Individual Feedback Data Formats

About Individual Feedback Data Formats

In XM Discover, individual feedback format is used for documents that present data as a single-row, or “flat” object. Each document may contain several verbatim fields, but usually these come from a single author (for example, when a survey contains more than one open-ended question).

Typically, an individual feedback document would contain a number of structured and unstructured data fields that represent a single response from a customer (for example, a survey response or an online review). Structured fields may contain dates, numbers, or text data with a high degree of organization (such as names of brands, cities, or hotels). Unstructured fields contain notes, comments, or responses to open-ended questions.

You can upload individual feedback via the following formats:

• CSV
• XLS or XLSX (Microsoft Excel)
• JSON

CSV and Excel Formatting for Individual Feedback

This section covers formatting for individual feedback for CSV and Excel files. The formatting and requirements for both file types are the same.

• Rows setup: Make sure the first row of your spreadsheet contains column names. These names will be used as attribute names and verbatim types. Each subsequent row is treated as a separate document. You can have a maximum of 100,000 rows in a single file.
  Attention: Do not include any empty rows between entries.
• Column names: When XM Discover converts column names to attribute names, it automatically adjusts them to match the supported format by removing unsupported characters and truncating long column names. Follow the below guidance:
  • Key names should only contain English letters (A-Z, a-z), numbers (0-9), underscores (_), dashes (-), square brackets [], and white spaces.
  • The ideal key name length is 21 characters or shorter. Key names should not exceed 99 characters as longer names will be automatically truncated.
• Maximum file size limits: The following file size limits are in place:
  • Maximum number of rows: 100,000 rows.
  • Maximum number of columns: 100 columns with attributes and 20 columns with verbatim feedback.
    Qtip: This is a soft limit to allow for project growth.
• Maximum field lengths: The following limits apply to field values:
  • Text attributes: 900 characters. If the value exceeds this limit, it is automatically truncated.
  • Multivalued attributes: 4,000 characters. If the value exceeds this limit, it is automatically truncated.
  • Verbatim fields: 512,000 characters. If the value exceeds this limit, the record is rejected.
• Required fields: The following fields are recommended in your file:
  • Date field: Having a date field in your data is crucial for reporting or analysis. The recommended date format is YYYY-MM-DDThh:mm:ssZ.
    Qtip: If your file does not have a date column, you can generate dates by following the instructions on Setting a Specific Document Name.
  • Unique ID: If there is a column with unique values for each row, you can use it as Natural IDs for your documents. This field is optional, but highly recommended since it lets you manage duplicate documents.
    Qtip: If your file does not have a unique ID column, you can generate them automatically by following the instructions on Automatically Generating Natural IDs.

JSON Formatting for Individual Feedback

This section covers JSON formatting requirements for individual feedback.

• Key names: When XM Discover converts key names to attribute names and verbatim types, it automatically adjusts them to match the supported format. Key names should follow these guidelines:
  • Key names should only contain English letters (A-Z, a-z), numbers (0-9), underscores (_), dashes (-), square brackets [], and white spaces.
  • The ideal key name length is 21 characters or shorter. Key names should not exceed 99 characters as longer names will be truncated.
• Objects: Your JSON file may contain either a single object or multiple objects.
  Attention: There is a limit of 100,000 objects in a single JSON file.
• Maximum field length: Different attributes have different value length limits:
  • Text attributes: Maximum 900 characters. If the value exceeds the limit, it is automatically truncated.
  • Multivalue attributes: Maximum 4,000 characters. If the value exceeds the limit, it is automatically truncated.
  • Verbatim fields: Maximum 512,000 characters. If the value exceeds the limit, the record is rejected.
• Required fields: The following fields are recommended in your file:
  • Date field: Having a date field in your data is crucial for reporting or analysis. The recommended date format is YYYY-MM-DDThh:mm:ssZ.
    Qtip: If your file does not have a date column, you can generate dates by following the instructions on Setting a Specific Document Name.
  • Unique ID: If there is a column with unique values for each row, you can use it as Natural IDs for your documents. This field is optional, but highly recommended since it lets you manage duplicate documents.
    Qtip: If your file does not have a unique ID column, you can generate them automatically by following the instructions on Automatically Generating Natural IDs.

Example

Below is an example JSON payload for individual feedback.

[
{
"ID": "1001",
"SOURCE": "HotelSurveys",
"CITY": "New York City",
"HOTEL": "Sheraton New York Hotel and Towers",
"AUTHOR": "Richard",
"REVIEW": "Hosted a meeting here and had a very fine experience. Would recommend highly.",
"GENDER": "M",
"LOYALTY LEVEL": "Gold",
"AGE": 51,
"CHECK-IN DATE": "2019-01-01T12:00:00Z",
"SCORE": 9
}
]