Product Recommendation Effect on Black Friday Weekend

Black Friday weekend is the most important sales weekend for many eCommerce stores.

Many invest in technology to help bolster sales. This year, in addition to
looking at shopping cart abandonment, we took a look at how product recommendations helped eCommerce stores across three key metrics.

Specifically, we wanted to know

  • Intent - Did recommendations effect the number of items customers placed in carts?
  • Conversion Rate - Did personalized product recommendations have an effect on purchase rate?
  • AOV - Did product recommendations have an effect on the total number of items purchased.

The Data: Where we got it and how we cleaned it

Our goal is to understand how product recommendations helped eCommerce stores.


We know that product recommendations work generally. As Wharton professor Kartik Hosanagar states, "We know these recommendations have a pretty significant impact on consumer choice. For example, at Amazon, they drive anywhere from a quarter to a third of the choices that consumers make."

The first step was gathering data. We aggregated session data from customers who use Barilliance's product recommendation engine.

Our goal is to understand how product recommendations helped eCommerce stores.

Product Recommendations Effect on Items in Cart

Our first question was whether seeing product recommendations effected the number of items customers placed in their cart.

Product recommendation effect on purchase items

Product recommendations improved the average number of items placed in the cart by an impressive 68.14%. Customers who were either not shown product recommendations or elected not to click product recommendations placed 2.21 items in their cart.

Conversely, those who engaged with product recommendations entered as many as 3.9 items.

As expected, the biggest improvement came from seeing 0 product recommendations to clicking 1.

Seeing a single product recommendation improved the average number of items added to the cart by 13.94%. Interestingly, clicking on a second product recommendation improved the number of items by a similar amount 13.11%, before the marginal improvement dropped off dramatically. Clicking on the third product recommendation still had a massive positive effect over simply clicking two, but at 7.29% increase is just more than half the improvement of clicking on a single recommendation had.

Product Recommendations Effect on Conversion Rate

A more important question is whether product recommendations had a direct effect on conversion rate.

For the purpose of this study, we define conversion rate as the percent of customers who completed their purchase. Did accurate, relevant product recommendations improve the chance that customers completed their order?

Product recommendation effect on conversion rates

The answer is a resounding yes.

Product recommendations increased conversion rates by as much as 320%. Customers who did not click on recommendations completed their purchase 4.95% of the time.

That number shot up by 74% if they clicked on a single recommendation, to 8.63%. Looking at the data, you see a clear improvement in conversion rate until the customer engages with 7 or more recommendations.

At that level, the effect on conversions plateau. Still, the cumulative improvement is impressive. Stores that are able to engage customers with multiple product recommendations close 3x as many carts as those that fail.

Product Recommendations Effect on Items Purchased

We see that product recommendations have a tremendous effect on both the items customers place in their carts and their chance of completing the order.

In other words, we know recommendations improve intent to purchase, and conversion rate. Our final question is whether product recommendations improve the average order volume.

To obtain this data, we filtered out all the sessions that did not purchase, and compared the final number of items in the cart by the number of recommendations clicked.

Product recommendations effect on items purchased

Given our previous results, we expected a dramatic improvement in the number of items purchased for customers who viewed multiple product recommendations.This was proven true.

Clicking a single product recommendation improved the average number of items purchased by 25%.

In aggregate, product recommendations improved the number of items purchased by as much as 113%, more than doubling the AOV of stores that did not show product recommendations!

Compared to items placed in cart and conversion rate, the effect on items purchased had significantly more variance.

One explanation is that customers still removed items from their cart before completing their purchase. This would shrink the total effect, increasing the expected variance we would see in the data.

Next Steps

Product recommendations work.

They increase the number of items customers place in their cart before purchase by 68.14%, they increase the percentage of customers complete their purchase by 320%, and increase the average order volume by an 113%.

As the data shows, it is not enough to simply show product recommendations. The more customers engage with your offers, the more impact you will have. It is imperative that stores take time to set up effective rules for product recommendations.

If you want to see for yourself how Barilliance helps our customer's achieve these results,
click here.