1. Home
  2. User Guides
  3. Miscellaneous
  4. Customer filters – In detail

Customer filters – In detail

This post is also available in: Czech

There are more than 60 customer filters available for you in Samba. In this article, we will explain what each filter means, what values it can take and what it is used for.

Let’s start by explaining how we distinguish between the values that customer filters can take on (you can find a detailed description here):

  • Exact value
  • Exact match or a substring of values
  • Numeric value

Customer filters can be categorized according to the information they contain.

General characteristics

Full name

When using this filter, select an exact match or substring of your selected customers first and last name.

The filter tracks the FIRST_NAME and LAST_NAME values from the customer feed.

Customer ID

Use this filter to select customers with a specific ID from your audience or multiple customers where at least part of their ID matches the substring you selected.

The filter works with the CUSTOMER_ID parameter in the customer feed.

Firts name

Use this filter to select customers with a specific first name from your audience or multiple customers where at least part of their first name matches the substring you selected.

Unlike the "Full Name" filter, here we work only with the FIRST_NAME parameter.

Email

Select customers with a specific email address from your audience, or multiple customers where at least part of their email address matches the substring you selected.

The filter works with the EMAIL parameter in the customer feed.

Phone number

With this filter, you can select your audience based on their phone number. Samba gets this information through the PHONE parameter in the customer feed.

You can either select a specific phone number or, alternatively, multiple results where at least part of the results corresponds to the substring you selected.

Registration date

Filter your audience based on when the customer registered on your store. The value for this filter is given by the REGISTRATION parameter in the customer feed.

When filtering, you can choose whether the registration date for your resulting audience should be Less than (and equal to), More than (and equal to), Equal to or Between the date you select.

Days since registration

Use this filter to select an audience based on how many days have passed since the customer registered on your store.

When filtering, you can choose whether your resulting audience should be Less than (and equal to), More than (and equal to), Equal to or Between the number you select.

Zip code

This filter allows you to select your audience based on their postcode. Samba gets this information through the ZIP_CODE parameter in the customer and order feed.

You can select either a specific zip code or, alternatively, multiple results where at least part of the zip code matches the substring you selected.

District

This filter allows you to select your audience based on what district they are from. The district is determined based on the most frequent ZIP_CODE values in the order or customer feed.

You can select either a specific district or, alternatively, multiple results where at least part of the district name matches the substring you selected.

Gender

Filter the audience based on whether they are female or male. Gender is automatically determined by parsing the full name as well as the customers email address.

Has name day today

In the case of this filter, you can select those customers who do or do not have a name day on a given day.

Thanks to the FIRST_NAME parameter from the customer feed, Samba is able to determine whether a given customer has a name day on a given day (this filter currently works for countries CZ, HU, PL, RO, SK).

The filter can only take 2 values:

  • Yes
  • No

When filtering, you select the exact filter value.

Days to name day

Select an audience based on how many days are left until the customers name day (this filter currently works for countries CZ, HU, PL, RO, SK).

When filtering, you can choose whether your resulting audience should have the number of days to name day Less than (and equal to), More than (and equal to), Equal to or Between the number you select.

Customer price category

Determine the customers price category according to the PRICE_CATEGORY attribute in the customer feed. This is the customer group to determine the special price - see Customer groups with different prices.

You can define different selling prices and original prices before discounts for different customers using the PRICES attribute. This can be useful, for example, if you use different discount programs or work with "VIP" and "guest" customers, etc.

When filtering, you select a specific parameter, inserted in your customer feed, and which corresponds to your desired customer price category.

Data permission

Filter customers by whether they have consented to purchase behaviour processing.

Samba works with the information in the DATA_PERMISSION parameter in the customer feed and distinguishes only 2 values:

  • Full - the customer has given full consent to the processing of purchase behaviour tracking information.
  • Do not personalize - the customer has not given consent to the behavioural processing and therefore does not wish to receive personalized advertising and newsletters.

When filtering, you select the exact filter value.

Custom customer parameter

In case you insert a custom parameter into your product feed in the PARAMETER field with the name NAME and value VALUE, you can then use it when filtering your products.

Custom Parameters: Days since date & Date

To use this filter it is necessary to set the parameter in the configuration of custom parameters to the "datetime" data type.
We have written a detailed description with all the information about these custom parameters in a separate article here.

Days to anniversary of custom parameter

If you set the custom parameter in the custom parameter configuration to the date type "datetime", you can use this filter to find the number of days until the next anniversary with respect to the value in the custom parameter.
Find out how to use the filter for your birthday card in our article.

Days since anniversary of custom parameter

If you set the custom parameter in the custom parameter configuration to the date type "datetime", you can use this filter to find the number of days since the last anniversary with respect to the value in the custom parameter.
Find out how to use the filter for your birthday card in our article.

Years since date of custom parameter

If you set the custom parameter to the "datetime" date type in the custom parameter configuration, you can use this filter to find the number of years since the date specified in the custom parameter.
Find out how to use the filter for your birthday card in our article.

Purchase activity

Recently purchased products

In this case, this is a so-called nested filter. When you select this filter, you can insert a product filter into your customer filter. Thus, by using the product filter, you define the products that your target audience should have purchased in the period you specified.

You can narrow your audience even further by specifying the total order price, the number of purchases, etc.

Number of purchases (Total)

With this filter, you can select an audience based on how many total purchases a customer has made in your e-shop.

When filtering, you can choose whether your resulting audience should have a total number of purchases Less than (and equal to), More than (and equal to), Equal to or Between the number you select.

Number of purchases (Last 365 days)

Use this filter to select an audience based on how many purchases a customer has made in your e-shop in the last 365 days.

When filtering, you can select whether your resulting audience should have the number of purchases in the last year Less than (and equal to), More than (and equal to), Equal to or Between the number you select.

Average order value (Total)

Use this filter to select an audience based on the average order value for a given customer in your e-shop.

When filtering, you can choose whether your resulting audience average order size should be Less than (and equal to), More than (and equal to), Equal or Between the number you select.

Average order value (Last 365 days)

With this filter, you can select an audience based on what the average order value was for that customer in your e-shop over the last 365 days.

When filtering, you can choose whether your resulting audience should have an average order size of Less than (and equal to), More than (and equal to), Equal to or Between the number you select.

Number of purchased products (Total)

In the case of this filter, you can define the audience of customers who have purchased the total number of products you specify.

When filtering, you can choose whether the total number of products purchased for your resulting audience should be Less than (and equal to), More than (and equal to), Equal to or Between the number you select.

Number of purchased products (365 days)

With this filter, you can define an audience of customers who have purchased the number of products you specify in the last 365 days.

When filtering, you can choose whether the annual number of products purchased for your resulting audience should be Less than (and equal to), More than (and equal to), Equal to or Between the number you select.

Spending (Total)

Use this filter to select an audience based on how much the customer spent in total on your store.

When filtering, you can choose whether your resulting audience total spending should be Less than (and equal to), More than (and equal to), Equal or Between the number you select.

Spending (Last 365 days)

Use this filter to select an audience based on how much a customer has spent in your e-shop in total over the last 365 days.

When filtering, you can choose whether your resulting audience total annual spending should be Less than (and equal to), More than (and equal to), Equal or Between the number you select.

Date of last purchase

Filter the audience by when the customer last made a purchase in your e-shop.

For this filter, Samba takes the most recent value for the FINISHED_ON parameter in the order feed.

When filtering, you can choose whether the date of the last order on your e-shop should be Less than (and equal to), More than (and equal to), Equal to or Between the date you select for your resulting audience.

Days since last purchase

Use this filter to select an audience based on how many days have passed since the customer last order in your e-shop.

When filtering, you can select whether the number of days since the last order should be Less than (and equal to), More than (and equal to), Equal to or Between the number you select.

Date of first purchase

Filters the audience by when the customer first made a purchase on your store.

Unlike the previous filter, for this filter Samba takes the oldest value for the FINISHED_ON parameter in the order feed.

When filtering, you can choose whether the date of the first order on your e-shop should be Less than (and equal to), More than (and equal to), Equal to or Between the date you select for your resulting audience.

Days between last two purchases

With this filter, you can select an audience based on how many days after which a customer repeated their last order in your e-shop.

When filtering, you can choose whether your resulting audience should have the number of days between the last two orders Less than (and equal to), More than (and equal to), Equal to or Between the number you select.

Purchased product ID

Use this filter to select customers from your audience who have purchased a product with a specific ID, or multiple products where at least part of their ID matches the substring you selected.

Samba gets the value for this filter from the product feed from the PRODUCT_ID parameter.

Purchased product title

You can also filter customers from your database based on whether they purchased a product with a specific name you selected. Alternatively, you can select multiple products where at least part of their name matches the substring you selected.

In the case of this filter, the value of the TITLE parameter is retrieved from the product feed.

Web activity

In all filters, visitor activity on the site for the last 30 days is available.

Recently visited products

As with the "Recently Purchased" filter, the product filter is nested within your customer filter here. With its help, you define the products that your target audience must have visited in the period you specify.

You can narrow your audience even further by specifying the number of days since the visit and the number of visits.

Total number of visits

Within this filter, you can select how many total visits to your site (in the last 30 days) you want the resulting audience to have.

When filtering, you can select whether the resulting audience should have a number of visits Less than (and equal to), More than (and equal to), Equal to or Between the value you select.

Days between last two visits

With this filter, you can select an audience based on how many days it has been since a given customer returned to your store since their last visit.

When filtering, you can select whether your resulting audience should have the number of days between two visits as Less than (and equal to), More than (and equal to), Equal to or Between the number you select.

Days since last visit

Use this filter to select an audience based on how many days have passed since the customer last visited your store.

When filtering, you can choose whether your resulting audience should have the number of days since the last visit as Less than (and equal to), More than (and equal to), Equal to or Between the number you select.

Email channel

Newsletter frequency

Within the NEWSLETTER_FREQUENCY attribute in your customer feed, customers can reach three values:

  • Every day - These customers are subscribed to your newsletter and want to receive regular updates.
  • Special Occasions - Customers in this category are also subscribed to newsletters but do not wish to receive emails every day, but only want to be notified when there are "special" events, etc.
  • Never - Customers who are unsubscribed from newsletters. These customers do not wish to receive any promotional or business-type emails. You are only allowed to target these customers if you communicate important information that does not convey a marketing message.

When filtering, you select the exact filter value.

Date when last mail was received

Filters the audience based on when your email was last delivered to the customer.

When filtering, you can select whether the last email delivery date for your resulting audience should be Less than (and equal to), More than (and equal to), Equal to, or Between the date you select.

Date of last mail open

Filters the audience based on when the customer last opened your email.

When filtering, you can choose whether the last email open date for your resulting audience should be Less than (and equal to), More than (and equal to), Equal to, or Between the date you select.

Date of last mail click

Filter the audience by when the customer last clicked through your email.

When filtering, you can choose whether the last email click date for your resulting audience should be Less than (and equal to), More than (and equal to), Equal to, or Between the date you select.

Days since last mail received

Filter the audience by how many days have passed since your email was last delivered to the customer.

When filtering, you can choose whether the number of days since the last email delivery for your resulting audience should be Less than (and equal to), More than (and equal to), Equal to or Between the number you select.

Days since last mail open

Filter the audience by how many days have passed since a customer last opened your email.

When filtering, you can choose whether the number of days since the last email was opened for your resulting audience should be Less than (and equal to), More than (and equal to), Equal or Between the number you select.

Days since last mail clicked

Filters the audience by how many days have passed since your customer last clicked through your email.

When filtering, you can choose whether the number of days since the last email click should be Less than (and equal to), More than (and equal to), Equal to or Between the number you select for your resulting audience.

SMS channel

SMS frequency

Within the SMS_FREQUENCY parameter in your customer feed, customers can have three values:

  • Every day - These customers are subscribed to receive business and promotional messages via SMS and want to receive regular updates.
  • Special Occasions - Customers in this category are also subscribed to SMS but do not wish to receive messages every day, but only want to be notified when there are "special" events, etc.
  • Never - Customers who are unsubscribed from receiving SMS. These customers do not wish to receive any promotional or commercial-type messages. You are only allowed to target these customers if you are communicating important information that does not carry a marketing message.

When filtering, you select the exact filter value.

Date of last sent sms

Filters the audience based on when your SMS was last sent to the customer. When filtering, you can choose whether the last SMS send date for your resulting audience should be Less than (and equal to), More than (and equal to), Equal to or Between the date you select.

Days since last sent sms

Filter the audience by how many days have passed since the last time your customer sent an SMS. When filtering, you can choose whether the number of days since the last SMS sent to your resulting audience should be Less than (and equal to), More than (and equal to), Equal to or Between the number you select.

Favorites

Popularity is defined as a customer’s most frequent purchase of a given category/subcategory/brand/product line among all their purchases.

Favorite category
Use this filter to select customers from your audience for whom the category you selected is their favorite. Alternatively, you can increase the number of selected customers by selecting more categories, where at least part of their name corresponds to the substring you selected.
Favorite subcategory
With this filter you can select customers from your audience for whom the most purchased subcategory is the category you selected. Alternatively, you can increase the number of selected customers by selecting more subcategories, where at least part of their name corresponds to the substring you selected.
Favorite product line
With this filter you can select customers from your audience for whom the most purchased product line is the product line you selected. Alternatively, you can increase the number of selected customers by selecting multiple product lines, where at least part of their name corresponds to the substring you selected.
Favorite brand
With this filter you can select customers from your audience for whom the favourite product brand is the product brand you selected. Alternatively, you can increase the number of selected customers by selecting multiple brands where at least part of their name matches the substring you selected.

CLV model

Samba has developed a unique probabilistic Customer Lifetime Value (CLV) model that analyses customer buying behaviour and predicts future behaviour based on this.

Unlike, for example, the standard RFM model, which works with fixed boundaries to determine the score (although in Samba the boundaries are determined dynamically based on business-specific data), CLV model takes into account the behaviour of each individual customer. This makes the resulting parametric model much more flexible and accurate.

Churn category

Depending on the purchase behaviour (customer must have made at least 1 purchase), customers are divided into 3 categories:

  • Active - a customer who repeatedly and frequently purchases on your e-shop = high probability of repeat purchase
    • E.g. a customer who shops regularly every 14 days has 7 days since their last order. The model, therefore, evaluates that the customer will buy again within 7 days.
  • Possibly churned - a customer who has not purchased for a while. This is a potentially lost customer
    • E.g. a customer who used to buy regularly every 14 days has 30 days since their last order. The model, therefore, evaluates that the customer may no longer be active.
  • Churned - the lost customer. This customer has not purchased any product for a long time = low probability of repurchasing
    • E.g. a customer who used to buy regularly every 14 days has 60 days since their last order. The model, therefore, evaluates that the customer is no longer active.

The numbers in the examples above always depend on the overall pattern of behaviour of all customers that the model learns from - this makes it robust to anomalous fluctuations.

When filtering, you select the exact filter value.

Predictive CLV category (Next 365 days)

Based on the Predicted CLV (365 days), Samba classifies the customer into one of the following categories:

  • Low
    • Customers make up the bottom 20% based on the Predicted CLV of all customers.
  • Mid
    • Customers in the middle 60% based on Predicted CLV of all customers.
  • High
    • Customers make up the top 20% based on the Predicted CLV of all customers.

Predictive CLV (Next 365 days)

Using purchase behaviour, Samba can predict how much a customer will spend with you in the next year (365 days into the future). So you can filter your audience by predicted spending in the coming year. In doing so, the model takes into account not only the expected order size but also the likelihood of purchase based on Customer Activity.

When filtering, you can choose whether your resulting audience predicted spending should be Less than (and equal to), More than (and equal to), Equal to or Between the number you select.

Expected purchases (Next 365 days)

Using information about customer buying behaviour, Samba can predict how many orders the customer will place in the future (the prediction covers 365 days into the future).

Using this filter, you can therefore select an audience based on how many future orders a customer will place in your e-shop.

When filtering, you can choose whether your resulting audience should have the number of future orders Less than (and equal to), More than (and equal to), Equal to or Between the number you select.

Days to next predicted purchase

As with the previous filter, Samba can also predict how long it will take for a customer to purchase from your store again. Thus, with this filter, you can select your audience based on how many days are left until the next predicted order.

When filtering, you can select whether your resulting audience should have the number of days until the next predicted order Less than (and equal to), More than (and equal to), Equal to or Between the number you select.

RFM model

RFM is an advanced marketing methodology focused on evaluating the value of your customers.

All the detailed information can be found on our blog in these articles:

RFM segment

Samba assigns a score to each customer based on their activity on your e-shop. It then divides customers into the following groups based on this score:

  • Champions

  • Loyal customers

  • Potential loyalists

  • New customers

  • Promising

  • Need attention

  • About to sleep
  • At risk

  • Cannot lose them

  • Hibernating

  • Lost

When filtering, you select the specific segment your audience is in.

RFM score
Recency score

This filter gives customers a score of 1-5 based on the time since their last purchase.

When filtering, you can choose whether your resulting audience recency score should be Less than (and equal to), More than (and equal to), Equal to or Between a number you select.

Frequency score

This filter gives customers a score of 1-5 based on the total number of their purchases.

When filtering, you can choose whether your resulting audience should have a frequency score of Less than (and equal to), More than (and equal to), Equal to or Between the number you select.

Monetary score

This filter gives customers a score of 1-5 based on the average value of their order.

When filtering, you can choose whether your resulting audience monetary score should be Less than (and equal to), More than (and equal to), Equal to or Between the number you select.

Solvency model

The customer’s credit rating is automatically analyzed based on their purchases with respect to the price score of the purchased products.

Depending on the price range at which the customer purchases products, a Solvency score is assigned to the customer. The model also takes into account the weight of each category so that it is robust to, for example, sporadic purchases of products with a high price score but generally in a lower price category (e.g., complementary goods). Based on this, it is then assigned to one of the three categories:

  • Economy – customers who purchase lower-priced products
    • Customers make up the bottom 20% based on the solvency score of all customers.
  • Standard – customers shopping in the average price range. These are usually the largest category of customers.
    • Customers who are in the middle 60% based on the solvency score of all customers.
  • Rich – your most valuable customer group. These are the shoppers who order expensive products.
    • Customers make up the top 20% based on the solvency score of all customers.

In the case of electronics, for example, Xiaomi phone shoppers would likely fall into the Economy group, whereas Apple iPhone shoppers would fall into the Rich group.

Solvency

Economy/Standard/Rich. When filtering, you select the exact filter value.

Solvency score

When filtering, you can select whether the resulting audience should have a bonity score of Less than (and equal to), More than (and equal to), Equal or Between the value you select.

Other

In audience list

If you have created a custom audience list, you can select a given group of customers using this filter, or in conjunction with other filters, further, edit and segment this segment. When filtering, you select a specific audience list. More information about Audience Lists and their creation can be found here.

In segment

As with Audience Lists, you can filter your resulting Audience List by whether it is in the segment you have previously saved.

A/B test audience split

We have written a detailed, standalone article on the explanation, operation, and audience splitting A/B test, which can be found here.

This post is also available in: Czech

Updated on March 15, 2024

Was this article helpful?

Related Articles