Pandas groupby two columns and create a new column based on difference in days from third column

I have a dataframe with creation_timestamps and personal_id's. Each personal_id has one or more application_id's. An application_id can have a bunch of rows, but they all have the same creation_timestamp.

I want to create a column that indicates the days between two application_id's for a given personal_id. All the offer rows within the application_id should show the same days between the two application_id's.

Here's the code I've tried to modify in a bunch of ways:

# Calculate days between consecutive applications for each personal_iddf['days_between_applications'] = (    df.groupby(['personal_id'])['creation_timestamp']    .diff()    .dt.days)

Here's a preview of the current result:

This is what I'm trying to achieve:

I've tried to forward fill, sort and then forward fill, groupby personal_id and application_id etc, but nothing has worked.