Group 1 receives one catalog, and each of two weekly e-mail campaigns.
Group 2 receives one catalog, but does not receive any of the weekly e-mail campaigns.
Group 3 does not receive the catalog, but does receive each of the two weekly e-mail campaigns.
Group 4 does not receive the catalog, and does not receive either of the two weekly e-mail campaigns
The table at the bottom of this article illustrates the findings of the experiment. Which strategy would you employ (the strategy from Group 1, Group 2, Group 3, or Group 4), and why? Or, do you recommend a different strategy? Are you missing any information that you need to make this decision?
|Group 1 = Catalog + E-Mail||$4.39||$5.94||$10.33||$1.56|
|Group 2 = Catalog||$4.62||$5.71||$10.33||$1.58|
|Group 3 = E-Mail||$0.19||$4.46||$4.65||$1.14|
|Group 4 = No Marketing||$0.20||$3.99||$4.19||$1.05|
At first pass, I like option 1. The top line is tied for best, and while profit is slightly lower, I wouldn't want to abandon the email program. Email is an important part of the marketing mix. A contact strategy test with emails is difficult to measure, as our email messages vary widely. (what if the emails were poor performers, or not hard sells?)ReplyDelete
I'm most intrigued with the Group 2's 'Phone Demand.' If the catalogue is unsuccessful at driving web demand - maybe we should consider having the analysts develop a phone propensity model and send these folks the catalogue. This speaks to catalogue strategy... 1.) Send the best customers a catalogue or 2.) Send it to customers who display the biggest incremental lift.
One more thought - what to do with Group 4? The $4.19, to some degree, is an accumulation of prior marketing. Should these dollars really float outside of any marketing vehicle's "claim" and therefore not be budgeted against? We consistenly see > 30% orders come in without a marketing source, and generally we've allocated this revenue to the catalogue through a match-back process.
Simon, you hit upon an important point in evaluating Group 4!!ReplyDelete
So often, we are eager to allocate business back to a marketing activity.
Frequently, those sales simply "happened". No marketing activity caused them.
The important thing to do is to relentlessly test marketing vehicles. Some marketers will withhold marketing from customers for six-months at a time, to eliminate the impact of marketing "spilling over" into the test period. When you do this, you still see organic sales happening online. Those sales cannot be matched-back to any marketing activity, when designing six-month long tests.
When viewed in that light, the profitability of marketing is not as good as originally thought to be.
Way to identify a key issue, Simon!!