Filters and boosters – precise recommendation targeting

In your recommendations, you can use additional filters to narrow down the display of recommended products (for example, to specific categories) and boosters that increase the likelihood of recommending selected products.

To use these parameters:

Your product catalog must contain the attributes you want to use as selection criteria.

You need to use the proper recommendation system syntax (detailed guidelines).

Where can I find filters and boosters in recommendations?

Go to the editor of your selected content, then:
Click the Content tab in the left panel.
- Find the Recommendations block and drag it to the workspace.
If you’re creating pop-up, banner, or email content in a scenario:
- Find the Products block and drag it to the workspace.
- In the right panel, select Recommendations in the Product type section.
Click Recommendations on the top toolbar.
Go to the Advanced settings section.

Filters

Filters let you narrow down product displays according to specific criteria. With filters, you can:

Display products from specific categories.
Exclude certain products from recommendations.
Limit product displays, for example, based on age categories.

Filters work on true/false logic. When a product meets the filter condition, it’s included in the recommendations. When it doesn’t meet the condition, it’s skipped. You can apply several filters simultaneously in one expression.

Filter examples

Example 1: Recommending products from a summer promotion

Let’s say your catalog has the following products:

Name	Available	Price	Sale price	Category	Summer 2025 promotion
Swimsuit	true	199		Women’s clothing	true
Beach tunic	true	99		Women’s clothing	true
Short cotton kimono	true	120	99	Women’s clothing	false
SPF 50 sunscreen	true	59		Cosmetics	true
Straw hat	true	50		Accessories	true
Towel 80 x 160 cm	true	79		Accessories	false

To show only products from the summer promotion, enter this in the Filter field:

'Summer 2025 promotion' == true

This way, visitors will see recommended products from different categories that are part of the summer promotion.

Example 2: Filtering by multiple conditions

If you want to display only products that are both in the summer promotion and belong to the “Women’s clothing” category, enter this in the Filter field:

'Summer 2025 promotion' == true and "Women's clothing” in 'Category'

Remember to use quotation marks correctly:

Single quotes (‘) for true/false values
Double quotes (“) for text type attributes.

Example 3: Filtering products by price

If you want to display only products with a price exceeding a certain value (e.g., 80), enter this in the Filter field:

'Price' > 80

The result of applying this filter:

Name	Available	Price	Sale price	Category	Summer 2025 promotion	Displayed?
Swimsuit	true	199		Women’s clothing	true	yes
Beach tunic	true	99		Women’s clothing	true	yes
Short cotton kimono	true	120	99	Women’s clothing	false	yes
SPF 50 sunscreen	true	59		Cosmetics	true	no
Straw hat	true	50		Accessories	true	no
Towel 80 x 160 cm	true	79		Accessories	false	no

As a result, the recipient will see only products with a price higher than 80, namely the swimsuit, beach tunic, and short kimono.

Example 4: Filtering discounted products

To display only discounted products (with a sale price greater than 0), enter this in the Filter field:

'Sale price' > 0

Alternatively, you can also use:

'Sale price' is not null

The result of applying this filter:

Name	Available	Price	Sale price	Category	Summer 2025 promotion	Displayed?
Swimsuit	true	199		Women’s clothing	true	no
Beach tunic	true	99		Women’s clothing	true	no
Short cotton kimono	true	120	99	Women’s clothing	false	yes
SPF 50 sunscreen	true	59		Cosmetics	true	no
Straw hat	true	50		Accessories	true	no
Towel 80 x 160 cm	true	79		Accessories	false	no

After applying this filter, the recipient will see only the short cotton kimono, as it’s the only product in the catalog with a sale price.

Boosters

Boosters let you increase the likelihood of recommending selected products. Unlike filters, which can completely exclude products, boosters only prefer certain products to others.

By default, the recommendation system independently selects the most relevant products. However, you can intentionally influence the system to better achieve your business goals.

Booster expressions return positive numbers (coefficients) by which the internal scores of the recommendation system are multiplied.

Booster examples

Example 1: Boosting products from the ‘Summer 2025 promotion’ campaign

This booster increases the ranking of all products marked as ‘Summer 2025 promotion = true’.

if 'Summer 2025 promotion' == true then 2 else 1

This expression means that every product from the summer promotion receives a multiplier of 2 for its recommendation score.

Example 2: Boosting only discounted products

Let’s say you want to promote discounted products. In our sample product catalog, the Sale price field either has a number or is empty. In the Booster field, enter:

if 'Sale price' != null then 1.5 else 1

This expression highlights products that have a value in the Sale price field.

Scenarios

Scenarios help differentiate various recommendation contexts – when, where, and which products are recommended. With them, you can:

Specify where the recommendation should appear, for example on the homepage, product page, or in an email.
Train the recommendation model in a specific context, for example what works better on the homepage versus in the shopping cart.
Analyze effectiveness separately for each scenario (CTR, conversions).

Launching scenarios requires support from our customer service department.
Contact your ExpertSender account manager to activate this feature.

Check out other articles on recommendations: