Stat. Sig of a dollar amount

Need help measuring the statistical significance of when a customer responds to a solicitation, & the size of their purchase.

This is a direct marketing scenario where we mail 5,000 solicitations to customers & test a different package to another 5,000 customers. (We split our 10,000 customers randomly into the 2 groups)

I can easily measure the statistical significance of the # of Responses. 200 respondents in population A & 150 in population B. But I also need to factor in the size of the purchase. Avg gift of population A was $25 & the Avg gift of population B was $45.
Dollar amounts are just numbers and can be treated as such. You could take the two samples who gave (150 and 200) and compare them directly with t-tests. Alternatively, you could take all 5000 in each sample and do the same computation. The company might be interested in an average-dollars-per-contact figure, which would 1.00 for Group A and 1.35 for B. So, in a simpleminded way, the Group B methodology has a 35% advantage.