Outfit Emails

  • 3 minute(s) read
Prev Next

Overview

Outfit Emails is a machine learning-powered recommendation feature that suggests complete outfits in email messages based on a specific garment that has been purchased. It is typically used in post-purchase emails to increase customer engagement and drive additional revenue by showing how a purchased item can be styled in different ways.

Each email typically features two distinct outfits centered around the same garment. If a customer buys multiple items, you are encouraged to select the highest-valued item as the basis for the outfits. Your email tool passes the relevant item information to the Outfit Emails service, which then generates and displays personalized outfit recommendations.

How it works

  • Personalized outfit recommendations: Uses machine learning to generate outfits around a defined initial item, usually a purchased garment.

  • Two outfits per email: Supports up to two distinct outfits in a single email, each with its own set of recommendation slots.

  • Multiple recommendation slots: Supports up to 6 recommendation slots per email. In the example, each outfit uses two slots.

  • Consistent identifiers per outfit: The :sku and :send_id values must be the same for all slots within a single outfit to ensure that all items belong to the same outfit.

  • Unique Send ID per email: Each email must have a unique :send_id so that fresh recommendations are generated for every new email sent to a customer. For post-purchase emails, this is typically the order ID.

  • Customer-level personalization: Recommendations are generated for a specific :customer_id, which must match the identifier used in your customer journey tracking and/or transaction feed.

  • Stable recommendations per email open: Recommendations are generated the first time the recipient opens the email and are then persisted for future use. Subsequent opens of the same email show the same recommendations.

Configuration and Parameters

HTML template structure

To use Outfit Emails, you work with the creative team to align the design, then add a specific HTML template to your email tool. The email tool must pass the garment on which the outfits should be based.

Example HTML for two outfits with two slots each:

<a href="https://dressipi-emails.com/outfits/:namespace_id/:send_id/:sku/:customer_id/link/outfit/1/slot/1">
  <img src="https://dressipi-emails.com/outfits/:namespace_id/:send_id/:sku/:customer_id/image/outfit/1/slot/1">
</a>
<a href="https://dressipi-emails.com/outfits/:namespace_id/:send_id/:sku/:customer_id/link/outfit/1/slot/2">
  <img src="https://dressipi-emails.com/outfits/:namespace_id/:send_id/:sku/:customer_id/image/outfit/1/slot/2">
</a>
<a href="https://dressipi-emails.com/outfits/:namespace_id/:send_id/:sku/:customer_id/link/outfit/2/slot/1">
  <img src="https://dressipi-emails.com/outfits/:namespace_id/:send_id/:sku/:customer_id/image/outfit/2/slot/1">
</a>
<a href="https://dressipi-emails.com/outfits/:namespace_id/:send_id/:sku/:customer_id/link/outfit/2/slot/2">
  <img src="https://dressipi-emails.com/outfits/:namespace_id/:send_id/:sku/:customer_id/image/outfit/2/slot/2">
</a>

Configuration rules:

  • SKU consistency: :sku must be the same in all href and src attributes within an email to keep all items in the outfit aligned with the chosen garment.

  • Send ID consistency per email: Use the same :send_id for all slots in a single email. Typically, for post-purchase emails, this can be the order ID.

  • Slot 1 as base garment: Slot 1 in each outfit always corresponds to the garment referenced by :sku.

  • Maximum recommendations: You can add more slots if you want to show additional recommended items in the email, up to a maximum of 6 slots per email.

  • Region domain: If you are based in the Australia region, replace dressipi-emails.com with au.dressipi-emails.com in all URLs.

URL parameters

The URLs in the href and src attributes use the following parameters:

Parameter

Value

Purpose

:namespace_id

string

Mapp Fashion-provided identifier for your namespace.

:send_id

string

Unique identifier for this email. For post-purchase emails, this is typically the order ID.

:sku

string

SKU of the garment used to build outfits. It must be the same for all href and src attributes in the email, so all items from the outfit.

:customer_id

string

User identifier for whom the recommendations are generated. This must match the customer_id used in your customer journey tracking and/or transaction feed.

:outfit_id

integer, either 1 or 2

Identifies which outfit to show. Up to two outfits per email are supported.

:slot_id

integer, maximum 6

Identifies the recommendation slot within the outfit. In the example, slots 1 and 2 are shown. You can add up to 6 slots per email.

Recommendation behavior

  • When a customer opens an email for the first time, Outfit Emails generates fresh recommendations and stores them.

  • Every subsequent opening of the same email shows the same stored recommendations as on the first open.

  • The lookup for recommendations uses both :send_id and :customer_id.

  • Each new email sent to a customer must use a new :send_id so that new recommendations are generated for that email.

  • The same :send_id value can be reused across different customers if needed, because the combination of :send_id and :customer_id is used to retrieve recommendations.