How does it work?

Put simply, when sounds look the same on the lips (as they are spoken) an accompanying handshape or position will make each lip-pattern look different.  For example the sounds /p/ /b/ and /m/ sound quite different to hearing people, but they are indistinguishable by watching the lips.  So people who rely on lip-reading alone have no way of distinguishing words such as 'baby' and 'maybe' or 'pay', 'bay' and 'may'.  When you use CS each consonant sound has a different accompanying handshape so each sound now looks quite different.  Vowel sounds with confusing lip-patterns are clarified by positions.

When people speak they join sounds to make words.  Similarly, with Cued Speech the handshapes and positions are joined to clarify a word.

This film gives you a demonstration of how difficult lip reading is because so many lip-patterns are the same for different words and how using  CS overcomes this problem.