In this lesson, we walk through how to run tests to see if your call to action/copy/imagery is working for a segment in response modeling. We then show how to target effectively and efficiently using historical response modeling, or receptivity analysis. Finally we estimate campaign ROI to ensure we are making the best use of our marketing dollars.
In this example we are testing copy/creative to the #Country group which we identified as a high spending segment earlier in the Analyze step. We pull a small subset of customers in the #Country segment family and run a control group, a test group A that received a certain creative and copy targeted toward the #Country segment, and a test group B that received a different set of imagery and copy.
These imagery and copy sets were driven from insights we gleaned from the segment taxonomy. We set our baseline target at 20% click through rate which is the typical historical click through rate for this type of email. We found the imagery and copy from Group A provided a 12% lift over baseline with a clickthrough rate of 32%. Now we can refine further or feel comfortable releasing that email campaign to the other 367k contacts that are in this segment.
Companies with previous campaign data can analyze previous campaigns by which segments responded and which didn’t. We can then use that to build efficiency into our campaigns by only contacting the specific set of customers that respond best to a specific type of offering. In this example ABC Insurance (our fictitious company) used this to expand a line of premium insurance products with historically low conversion rates (.82%). They have a contact database of ~ 5 million contacts, based on this conversion rate they could expect ~ 250m in total potential insurance sales. But they do not have the budget to inundate their entire database with an ad that is only relevant to a small portion.
ABC insurance needs to target effectively but capture as much of that 250m as they can. By analyzing response rates by segment we find that the .82% response rate is not equally distributed. Some segments respond at a rate of 4.5% whereas many do not respond at all. By sorting the top 7 segments in terms of response ABC Insurance did a tiered approach, only hitting the segments that responded the best to this insurance product first and on down the line until they met their budget for this campaign. Using this method ABC insurance slashed 81% of their total advertising cost while experiencing a 63% retention in total sales.
Using historical data on conversion, we can find historical conversion rate per segment. We also have average spend per segment. All we have to do to estimate campaign ROI is multiply the (Historical Response Rate * Total Contacts In This Segment * Average Spend For This Segment = Estimate ROI).
Now that you have analyzed customers, created personalized campaigns, and estimated ROI you’ll want to Activate these segments via Social, Google, or Display advertising. Head on over to the Facebook Activation or Google Activation tutorials to learn more.