Tally Webhook → How to correctly calculate average from multiple LINEAR_SCALE fields in Make?

[Jochen_Henke]

What is your goal?

Automation

What is the problem & what have you tried?

Hi everyone,
I’m receiving form responses from Tally via a Custom Webhook in Make.
Inside data → fields, I have multiple LINEAR_SCALE questions with the following labels:
sinn_bedeutung
sinn_identitaet
sinn_standhaftigkeit
sinn_langfrist
Each field contains a numeric value (e.g., 1–5).
I want to calculate the average of these four values inside a Tools → Set variable module.
My current expression is:
Code kopieren

(number(first(map(2.data.fields; value; label; “sinn_bedeutung”)))) +
(number(first(map(2.data.fields; value; label; “sinn_identitaet”)))) +
(number(first(map(2.data.fields; value; label; “sinn_standhaftigkeit”)))) +
(number(first(map(2.data.fields; value; label; “sinn_langfrist”)))) / 4
However, the output returns the full expression instead of the calculated numeric result.
From the webhook output I can confirm:
The labels match exactly
The values are numeric
The fields exist in the array
What is the correct way to extract and calculate the average from these labeled fields inside an array in Make?
Thanks in advance!

Screenshots (scenario flow, module settings, errors)

[Dominique_Djagbassou]

Use the Average function in make and map the value table to calculate the average

I hope I have answered your question correctly.

[Stoyan_Vatov]

You are not using functions here, that’s just free text. Select the functions from the menu instead and map the variables inside them.

Also it’s parseNumber().

[Jochen_Henke]

Hi everyone, thanks for the feedback. When I try to calculate the average here, it’s not displayed correctly; I’ve tried both using the symbols and the text. What am I doing wrong?

[Jochen_Henke]

Ok, i got it Chat gpt ist a ….

[Azariah786]

If you’re receiving multiple LINEAR_SCALE values from a Tally webhook in Make, the easiest way is to first collect those values into an array and then calculate the average after the bundle processing step. I usually map the scale responses into numeric fields and then divide the total by the number of responses.

One small tip: make sure empty fields are filtered out first, otherwise the average can become inaccurate.

I’ve dealt with similar logic before when building small calculation tools and forms, especially where multiple inputs need to produce one final result. Even simple utilities like age or number calculators follow the same principle — gather the values, clean them, then run the final calculation.