hi all,

I have a 200 person sales force selling product A. However, we acquired another services company which provide services to product A. Now our sales force is expected to sell product A and also try selling as many services as possible (If a sales person is selling X units of products he/she can sell max. X units of services as well - we call this as cross-sell rates)

1) In Q1, we had no incentive program to reward sales person to sell service when they sell product A. However, few sales people still went ahead and sold services to customers when they were selling them product A (with varying degrees of success). Average cross-sell rate was 22.5%

2) in Q2, we introduced an incentive program which rewarded sales people who were cross-selling services with product A (higher the % cross sell; higher was the payout). Average cross-sell rate was 48%

3) In Q3, we tweaked the incentive. Results are just in and average cross-sell rate is 46%.

Now, people are claiming that Q3 incentive plans were flawed hence the lower cross-sell rate but i want to statistically understand if this statement holds?

The data i have is, for each quarter, for every sales person - units of product A sold, units of services sold, average unit value of services sold.

Which test can i undertake to prove/disprove that dip in cross-sell rate was random and can be explained by chance?

Thanks
Statsnovie