What is PopKart

While researching other domains on the web, we found that recommendation engines are one of the strongest modules used to engage website audience with new and exciting content. The method used to find relevant content that is displayed to said audience can be used in online carts as well. A customer who did not complete a checkout process on one site can be tagged and then reminded while viewing a different site's checkout process that he has not completed the previous checkout. This can be done if both sited implement the same code which is the case with the Rapyd embedded checkout. PopKart wraps around the Rapyd code with the PopKart recommendation engine.

The advantage to the sites is that the each site gains exposure for reminding the customer of the abandoned cart by both of them integrating the PopKart engine.

How we built it

We started by building two separate embeddable checkout processes using the Rapyd embedded checkout. Every checkout page integrated our SDK with the Rapyd checkout ID. We then detected when a customer did not complete the checkout process and stored a cookie on the customer's device indicating so. When the customer visits the next site that is using our engine, and gets to the checkout process, we identify the cookie and display a recommendation to the previous site with a quick "Finish checkout" call to action.

Challenges we ran into

Since the customer id is different between organizations, we found it hard to be able to track the customer between checkouts. However, we found a solution by mimicking the process that is used by analytic tracking services.

What's next for PopKart

There is a lot of potential for PopKart. A few items we see this going are the following:

  1. Smarter recommendation - Not every website would like to have the recommendation running on their website nor would they want "competing" sites listed. We would like to develop this smarter to avoid "stealing" of the customer.

  2. Cart Items - We would like to have in the future the ability to show cart items in the recommendation boxes and not only the store logo as we do now. This can be done by allowing functionality to store the cart items together with the session when created

  3. Management dashboard - A dashboard for every customer to be able to view the efficiency of the recommendation engine as well as set different settings.

  4. Better UI - With the limits we had in time and team capabilities, we believe we got to a good UI but can do better.

There is of course many other functionalities that can be implemented and we look forward to it.


To test our recommendation engine, follow the following steps:

  1. Go over to http://popkart.benchlabs.xyz/files/checkout_1.html.
  2. Once you have viewed the page, you can now go over to http://popkart.benchlabs.xyz/files/checkout_2.html you should see the recommendation for the first checkout.
  3. If you would like to experience the recommendation engine after the checkout, you can go over to http://popkart.benchlabs.xyz/files/checkout_3.html and checkout by entering one of the demo credit cards available by Rapyd. Notice: You will be able to checkout only once!
Share this project: