Drive search is a filename search with a full-text index bolted on, which is why "the deck with the pricing model in it" is so hard to find and "Untitled document (3)" is so easy to lose. The content is all there. The way in is the problem. The Google Drive connector is part of askFinz Knowledge, which is still rolling out; here is what it will do, and what it will not.
Find a document by what is inside it
Describe the content — the argument, the number, the client — rather than the filename, and get the document, named and linked so you can open it in Drive.
Read Docs, Sheets and Slides as content
A point made on slide 14 or a note in a sheet's comment column is content, not an attachment. Those are searched as text, so a deck is as findable as a doc.
Keep everyone's access exactly as it is
Connecting a shared drive does not widen who can see what. Each person's results are limited to the files they could already open.
What is copied, and what is not
What crosses, and what stays in Drive
Read
the text of documents in the folders and shared drives you select.
Written back
nonenothing. No files created, edited, moved or shared.
Stored
a searchable copy of that text, plus a link back to the file.
Never read
folders you did not select, and files whose access you do not have.
What you will be granting
Read-only access to the folders and shared drives you pick, with no permission to create, change or share anything in Drive.
Setting it up
Planned — not available yetWhat connecting it will look like once the connector arrives.
- 01
Sign in with Google from Knowledge.
- 02
Pick the folders and shared drives to include. Nothing else is requested.
- 03
Decide the mode: your own Drive, shared drives, or both.
- 04
Set how often to re-check for changes.
Limits, honestly
- One-way and read-only. Drive stays the place files live.
- Answers reflect the last refresh, not the edit you made a minute ago.
- Very large media files are skipped; the documents around them are read.
- Files in the trash and version history are out of scope.
- Scanned PDFs are read only where the file already carries a text layer.