Data Action - Zendesk Ticket Field Mapping - Unexpected Null Values

We’re mapping Zendesk ticket properties to interaction data via a Data Action in an Architect flow - version 8.5.1. Similar to the workaround detailed in a recent community post regarding custom object properties, the issue presents as intermittent null values in Genesys Cloud reporting when Zendesk custom fields containing numeric data are processed. Zendesk stores these as strings, and the Data Action’s JSON parser seems to interpret empty strings as explicitly null, which then breaks the numeric aggregation in GC reporting - unlike Zendesk, where empty strings are treated as zero.

The Data Action’s JSON payload looks like this: {"ticket.priority": "{{ticket.priority}}", "ticket.customfield123": "{{ticket.customfield123}}"}. We’ve tried string-to-number conversion within the Data Action using Javascript, but it’s unreliable - sometimes the values come through, sometimes they don’t. The Zendesk API version is v2. Is there a more consistent way to handle this data transformation, or are we stuck pre-populating default values in Zendesk to prevent empty strings in the first place?

hello, i think maybe is problem with data type conversion. the parser in Data Action is strict - empty string is not number, so is null. try to use Zendesk function to replace empty string with “0” before send to Genesys Cloud - 2024-06-26T10:30:00Z. i am sorry for newbie question.

The suggestion above - using a Zendesk function to replace empty strings with “0” - is a good start, but it’s sidestepping the REAL issue.

Data Actions aren’t designed for this kind of string manipulation. You’re better off handling the conversion in a Lambda function BEFORE it hits the Data Action. That way, you control the parsing and avoid unexpected nulls.

Remember, you’ll still need to account for rate limits and potential data type overflows - ESPECIALLY with numeric conversions. ALWAYS validate the input.