-->

Product Recommendations

Basic Plan Pro Plan Enterprise Plan

You can use personalized product recommendations to encourage cross-selling and upselling within your store.

To recommend the most relevant products, Doofinder's recommendation model and algorithms are based on actual customer behavior. For this purpose, an artificial intelligence machine learning model is used, which dynamically learns what users visit together.

Product recommendations can be used for home and product pages. In order to collect data on customer behaviour it is necessary to have the widget script on the store pages. The recommendation model displays individual product recommendations based on products previously viewed by users with similar behavior through browsing and searching.

When you activate recommendations for the first time, the system will start learning from users' behavior. Once the widget is running on the website, the recommendations will initially be similar products. From that point on, the recommendation model will be collecting information about customer visits and generating recommendations based on the products that other customers have visited.

Recommendations screen

For customers visiting your store for the first time, Doofinder recommends globally attractive products on the home page and similar product recommendations for the product pages.

💡 Tip: Product Recommendations are not the same as the Query for recommended products used in the initial screen of your search. Find more information on Query for recommended products here.

How to activate Product Recommendations

To activate this feature, go to your Admin > Search Engines > Recommendations. Simply copy and paste the JavaScript code from the box into the corresponding area within your HTML code where you want the recommended products to be displayed:

Activate product recommendations

You can enable/disable the Recommendations, without removing the JavaScript from your code, by checking the ‘Status’ box.

Also, indicate the number of products you want to show in the carousel. By default, 10 recommended products are displayed. You can increase or decrease this number to suit your needs.

📌 Note: When you disable the recommendation widget from your site, it may continue to render up to 15 minutes later due to cache reasons.


Add currency

If for some reason your currency symbol or format is missing, you can add it to the recommendation script as follows:

<script>
<df-recommendations
  hashid="aeb873e432c45219ebf63b36c78811f0"
  total-products="10"
  region="us1"
  currency-symbol="₲"
  currency-format="%v %s"
></df-recommendations>
</script>

Symbol:currency-symbol="₲"

Format:currency-format="%v %s"

(%v = currency value)/(%s = currency symbol)

You can choose whether to type the symbol or value first and leave a space or not. For example:

100 € will be %v %S

$100 will be %s%v (no space between)

Magento 2 Recommendations Script

If you are using Magento 2, the recommendations script for require JS is different. You need to replace the first line in the default script found in the recommendations section with the following:

Default script

default recommendations script

New script line

<script>
  var dfUrl = 'https://cdn.doofinder.com/recommendations/js/doofinderRecommendation.min.js';
  (function(c,o,k,e){var r,t,i=setInterval(function(){t+=c;r=typeof(require)==='function';
  if(t>=o||r)clearInterval(i);if(r)require([k],e)},c)})(100, 10000, dfUrl);
</script>

Final script

The final script should look like this:

<script>
  var dfUrl = 'https://cdn.doofinder.com/recommendations/js/doofinderRecommendation.min.js';
  (function(c,o,k,e){var r,t,i=setInterval(function(){t+=c;r=typeof(require)==='function';
  if(t>=o||r)clearInterval(i);if(r)require([k],e)},c)})(100, 10000, dfUrl);
<df-recommendations
  hashid="5bdf17409acaa64fd36fc4fcf7c25820"
  total-products="10">
</df-recommendations>
</script>

Product Recommendations on Shopify

To implement Product Recommendations in your Shopify Store, you need to edit the existing code. Please click here to access the documentation.