The combination of a chat interface with source grounding makes NotebookLM an
excellent platform for help documentation or as a guide for new users, customers,
or colleagues trying to understand a product or an organization. NotebookLM can not
only answer factual questions based on the sources you supply, but it can also help
with freeform queries. This Welcome To NotebookLM notebook is a great example of
how this kind of approach can work in practice. You can ask NotebookLM very
specific questions, like “How many words can I include in a source?” But you can
also ask something like, “I’m trying to put together a legal brief that draws on
several existing court decisions. How can NotebookLM help with that?”
Creating a help center is really just a matter of assembling documents that explain
your product, service, or organization in a clear and comprehensive way. As always
with NotebookLM, you want to make sure those documents are formatted so that they
can be “read” by the AI. Straight text-based Google Docs are always the best bet.
FAQs can be a great way to “teach” NotebookLM about how your product or
organization works, though ideally you should format the FAQs so that the question
is in the same paragraph as the answer. But NotebookLM is smart enough to answer
questions even if you haven’t supplied FAQs in your sources.
Once you’ve assembled the sources, create a notebook and upload the sources. At
that point, NotebookLM will be able to answer questions based on information in
those documents. You can share the Notebook with others in “Viewer” mode, which
will allow them to ask questions, follow suggested questions, and read the sources,
but they will not be able to alter the collection of sources you’ve curated or take
notes.