---
title: "FAQ recommendation logic"
slug: "empfehlungslogik-1"
updated: 2025-01-16T13:39:57Z
published: 2025-01-16T13:39:57Z
canonical: "help.foxbase.de/empfehlungslogik-1"
---

> ## Documentation Index
> Fetch the complete documentation index at: https://help.foxbase.de/llms.txt
> Use this file to discover all available pages before exploring further.

# FAQ recommendation logic

**Which recommendation logic should I use?**

The logic that best suits your selector depends on the respective use case and the desired recommendation. During onboarding, the CSM team will help you determine the right logic.

**In principle, the following guidelines apply:**

- For particularly long questionnaires and / or a larger number of products, the use of **filters** is generally more sensible.
- **Filters** exclude unsuitable products and lead to 100% suitable recommendations. With the **matrix**, on the other hand, less suitable products can still be displayed as an option by entering smaller numerical values.
- As soon as calculations are required, you need the **Rules**. They are usually combined with the **matrix** or the **filters**. This allows you to filter the recommendations and calculate quantities, areas, etc. at the same time.

**What is the difference between filters and the matrix?**

**Filter** works on the principle of **either/or logic**: as with a kitchen sieve, unsuitable products are gradually filtered out with every question answered. The recommendations that fulfil all the user's criteria remain on the result page.

The **matrix**, on the other hand, enables to make a weighted recommendation: Here you enter numerical values per product and per answer in the workbench. The platform then calculates which product best meets the requirements by multiplying the values entered when the questionnaire is run through. The recommendation unit with the highest value is displayed first on the results page.

**How do I create a user input filter?**

1. In the **Filter** menu, click on **Add Filter > User Input Filter**.
2. Enter a title under **Filter Label** . The title is only visible to you in the backend and is not displayed to the selector user in the questionnaire.
3. Under **Answer - Conditions**, select the entries for which the filter should be activated. To do this, click on **Manage** and tick the answers in the questionnaire that you want to activate the filter.
4. Enter the question to be filtered under **Choice**. Under **Answer** , only select an answer if the filter should only be activated for a specific answer. If you want to configure a filter logic for each response, leave this field empty ( - ).
5. Enter the values from your imported product file in **Overwrite answer values for filter**.
6. Click on **Add Filter Function** and select the column from your imported product file under **Recommendation attribute**. The values that you have entered manually in Overwrite answer values for filter must match the values from the product column.
7. Select an **operator**.

**How do I add filters to the results page (Result Page Filter)?**

1. Click on **Filter >** **Add Filter**.
2. Select one of the options listed there (except User Input Filter).
3. Enter a title under **Filter label**. This will be shown on the Result Page.
4. Click on **Add Filter Function** and select the column from your imported product file under **Recommendation attribute**.
5. Select an **operator**.

**How do I achieve a weighted recommendation?**

You can obtain a weighted recommendation using the **matrix**. You store numerical values per product and per answer in the workbench. The platform then calculates which product best meets the requirements by multiplying the values entered when the questionnaire is run through. The recommendation unit with the highest value is placed first on the results page. A **combination of filters and matrix** also leads to a weighting.