Personalized sales

This post is also available in: Czech

Personalized sales are a very powerful marketing tool. Thanks to personalisation, you can tailor this type of promotion to each customer.

Want to personalize your sales?

If you are interested in this functionality, please contact us at sales@samba.ai.

What is Sale

Sale is a marketing strategy that involves one or more product items being offered at a reduced price, often for a limited time. These sales are usually represented by a unique identification number, title, image, URL link, validity information and a list of associated products. The aim is to attract customers and increase sales of specific products.

Data feed

The data feed of sales must generally meet the same requirements as other feeds, see this article. It includes the basic definition of sale including all additional custom parameters.

Supported sale parameters

NameTypeRequiredDescription
SALE_IDStringyesdatabase ID of sale
TITLEStringyesname of sale
URLStringyeslink to sale
IMAGEStringyesURL of the sale image approx. 300px x 300px
ACTIVE_FROMDatetime (as String)nostart of sale in the format YYYY-MM-DDThh:mm:ss.sTZD (more information)
ACTIVE_TODatetime (as String)noEnd of sale in the format YYYY-MM-DDThh:mm:ss.sTZD (more information)
PARAMETERS[NAME, VALUE]noother parameters of sale, NAME does not have to be unique see example
PRODUCTSsee belowyesassociated products (PRODUCT_ID) with the sale

If only the date is filled in for the ACTIVE_FROM or ACTIVE_TO attributes (i.e. the time is missing), it is automatically assumed that

  • ACTIVE_FROM: the action starts at midnight (0:00:00) of the day
  • ACTIVE_TO: action ends at midnight of the day (23:59:59)

If the validity of sale is not filled in, its unlimited validity is automatically assumed, namely:

  • Missing ACTIVE_FROM and ACTIVE_TO: unlimited validity
  • Only ACTIVE_TO is missing: the sale is valid from ACTIVE_FROM indefinitely
  • Only ACTIVE_FROM is missing: the sale is valid until ACTIVE_TO expires

A sale may include more than one product. One product can be covered by multiple sales at the same time. The product ID must match the PRODUCT_ID specified in the product feed.

In case of multiple occurrences of a parameter with the same NAME, all relevant VALUEs are processed. It is therefore possible to search among all values within the filtering.

Example of sale feed

<?xml version="1.0" encoding="utf-8"?>
<SALES>
<SALE>
<SALE_ID>484245864</SALE_ID>
<TITLE>Brand XXX sale</TITLE>
<URL>https://www.myeshop.com/path/to/the/sale</URL>
<IMAGE>http://www.myeshop.com/path/to/the/image.jpg</IMAGE>
<PRODUCTS>
<PRODUCT_ID>52129</PRODUCT_ID>
<PRODUCT_ID>17160</PRODUCT_ID>
<PRODUCT_ID>67721</PRODUCT_ID>
<PRODUCT_ID>51190</PRODUCT_ID>
<PRODUCT_ID>74843</PRODUCT_ID>
</PRODUCTS>
<ACTIVE_FROM>2024-07-01</ACTIVE_FROM>
<ACTIVE_TO>2024-07-31</ACTIVE_TO>
<PARAMETERS>
<PARAMETER>
<NAME>Type</NAME>
<VALUE>brand</VALUE>
</PARAMETER>
</PARAMETERS>
</SALE>
</SALES>

Sale recommendation algorithm

Samba.ai only selects from sales that meet all of the following conditions:

  • URL exists
  • IMAGE exists
  • is currently valid
  • has at least 1 product associated with it (the value of the STOCK attribute etc. is not checked, the existing reference in the product feed is sufficient)

The system uses its knowledge of product recommendations and uses advanced AI algorithms to select tailored sales for each customer based on their buying behavior. If there is no longer any sale in the database that meets the above basic requirements, a personalized email will not be sent.

Sale filters

Similarly to products, sale promotions can be filtered by the user using various criteria. This can be used in particular for user configuration of the recommendation algorithm, where its output can be limited to a certain set of sale actions that contain at least 1 product with a certain property, or on the contrary exclude from the recommendation all selected sale actions.

List of sale filters

Sale end date
When the sale will end.

Sale start date
When the sale will start.

Sale ID
Search by ID.

Title
Name of sale.

Contains products
Filter by the features of the products included in the sale using product filters. In the case of using negation, only those sales that do not contain even one product that meets the given property will be matched.

Days to sale end date
Filter by the number of days remaining until the end of the sale.

Days to sale start date
Filter by the number of days remaining until the start of sale.

Days since sale end date
Filter by the number of days elapsed since the end of the sale.

Days since sale start date
Filter by the number of days elapsed since the start of the sale.

Custom sale parameter
Filter by any custom parameter from the sale data feed.

Campaign and email template

Personalized sales can be sent in one-off email campaigns and also in an email node within Flow campaigns.

First you need to create a template of type “Sale”. The following merge tags can be used within the template:

* = 1, 2, …, 10

In the campaign settings section “Dynamic sales” you can customize the personalization output using sale filters. If no further sale meet the selected user criteria, further actions are recommended regardless of these criteria. However, the basic requirements for the validity of sale and its definition see above still apply.

Evaluation

In addition to the usual statistics available for campaigns, you can also use the Detailed reporting of sales. To activate it, you need to add the tracking parameter rk=sale* to the relevant merge tag of the sale URL, where * must correspond to the numbering of the sale merge tag. So, for example, it is sufficient to provide a link to the sale in this format:

  • {{sale1_url}}&rk=sale1
  • {{sale2_url}}&rk=sale2

After sending a campaign, clicking on the “Show Detailed Reporting” button will display a list of all sent personalization sales sorted by the number of occurrences in personalized emails for each customer.

The following data is available:

  • Image – image of the sale
  • ID – Sale ID
  • Title – Name of sale
  • Sent – How many emails recommended the sale
  • Open – How many opened emails contained the sale
  • Clicked – In how many emails the sale was clicked through

 

This post is also available in: Czech

Updated on July 12, 2024

Was this article helpful?

Related Articles