Truth is, I don't think the economic cycle makes any difference in setting realistic goals - the core principles remain the same:
1. Hire people who have a track record in achieving their goals.
Perennial under-achievers don't change - hire people you can trust will always give their best.
2. Set goals early, then revise the goals regularly in the light of hard data on actual performance.
Setting goals as a fixed line in the sand, then 'turning up the volume' in an attempt to magically cause your sales people to deliver those goals is a road to nowhere. Stubborn refusal to revise goals -worse, a belief that revising goals is weakness, is simply airheaded. [See recent developments in the financial markets for examples.]
Sales goals should be reviewed monthly and reviewed (up or down) in the light of whatever new information you have.
3. Use data for 70% of the goal-setting process, intuition/judgement 30%
4. Build goals from the bottom up, not the top down.
The people on the front line know more than anyone what is achievable. If you think you can't trust them, or that they'll sandbag you, go back to point 1.
5. Reward success, penalize failure.
Too many compensation schemes favor the first and under-play the latter. There should be consequences for sloppy forecasting - being over-optimistic just to please the manager is a terrible drain an organization's success, so put penalties in place as well as rewards.
These are just as important in an 'up' economy as in a down one.
