Browse the guide

What to do if a source can't be read

What the Can't read status means and how to get the content in another way.

A source marked Can’t read in Knowledge Base means we reached it but couldn’t get usable text out of it. Nothing is broken, and the rest of your knowledge base works normally.

Why this happens

Web pages

A web page usually fails for one of these reasons:

The text loads with scripts. Some sites render text in the browser instead of sending it in the page source, so the page looks empty to us.

The page requires a login. We can’t access anything behind a login, customer account, or password.

The page is mostly images. If the content is baked into graphics, there may be no text to extract.

The site blocked us. Some hosts block automated readers.

Documents

A document usually fails for one of these reasons:

The PDF is a scan. Scans and photos contain images, not selectable text. If you can’t highlight a line in the PDF, there’s probably no text for us to read.

The file is over 10 MB. Export a smaller version or split it into multiple files.

The file is password protected. Remove the password and upload it again.

The file type isn’t supported. Older .doc files and Excel workbooks can’t be read. Save as .docx, PDF or CSV and try again.

Google Drive imports

An import from Drive can be turned away before it starts. The message next to the file name says which of these it was:

You picked a folder. Open the folder and pick the files inside it.

It’s a Google Form, Drawing, Site or Jamboard. There’s no text in these for us to read.

Google says the file is too large to export. Google won’t export anything over 10 MB. Split the document, or save a smaller copy to Drive first.

You reached your page limit. Files earlier in the batch used up the allowance. Delete sources you no longer need, or move to a larger plan, then import the rest.

Google sign-in expired. Select Choose from Google Drive again to sign back in.

What to do next

Try these in order:

  1. Re-sync the source. Temporary failures happen. Open the row menu and choose Re-sync.
  2. Copy and paste the text yourself. Open the page or file, copy the text, and paste it into a Common question or a plain text file you upload. This is the most reliable fix.
  3. Use a different source. If one page fails, a similar page may work.
  4. Type in the essentials. For example, if a scanned menu won’t import, typing the items and prices often gives better results anyway.

Remove unreadable sources

If you can’t fix a source, it’s worth deleting it so the warning goes away. Select the failed rows and use Delete.

Still stuck?

Email hello@respondhelper.com with:

  • the page URL, or
  • the file you tried to upload

and we’ll take a look.

Related

Last updated 7 August 2026

Still stuck? Email us and a human will reply.