prev thread from support - This behavior is due to how we manage column source dependencies. Each column is tied to a specific data source, and when you imported the CSV, the dependency was set to that particular column from the CSV. Later, when you parsed your company search data into the same columns where the CSV data was living, that original dependency was overwritten.
Think of each column like a power outlet that can only be plugged into one power source at a time — if you switch the plug, the original power source gets disconnected. Similarly, switching the data source reassigns the dependency and overwrites what was there before.
This is a current limitation of the product which I will flag to our product team for addressing in the product. As a workaround, I recommend merging the outputs into a single set of columns (f.e. company name csv, merged with company name from the company search, into a single column using our merge functionality) when working with multiple sources — especially when one of them is a CSV import.
