Many times we need to bring a product into a new market and we try to guess the acceptance rate. It never seems to work out quite right. The trick here is to develop assumptions and monitor those assumptions closely. Be willing to discard those assumptions if they don't play out and exploit the ones that do. As you start to get a track record, the forecasts become a lot more accurate.
What brought this home was an article stating that laptop sales exceeded desktop sales. Back starting in 1990 I did some writing for Sam Whitmore at PC Week. Intel had just announced the SL chip designed specfically for notebook computers. The working title was "Death of the Desktops?" Many people were predicting that within 3-4 years notebooks would be more popular than desktops. I said it was a good 10 years out, if then. It's nice to be right once in a while.