn’t she/he?

Product recommendation tools are the e-sales assistants, who help us provide the right offer to the right person at the right time and optimize conversion rates. These tools continuously observe our customers, learn their needs and select the most appropriate product for them. This includes three main activities.

1 – Get Familiar with Your Products and Customers

This is actually the first activity expected from a sales manager: Learn what we sell, how we sell and to whom we sell. This is really a complex process and needs time for an assistant to learn them all. On the other hand, when this is left to a machine on the e-commerce side, the product recommendation tools start to learn both the products side and the customer side.

A sales assistant has the chance to observe a customer when they enter the store, talk to them to aggregate more information about them and behave accordingly.

To benefit from this best practice in e-commerce world, product recommendation algorithms use all available data such as the source our customers are coming from (Direct, Organic or Paid), the location information gathered from IP address, the device type (mobile, Desktop or Tablet), the time they visit the site and etc in the first row. In addition, every single activity of a customer on our sites such as viewing a product, investigating product descriptions and etc. These parameters are quite beneficial especially for getting familiar with the anonymous visitors.

2 – Personalised Service

Any piece of information is enough for a sales assistant to start a personalized conversation and generate a baseline to start a journey for the customers. And once the assistant aggregates enough information, the communication with the customer becomes more personalized based on the current needs and expectations of the customer.

In the e-commerce side, once the product recommendation is in use, the overall e-commerce experience provided to every single customer can be personalized.

This personalization uses every possible opportunity to recommend products based on the behavioral analytics underlying. Customized home-pages “Selections”, categories “Trending Products” lists, “You May Also Like” widgets in product detail pages are the most common examples of the personalization. The main idea behind using these widgets is to increase customer engagement, enable them to spend more time on our site and view more products. Although we have not mentioned conversion yet, research on customer engagement in e-commerce shows that a 5% increase in engagement could result in 95% in revenue. Now, this should sound interesting.

Again to highlight, some best practices that should be copied from sales assistant actions are impressive when we personalize “search result” pages and “product not found” pages.

A successful sales assistant never only says “sorry we do not have what you are looking for”? and they always add “but look we have these ones you might get interested” or “our customers who are looking for that product actually buy this one since this is much better”.

But you can see lots of e-commerce “empty search page results” which in practice says “You can leave now, I can not sell any products to you” which definitely kills the conversion rates.

Kohl's Empty Site Search Result

As a friendly advice, personalized product recommendations in an “empty search page results” results in a minimum of 15% CTR.

Similarly, the fancy “not found pages” are much better than “404” pages but again personalization of these nice fancy pages with selected products would keep our customers on-site.

Fancy Not Found Page

3 – Keep Tracking

And the final step in the way of a successful sales assistant is to learn continuously, based on your current and previous conversations with the customers. How they respond to your questions and alternative product offers are the key findings of an assistant that might be the most valuable opportunity in converting the visitor.

Definitely, this is a built-in function of product recommendation tools with using the positive and negative feedback which is, in other words, learning from the feedback of your customers to your product recommendations.


    Leave a Reply

    Related Stories

    Big Data 5 Ways it Can Assist Merchants
    February 19, 2016

    Big Data: 5 Ways it Can Assist Merchants