You are correct; it does sound like you need to add some identification conditions.  I can't offer too much assistance here as I don't know what the documents look like and what any of the pages contain, but hopefully I can guide you in the right direction. 

When setting up identification conditions, you can look for a certain area of text on the page that will not be included on any other pages (a document title, perhaps).  You can Zone OCR that section of the page and use an identification such as "Zone 1 contains [text]".  See below for an example:

If you are unable to find any text that is unique to the first page, there are many other ways to set up identification conditions.  You can read all about them in the help files.