Importing Order Data

Overview

In order to utilize Recommendations prior to the real-time data collection, order history data can be provided to the Recommendation engine. The order history data does NOT need to include any customer identification details. This document will describe the necessary feeds that are required.

Note this process can be run on an ongoing basis. This would be helpful in a scenario where significant sales activity occurs offline, but you would like to capture that information to help target recommendations on your website. You could continue to upload the file periodically for example. If you would like to automate this process by sending the files to an SFTP server, HawkSearch could run a process to pick it up and add in an automated manner. Please reach out to your HawkSearch contact to discuss.

Feed Formats for Order data

The Recommendations engine will utilize your website’s order history, broken out into two separate files, and representing activity from the last 30 days.

orders.txt

This file represents the summary of the orders from the last 30 days.

order_items.txt

This file represents the items contained within the orders defined in the orders.txt file.

Please review the data feed format for each file. If a particular column does not appear to apply to your data, please bring this to the attention of your Hawk Search implementation representative for further clarification.

File Format Requirements

File Size Limit

The maximum file size that can be imported is 4096 kB. Larger files needs to be spitted and imported separately.

Flat-File Format Properties

Item	Value
Encoding	UTF-8
Column Delimiter	Tab
Column Headers	Required, case insensitive Column Headers are allowed in plain text or in quotation marks. For example: OrderId, orderid, "orderid" or "OrderId"
Row Delimiter	Unix Format (\n as a row delimiter)
File Name	Lowercase and should match names specified above

Orders Data Feed

FILE NAME: orders.txt

The Orders Data Feed consists of order summary data from the site over the last 30 days, provided in a single file named orders.txt.

Required Data Columns

Name	Type	Data Description	Required	Sample Value
OrderId	String	Alphanumeric ID value representing the order identifier. This is usually generated by your ecommerce system and is unique.	Yes	AA123
OrderNo	String	Alphanumeric value representing the order number, if it is different than the OrderId. If it is the same, please use the OrderId in this field. Typically unique.	Yes	AA123
SubTotal	Decimal	The total of items in the order before taxes or shipping costs.	Yes	12.34
Tax	Decimal	The amount of tax associated with this order.	Yes	1.23
Total	Decimal	The order total, or the amount that includes shipping, taxes and all associated costs.	Yes	13.57
Currency	String	The currency format used by the order (USD, EURO, CAD)	Yes	USD
CreateDate	MM-DD-YYYY	The date the order was submitted and completed in UTC	Yes	01/23/2014
MemberId	String	The unique identifier representing the customer. This is usually generated by your ecommerce system. This usually will not contain details such as the customer name.	Yes	ZZ1234

Order Items Data Feed

FILE NAME: order_items.txt

The Order Items Data Feed will contain the items associated with each order defined in the Orders Data Feed. This items should be grouped by its order identifier, and provided in a single file named order_items.txt.

Required Data Columns

Name	Type	Data Description	Required	Sample Value
OrderId	String	Unique Alphanumeric ID. This should correspond to the OrderId column specified in the orders.txt file	Yes	AA123
ItemId	String	Alphanumeric identifier for the item. This should be the unique number or internal identifier used track item. This is typically ItemId, SKU or ProductID. This will be mapped to your primary key field in HawkSearch.	Yes	AABB123
Price	Decimal	The per unit item price.	Yes	123.45
Quantity	Integer	The number of items purchased in this order.	Yes	123

Sample Files

order_items.txt

orders.txt

How To Upload Offline Orders

The upload option is within the Workbench in the Recommendations section.