Analytics & Reports

Numeric Biasing

Numeric Biasing allows you to influence the order of results by using things like

  • sales rank
  • popularity
  • clickboost
  • margin

… and other numeric values, so that results are automatically influenced by the numeric ranks.

Numeric biasing is configured within a biasing profile, and can be a powerful way to merchandise your browse experience without having to manually curate each product.

To apply numeric biasing you’ll need to create a Biasing Profile, and associate it to the trigger on which you want it to apply, typically via a Rule.


When do you typically want to use this?

Typically, within biasing profiles you will specify business flags to use as influences. For example, bias out of stock items to the bottom, or bias in-house products higher. Numeric biasing lets you apply an additional factor on top of those biases - with everything being incorporated. So, something not in-house but with high margin will be higher than something with a lower margin, even if it’s in house. You can use the Strength control to decide how much any numeric bias influences the results.

Which Fields Can Be Used?

Any numeric fields, that are marked as a Float type at upload time within the given collection, can be used.

How Can These Fields Be Used?

You can use numeric values to bias in 2 ways:

  1. Use a field where larger numeric values are better
  2. Use a field where smaller numeric values are better

To bias by a numeric field, add it under the Numeric Boost section by typing in the field name. You will see autocomplete suggestions based off all the fields that are defined as numeric within your catalog.

Note: By default, larger values are boosted.

Records with no value will exert no boost or bury.

How to Boost Smaller Numbers Higher

Check “invert” to boost in reverse - smaller values in the field will boost the record higher, and bigger values in the field will lower it in the results.

Which Records Will Be Affected?

The numeric boost is context aware to the query and the results - that is, the values that are used for applying the boost are based off the range of the results. This means two things:

  1. Numeric boost will result in different level of boost, given a specific record-value, depending on what other records are returned.
  2. Because this is a context aware functionality there will be a slight performance impact at query time (around 20%).

The exact content of the field that you use for boosting (e.g., whether it contains “1”, or “100”, or “103323432”) doesn’t matter as much as the overall distribution of the values across the results.

Numeric Bias Strength

You can control how strongly each numeric bias will influence the rank of any given product/record, by setting the Strength field to a value between 0 and 10.

This lets you balance multiple different numeric biases so that not all fields are weighted equally.

Set this value to an effective weight / rank to reflect the importance of a given field within your desired business rule.eventsapi

Numeric Bias Example

Let’s say we have a basic set of 5 products and 2 fields we can use for biasing:

  • Inventory
  • Rating
Biasing Fields

If we want to bias them by inventory (more is better):

You will see the order change like so: