Negative ROMI. How can that be possible?

I was speaking with a bank campaign marketing team and they had a very unusual question for me:  How is it possible that if we’ve set up an test-control group marketing experiment correctly that our response in the test group (which received the promotional offer) could be worse than the control group (which didn’t receive the promotional offer).  This lead to a negative calculated ROMI.  How could that be possible?

Checking the experimental design of your marketing test

Certainly the first thing to check is whether the experiment was properly designed.  Was the test timeframe correct?  Could it have been possible that some holiday or seasonal anomalies crept into the test?  Did the direct mail piece get accurately produced?  Was there a typo or some other error in the design?  Was the sample size properly determined?  Was the test sample properly selected?  All of these are good questions to start off with, but what if the experimental design still passed muster?

Uncompetitive advertising

Then it is possible you did something that most marketers dread and that is to advertise a promotion that consumers don’t find competitive.  That is, interest in the offering was awakened but because the offer wasn’t competitive, those persons receiving the offer checked with the competition and purchased from them.  The competitive offer was better than your promoted offer and upon further consideration by the consumer they decided to purchase the competitive offering.  Whoops - nothing worse than advertising an uncompetitive offering.

The greatest thing about measuring the response to all your marketing activities is that you can go back and now have a way to diagnose what went wrong and what went right. 

If this ever happened to you please let me know.