Is there a way to flatten a pandas column of nested json strings with a varying number of objects?

I have a pandas dataframe that is formatted similarly to this:

What I've been trying to do is unnest the selections field such that each selection occupies its own row with the order_guid as a foreign key to a separate orders table, similar to this:

The main thing giving me trouble is the varying number of nested json elements. My initial instinct was that I could do something like to_json() ->json.loads()->json_normalize, but I can't make that work with the order_guid and the multiple selections. Looping seems silly. It feels like this is a simple enough problem that someone else should've encountered it, but this is the closest I've seen to the issue.

Is there a pandas method for this? Or something obvious I'm missing?

You can first use df.explode() before using json_normalize(). Hope this is what you meant.

df = df.explode('selections').reset_index(drop=True)
result_df = pd.concat([df['order_guid'], pd.json_normalize(df.selections)], axis=1)
result_df['guid'] = result_df['guid'].fillna(result_df['id'])
result_df = result_df[['order_guid', 'guid', 'item']].rename(columns={'guid': 'selection_guid', 'item': 'selection_item'})