Here is a file which shows how to process this type of data. You will have to type the brands into the Brands table but since it produces results immediately, you can see ones you still need to enter. Once you have the brands, getting the portion of the product which is NOT the brand becomes simple.
I hope this is helpful to work through it. Also, if you happen to have a Brands list somewhere you can use it instead of the related table. As you enter the Brands and the cProductOnly portion fills in, you can easily perform a find on cProductOnly for ERROR to find those which need to still be entered. Note that if there is potential of two matches, it produces an error as well in which case you must manually decide. An example would be if there is a Brand called 'Hungry" and there is a brand called "Hungry Jacks" you will get ERROR because there will be two potential matches for a product called Hungry Jacks Pancakes.