A recent MarketingSherpa article highlighted results from a survey of 400 subscription marketers who rated these tests as delivering the best ROI:
1. Multivariate testing of site pages
2. Email list cleansing to improve quality of responses
3. Usability testing of registration and/or shopping cart
4. Implementing high-end Web analytics
5. Shopping cart design/functionality tests
My reaction: These marketers are wearing Chanel! I'm impressed they're employing more sophisticated techniques like multivariate testing (these people don't have time for A/B testing!) and usability testing, but my only criticism is that they're qualifying email list cleansing as a test to deliver high ROI.
Isn't list cleansing common sense? If you spend less in email delivery costs to reach the same amount of engaged customers (because you're paring down the 20% of your list that are deadbeats), then sure, anyone will achieve a higher ROI.
A better test would be to test offers to re-engage the deadbeats on your list. Test various offers within a set time period and if they don't bite, then hit "delete."