I’m using n8n v2.7.4 with the Google Firebase Firestore node v1.1.
My problem: I want to create a document in Firestore from a JSON object, but the node does not save the fields properly. Instead of separate fields, everything gets merged into one key and all values are null.
Unlike Zapier, the n8n node doesn’t give me a proper table to fill fields. I tried:
Comma-separated key=value in Columns / Attributes
Multi-line input
{{$json}}
…but the result is always null values or errors like columns.split is not a function.
I want to create Firestore documents dynamically from JSON with all fields correctly saved. Has anyone solved this in n8n v2.7.4 Cloud?
The current result in Firebase:
ResponseID=68qbE1Ynull
I’m not familiar with Firestore, but from the node perspective, it seems, that you will need to specify the comma-separated list of column names, where in the input you have them as keys.
E.g.
Hello @Alex_Schtelbe ,
try this :
In the Columns field, simply list the exact names of the keys you want to keep, separated by commas. Do not add values or equal signs.
Copy and paste this exactly:ResponseID, firstName, isConfirmed, Turnover, company, createdAt, email, jobTitle, lastName, phone
try inserting an Edit Fields node between your Tally Trigger and the Firestore node. Create the clean fields you want on the left, and drag the specific values you need from the Tally output on the right.
You are incredibly close—your “Edit Fields” node is successfully generating the clean JSON , but the Firestore node is failing because the Columns field strictly expects a list of key names, not their values. By entering expressions like {{ $json.Name }} in that field, you are effectively asking n8n to find a key literally named “Alex,” which does not exist, resulting in null values.
To fix this replace the entire content of the Columns / Attributes field with the expression {{ Object.keys($json).join(',') }}; this will automatically grab all your clean keys (id, Name, Surname, email...) and map them correctly to their values, resolving the error instantly.