Integration of the Recolize Conversion Tracking


The Recolize conversion tracking is one possibility to track the success of your recommendations. It can e.g. be integrated into the order success page of a shop or be activated in blogs or CMS after a specific visiting time.
If you already use a Recolize standard plugin for e.g. Magento or Shopify the conversion tracking is already integrated and you don’t have to add anything.

Technical Integration


Description: Integration on an order success page of a shop:

orderPositions: contains the order positions.
orderPosition.productId: the product id, which is also used in the Recolize product feed (e.g. also ‘id’ from the Google Merchant Feed)
orderPosition.netSum: net sum of an order position. Also ordered quantity of the order item * net value of the purchased product.


<script type="text/javascript">
var RecolizeParameters = RecolizeParameters || {};
RecolizeParameters['itemAction'] = 'sale';
RecolizeParameters['saleData'] = {};

for (orderPosition in orderPositions) {
RecolizeParameters['saleData'][orderPositions[orderPosition].productId] = RecolizeParameters['saleData'][orderPositions[orderPosition].productId] || 0.00;
RecolizeParameters['saleData'][orderPositions[orderPosition].productId] += parseFloat(orderPositions[orderPosition].netSum);
<script type="text/javascript">
    var RecolizeParameters = RecolizeParameters || {};
    RecolizeParameters['itemAction'] = 'sale';
    RecolizeParameters['saleData'] = {};
    RecolizeParameters['saleData']['SKU-100'] = 199.50;
    RecolizeParameters['saleData']['SKU-222'] = 12.67;


Description: integration on every page of a website which is contained in the Recolize feed. For blogs e.g. every blog article page from the RSS feed.

contentId: ID of the content in the Recolize feed. In RSS feeds of WordPress blogs this is e.g. the non-SEO-optimized URL of a page like
netSum: page value or optional “0” if unknown. With the value 0 on the quantity of conversions is tracked.

<script type="text/javascript">
    var RecolizeParameters = RecolizeParameters || {};
    RecolizeParameters['itemAction'] = 'sale';
    RecolizeParameters['saleData'] = {};
    RecolizeParameters['saleData'][contentId] = netSum;
<script type="text/javascript">
    var RecolizeParameters = RecolizeParameters || {};
    RecolizeParameters['itemAction'] = 'sale';
    RecolizeParameters['saleData'] = {};
    RecolizeParameters['saleData'][''] = 90.99;
How does the Recolize article feed format look like?

Recolize supports several ways to provide us with your product data that we need in order to generate personalized product recommendations.
Next to the Magento feed format, Google product or feeds we have a custom Recolize feed format that you can use.


Format: CSV
Encoding: UTF-8
Important: CSV file may not contain a BOM
Delimiter: semicolon (;)

Mandatory Fields:

  • id
  • name
  • url

Optional Fields:

  • image_url
  • brand
  • categories
  • price
  • special_price
  • price_currency
  • updated_at

Additional Fields:

Next to the mandatory and optional fields you can add as many additional columns to the feed as you want to. All fields will then be available in the Recolize Tool for filtering.


1;"Recolize Testproduct 1";;;Recolize;"Kategorie 1,Kategorie2";1.99;0.99;EUR;"2018-08-01 12:45:21"
Is it possible to change the HTML structure of the Recolize Carousel

We have made extensive research and evaluated user feedback on what is the best format and style of the product recommendations. And the result was our product carousel that you get with Recolize out-of-the-box. So for the huge majority of cases we recommend our customers to use this carousel and the presented elements as it is.

In most of the cases it is not necessary to change the structure of the Recolize Carousel HTML. You can use the CSS pseudo classes ::before or ::after to nearly do anything you want without changing the structure.

Therefore we do not recommend to change the HTML structure.

If you really still want to change the structure, you can do this in our success callback:

var RecolizeParameters = RecolizeParameters || {};
RecolizeParameters.Api = RecolizeParameters.Api || {};
RecolizeParameters.Api.successCallback = function(resultObject) {
    for (var widgetPosition in resultObject.widgets) {
        // add your structure changes here in resultObject.widgets[widgetPosition].content[0] 

How to use Context Filters to improve fashion recommendations for the chosen size of the user (or any other attribute)?

A common problem for product recommendations e.g. in the fashion market is to filter the recommendation carousels for the size of the user.

Given that we have a shoe online shop and the user is able to specify his shoe size in his account settings, we can use the following steps to filter the recommendations for that specified shoe size:

  • Save the user shoe size either in the session, browser cookie or browser local storage and define the appropriate Recolize parameters via JavaScript on every page before the general Recolize JavaScript snippet is defined:
var RecolizeParameters = RecolizeParameters || {};
RecolizeParameters.Context = RecolizeParameters.Context = {};
RecolizeParameters.Context.CurrentUser = RecolizeParameters.Context.CurrentUser || {};
RecolizeParameters.Context.CurrentUser.available_sizes = '#### PUT THE COMMA-SEPARATED LIST OF CHOSEN USER SHOE SIZES HERE ####';
  • Assure that in your Recolize product feed you provide the information of the shoe size either as a specific value (e.g. size = 36) or as a list of comma-separated values (e.g. size = 34,36,38). For now we assume that you have an additional column called “available_sizes” in your feed that contains a comma-separated list of shoe sizes that are available for the shoe.
  • Now go to the Recolize Tool and define a new Recommendation configuration with the following filter combination:

The interesting part here is the %CurrentUser.available_sizes% variable that gets filled with the chosen shoe sizes of the user.

The cool thing is basically you can define any attribute that you like, i.e. the whole JavaScript array below RecolizeParameters.Context is available in the filter of the recommendation configuration.

> Read more

WordPress and WooCommerce

How can I include images into the WordPress RSS feed?

Our Recolize WordPress plugin does out-of-the-box include article images into the default WordPress RSS feed.

The WordPress RSS feed url is e.g.

How can I integrate Recolize product recommendations into my WooCommerce shop?

Integrating personalized product recommendations generated by Recolize into your WooCommerce shop is very easy.
You just need to complete the following steps:

  1. Install the Recolize plugin for WordPress
  2. Register a Recolize account on
  3. In the Recolize tool create a new domain with your WooCommerce shop url and add the following feed URL: (of course you have to replace the domain with your shop domain)
  4. After saving the domain please copy the generated JavaScript snippet into the appropriate textbox in the Recolize settings of your WordPress installation.
  5. In the Recolize tool on your created domain add these CSS selectors (for the default WooCommerce storefront, this may vary on your chosen theme):
  6. Finally the Recolize conversion tracking pixel has to be integrated into your order success page. There see our FAQ for the conversion pixel integration.
WordPress and WooCommerce
> Read more


How to import a feed from a password protected url?

If you are e.g. using a development, testing or staging environment you may have a password protected website. In this case you have to enter the access credentials in the Recolize Tool for us being able to access your article feed url.

On the domain page in the Recolize Tool please use the following url scheme to import a Recolize feed from a password protected url:
> Read more


Can I add ratings or other elements to my recommendations?

Yes, of course. You can use our success callback to do so:

var RecolizeParameters = RecolizeParameters || {};<br />
RecolizeParameters.Api = RecolizeParameters.Api || {};<br />
RecolizeParameters.Api.successCallback = function(resultObject) {<br />
Recolize.Recommendation.Api.render(resultObject);<br />
/***** ADD YOUR CODE HERE *****/<br />
<div class="rating-box"></div>
<p>");<br />
Can I show banners in the Recolize carousel instead of products?

Yes, this is possible as Recolize works content-independent. So basically you can recommend products, articles, banners and anything else. You can also arbitrary combine these elements.

All you have to do is provide us an item feed that contains your banners and links. So you have to generate a CSV feed on your side that matches the specification that is described in our FAQ here.

In that CSV feed you just put your URLs for the images and the URLs for the links and a title (which can also be hidden in frontend via CSS) and you’re done. Now you can enter that generated feed url into our Recolize tool.

> Read more

Magento 2

How to display recommendations in Magento 2?

To display the recommendations on the following pages

  • Product pages
  • Category pages
  • Search result pages
  • Cart
  • Checkout success page
  • 404 page

use the following CSS selectors

.catalog-product-view .column.main
.catalog-category-view .column.main
.catalogsearch-result-index .column.main
.checkout-cart-index .column.main
.checkout-onepage-success .column.main
.cms-noroute-index .column.main

Add the CSS selectors to your domain in the Recolize Tool under Domains > Edit > Configuration > Please enter the CSS selectors for your page.

How can I create the product feed manually in Magento 2?

The product feed is generated automatically by our Magento 2 extension via the standard Magento cron.

In case you want to trigger the product feed generation manually, you can execute the following command in the directory of your Magento installation

bin/magento recolize:feed:generate

If you only want to create the product feed for a certain store, you can use the following command:

bin/magento recolize:feed:generate 
Magento 2
> Read more

Magento 1

How can I create the product feed export manually in Magento admin?

The product feed in Magento can be generated manually by executing the Dataflow profile:

  1. In the Magento backend go to System > Import/Export > DataFlow – Advanced Profiles and choose the appropriate Recolize Feed profile. Then click on the profile (all Recolize profiles start with “Recolize: ” in the name) and in the left tab “Start Profile” click on “Run Profile in Popup”.
  2. In the Recolize-Tool choose the appropriate domain und save it again to initiate the import.

If the Magento cron job is running correctly the Recolize Feed is generated daily and imported into Recolize.

If the problem persists it has to be checked if the PHP memory limit and maximum execution time is sufficent and increase the appropriate values accordingly (PHP setting memory_limit and max_execution_time).

Magento shows “Service Temporarily Unavailable” after extension installation via Magento Marketplace

If you install our Recolize Magento Extension via the Magento Marketplace downloader in the Magento backend you may experience an issue that your site shows an error message like “Service Temporarily Unavailable”.

This is not specific to our Magento extension and may occur with any extension installation. The reason is that you can opt in to set Magento into a maintenance mode during extension installation. Unfortunately there may be cases where this maintenance mode is not quit correctly after the extension installation has been completed.

To solve this issue just go to your Magento root directory and delete the file maintenance.flag. Then your shop is accessible again.

Where do I find the Recolize feed URL in Magento?

Navigate in Magento Admin to
System > Configuration > Recolize Recommendation Engine > Recolize Product Feed Settings.

Here you find a feed url for every Magento store view. Please create a separate Recolize domain for every store in Magento in the Recolize tool under

Magento 1
> Read more


Why is my filter configuration not working?

You’ve created a specific filter configuration but the users still get recommended other products? If this is your problem then the personal recommendations of the current user don’t fit to your filter configuration.
If your filter configuration is very restrictive and no matching recommendations are contained in the user’s profile (e.g. the user is only interested in pants but your recommendation is configured to show only products with the category shoes) then the product recommendations of the next matching recommendation are offered. If none of the configured recommendations match for a user then the recommendations are generated by the Recolize Brain.

How can I add a filter to a filter combination?

Choose the filter in the tool according to the description in the right column. Then the filter can be added per drag & drop (click with the mouse, drag the element and drop it) to the filter combination box. After these steps your filter was successfully added.

How can I show pre-defined items in the Recolize carousels?

By default Recolize generates personalized product recommendations, i.e. they are automatically calculated to be the best products for the individual visitor. This works right from the start, that means after the installation you do not have to do anything for this to work.
But of course there are several ways you can influence the recommended items to be able to show pre-defined ones:

1. Go to My Recommendations > Add New Recommendation and configure a new recommendation configuration. Under point “5. Please Choose a Data Pool” you can choose between various options for the data pool:

  1. Personalized (User-to-User): Automatically calculate the individual recommendations for your visitors based on other visitors. This is the default setting.
  2. Cross-Sells (Item-to-Item): Automatically calculate the individual recommendations based on the current item.
  3. Bestseller
  4. Last Viewed Items
  5. All Items: This data pool allows static item lists by additionally defining filters. Please select this one.

2. Now you can define the item filters under point “4. Please Define Your Filters”. For example you can limit the static item pool to certain SKUs or items with a price below 10 Euro and something like that.

Please note:
The number of shown articles in our carousel is currently set to 10 because we have made several experiments with our customers and this proved to be the ideal size.

My recommendations show items out of stock. How to solve this?

You can exclude items that are out of stock by adding a new Recommendation in the Recolize Tool. Select the “All Pages” URL and add the filter “Is item in stock = Yes”.

How can I control the position that the recommendations widgets will be displayed?

You are completely free to position the recommendation widgets wherever you want. Out of the box for most systems like Shopify and Magento we have a default set of positions like e.g. on the product detail page, in cart and on checkout success page.

The position of the recommendation widgets is determined by the configured CSS selectors in our Recolize tool.

Every widget is a new line in the configuration section. You can use all the available css selectors as e.g. documented here.

We assume a configuration as follows:
.main-content .recommendations

In this case the recommendation widgets will be displayed for every HTML element that matches “.main-content .recommendations” and “.footer”.
Let’s assume that there are 2 positions matching those selectors on your website. Then the first match gets the position 1 and the second one position 2 for the recommendation configuration.

How can I control the look’n’feel of the recommendations via CSS?

Next to the configuration and styling possibities in our Recolize Tool you can further customize the recommendation widgets via plain CSS.

Therefore all elements in our carousels have explicit CSS selectors that you can use to create your own styling.
These CSS styles can be added in your domain configuration in the Recolize Tool.


.recolizePrice {
    font-size: larger;
Which recommendation algorithms are supported?

Recolize supports the following recommendation algorithms:

  • User-to-user (“Customers, who bought this, also bought”)
  • Item-to-item (“Cross Sells”)
  • Top items / Bestseller
  • Last viewed
  • Static product lists

The algorithms are chosen automatically dependent on the current visited page and the current user to always generate the best matching recommendations.

What are useful CSS selectors for positioning the Recolize recommendation widgets?

Recolize uses CSS selectors to control the positions of the Recolize recommendation widgets. The selectors can be configured in the Recolize tool.

A full list of available selectors can be found here. Below we have highlighted some useful ones.

Selector Description
p:contains(‘Your cart is empty’) Recommendation Widgets are added right after the text “Your cart is empty”.

What is the best time interval to set for the import of your article feed?

By default we recommend importing the article feed every 12 hours. The value mainly depends on how often your articles change in your system and how fast you want the changes to be reflected in the recommended items. It also depends on how often the product feed is generated on your side.

How to add my payment data in the Recolize Tool?
  1. Login with your email address under
  2. Click on “My Account”
  3. Click on “Edit Data”
  4. Add your payment details
What is the difference between Up-selling and Cross-selling?

We often see confusion regarding the 2 topics “Up-selling” on the one hand and “Cross-selling” on the other hand, although there are major differences between both. In addition, also E-Commerce systems like Magento do not distinguish properly between these two which increase the uncertainty about them.
Generally both terms describe the process of providing recommendations for other products on the basis of one or more products. The difference between Up-selling and Cross-selling will be examined in the following paragraphs.

In the process of Cross-selling several other products are recommended that are similar to the current one and may be exchanged with it. A popular example is a bike tire of brand A that can be exchanged with a tire of brand B that has similar properties (e.g. kind of terrain, pricing, etc.).
In order to be able to display those recommendations Recolize compares and weights the product attributes of the current tire with all available other products and automatically learns and creates those cross-linking connections between them.
For more information about Cross-selling please also see our blog articles about 3 Ways to Integrate Cross-Sells on Product Pages and 15 % Conversion Uplift: Automated vs. Manual Cross-Sells.

On the other hand, Up-selling describes the process of recommending products that are more expensive than the current one(s). Imagine for example you added a watch in the mid-price region to the shopping cart and on the cart page you will be displayed with luxury watches that have a much higher price.
Up-selling in Recolize can be easily done with one click by using its flexible filtering engine.
More use cases for Up-selling can also be found in our blog articles 5 Advantages of Recommendations in Transactional Emails and The Power of Real Individual Recommendations.

> Read more


Which payment methods do you offer?

Recolize is a Software-as-a-Service with monthly payment of your used clicks.
We offer the following payment methods:

  • PayPal
  • Debit
  • Credit Card

You can find further information to PayPal here.

What is the Recolize Brain?

The Recolize Brain is the standard recommendation for all of your users and is active on all pages in your shop. If you didn’t create a more specific recommendation with e.g. additional filters then the Recolize Brain generates the personalized recommendations for your users.

Where is all the data saved?

Data security has the highest priority for us. Therefore the complete data (e.g. visitor and customer tracking data) is saved in data centers in Germany.

How about performance of the recommendations?

Recolize has been especially optimized for performance so that the personalized product recommendations can be delivered in under 100 ms in general.

As Recolize is being served as a Software-as-a-Service, it’s completely independant of your hosting provider and does not influence your page performance or loading time of your website or shop at all.

Images in the carousel are not displayed, what can I do?

If the images in the Recolize carousels are not shown, the most likely reason is that the images are no longer available at the URLs that you have provided in the Recolize product feed. Depending on your system there are several points to check.


Possible causes in Magento are

  • the image cache has been manually flushed in the Magento backend
  • the Magento cron job is not running (any more)
  • the PHP memory limit for the process of the thumbnail generation has been reached

As a solution the product feed in Magento has to be regenerated and the import has been triggered in the Recolize tool. See our appropriate FAQ article.

If the Magento cron job is running correctly the Recolize Feed is generated daily and imported into Recolize, i.e. at the latest on the next day the problem is also solved automatically.

If the problem persists it has to be checked if the PHP memory limit is sufficent and increase the appropriate value accordingly (PHP setting memory_limit).

Which data is used to generate personalized recommendations?

If a shop visitor is anonymous and visits the page e.g. for the first time, then his clicks on products, add to carts etc. is tracked. Every time the visitor clicks or moves from page to page, his profile is updated and his recommendations get updated.
If the user logs in or is already logged in the data is saved in his customer profile

What are the benefits of a Software-as-a-Service model?

Recolize is provided as a Software-as-a-Service that means you do not have to take care of any infrastructure issues, hardware failures, software updates, etc. We do everything for you and you get in return the generated personalized product recommendations.

In contrast to a so-called “on premise” installation the Software-as-a-Service model has lots of benefits for you:

  • We constantly update our software and deliver you the latest highly innovative features – and there’s nothing you have to to do for that: you’ll get these for free and right when they become available
  • No software or hardware updates required
  • Unlimited scalability: No matter if you have hundreds, thousands or millions of visitors on your website the Recolize recommendations will always be calculated extremely fast within around 100 milliseconds.
  • High availability: the Recolize service has an uptime above 99,99 % so every visitor of your page can be served with personalized product recommendations
  • Our German datacenters are highly secured to confirm with the German data privacy laws.

So what are you waiting for? Register for free and check these points out.

Where can I download my Recolize PDF invoices?

You can download all invoices in the Recolize Tool:
Your domains > Click on the domain > Download

How to integrate the Recolize Opt-out?

For generating personalized recommendations Recolize creates pseudonymous user profiles, which exclusively can be connected to an end user with a correspondent cookie on the system of the end user and are only usable on your own website respectively your own domain. This cookie contains a randomly generated user id and can be deactivated with the Recolize Opt-out by the website visitor. Please add a correspondent advise to your privacy policy.
Recolize Recommendations also work with active Opt-out without any issues. Then non-personalized products are delivered, that is only product-based cross sells and best sellers, which increase the conversion rate of your website.

To integrate the Recolize Opt-out, you just have to add the following link to your privacy policy. This link is working completely automatically and displays the appropriate text depending on the current Recolize Opt-out status of the website visitor:

<a class="recolizeOptOut" href="#" onclick="return false;" data-activate-opt-out-text="Activate Opt-out" data-deactivate-opt-out-text="Deactivate Opt-out"></a>

Please replace the colored text parts with your individual texts if you want to. The text will automatically switch:

  • “Activate Opt-out”: this text will be displayed if the Recolize Opt-out is not active. With clicking on the link, the Recolize Opt-out will be activated and the link text will switch to “Deactivate Opt-out”.
  • “Deactivate Opt-out”: this text will be displayed if the Recolize Opt-out is already active. With clicking on the link, the Recolize Opt-out will be deactivated and the link text will switch to “Activate Opt-out”.
When do you count a sale as conversion?

Recolize uses the so called conversion conversion tracking to determine, which articles were bought over the personalized recommendations by your customers. Therewith we have a powerful tool to optimize your recommendation configuration in the Recolize Tool and to generate more revenue with your product recommendations.

A sale will only be counted as Recolize conversion when a user clicked exactly the bought product in a recommendation widget.

The sale of an article will be tracked as Recolize conversion when it is in a close temporal connection to an interaction with a recommendation.

Recolize only tracks the information for bought articles and of course no personal data of your customer like addresses, email address etc.

How Does Recolize Determine the Top 10 URLs in the Dashboard?

In the Recolize Tool you can find 2 widgets in the dashboard displaying the top and flop 10 URLs. Here URLs are displayed where the personalized recommendations perform best respectively worst.

If and which URLs are listed here Recolize automatically defines by intelligent logics. Therefore conversions, revenue, click count and views etc. are combined with complex algorithms and rated.

Especially you should monitor the flop 10 URLs. With one click on the URL you get further information in the tool and recommended actions by our smart Customer Intelligence logic.
For instance it can make sense to adapt the recommendation configuration and only display low-priced products or only articles from a specific category.

> Read more