Having to regex a list of JSON text records with 1000-2000 items each can create a very noticeable delay in the frontend, with the consequent degradation in UX.
Having the data properly structured in DB allows for faster searches, especially if they are frequent and table indexes are created internally by Bubble. Also, allows for search constraints, that reduce the number of returned results and so the WU spent.
How you design the system depends on how frequently you expect the user will search through the DB records and how big you expect the table is to become. The bigger the latter two factors, the more reasons you have to save it all structured in DB.
For your particular use case, as you mentioned that the user is performing LinkedIn searches under the hood, it sounds to me like volatile searches, i.e. records consulted once and never again. So, using JSON and regex, as long as you properly handle the search delays in the frontend, sounds like a suitable solution to me.
1 Like