As a campaign metric, ROI beats response rate every time

by PaulRaca on September 25, 2009

Paul Raca

Paul Raca

Too often a simple response rate calculation is used to evaluate whether one campaign or test cell has performed better than another.  It’s not surprising – response rate is so easy to understand…What percent of those I contacted got back to me or took me up on my offer?  There are even “standards” or benchmarks for the typical response rates to expect from prospects or existing customers that help me feel good about the rate I observe.

However, the calculation of ROI, or Return on Investment, brings in additional “confounding” but very important factors:  cost and revenue.

Cost:  For my money, cost absolutely needs to be part of the evaluation because I am allocated marketing budgets  in terms of dollars, not pieces.  So if my test cells are the same size – say 25,000 pieces in each – I need to account for the differences in cost of producing the pieces for each cell.

If I can send a piece that costs less than my champion package without seeing a decrease in response rate, I’ve got a new champion.  In fact, even if I observe a DECREASE in response, I may have a new champion if I have a GREATER decrease in cost.  With the money I save, I can mail more pieces and my overall number of responders is higher for the money I spend.  In this way, a metric such as Cost per Acquisition that accounts for cost (campaign cost/# of responders) is a better metric than Response Rate.

Revenue:  Measuring responses is often not enough.  All responders are not necessarily created equal when translated into revenue.  For instance, if the offers are different between my test cells, I’ll want to account for that difference in cost in measuring success.  I am likely to get more responders with an offer of 10% off, but the question is whether the increase in response makes up for the 10% cut in resulting revenue.

By the same token, if I send a coupon, I’m hoping that the responder will purchase more than just the discounted item.  So, my coupon responders may have a higher average spend than my other responders and I want to account for that in the test too. I need to account for the increased value of these new customers.

A simple ROI calculation does it all:  Test cell revenue/test cell cost. Run the numbers on your next campaign and you may find some fascinating new insights about the results of your programs.

About the Author:

Paul Raca is the Vice President of Marketing Analytics at SIGMA Marketing Group.  Connect with Paul on , or follow him on .

Leave a Comment

{ 1 trackback }

Next post: