I like your approach - it's cautious which seems like a plus!
In our planning, we've been using stats in part - there are a lot of fantastic blogs out there from people who have successfully funded and who talk about their approach. One such blog that I found helpful was this:
http://craigmod.com/journal/kickstartup/He goes into a lot of detail, including the stats for various reward tiers from the top campaigns. It's a bit dated so you'd have to crunch the numbers if you wanted a current set, but based off of his numbers, we set up our reward tiers and made some guestimates as to what percentages we would have at each tier.
Of course it also probably depends on the specific rewards. Anything tangible with a production cost and time you'll need to factor that in to ensure timely delivery. On a related note it's good to make sure you don't have any reward tiers that will leave you owing more than you'd like. I imagine it would be sad to promise a large reward for pledges of $2500 only to then exceed beyond your wildest expectations, get a lot of $2500 pledges, but have your net from those pledges be small. We had one tier planned that we realized we'd be spending too much of the pledge on the reward, so modified it.