Using Ad Curves to Measure ROI
If you're like many retailers who heavily rely on a printed catalog, you’ve probably been faced with a difficult choice. Maybe you want to decrease the number of pages in your catalog, from, say, 124 to 104. Or you want to double your email contact frequency, from one campaign per week to two campaigns per week. To that, I say you have three choices.
1. You can “test” your strategy. Create two versions of your catalog — one with 124 pages, the other with 104 pages. Test your email contact strategy to a limited audience for three months to see what impact it has.
2. Try doing what most of us would do: Guess what impact your decision will have on demand, then simply execute the strategy you choose, hoping to somehow estimate what happened on the back-end via company reporting.
3. Employ 'ad curves' to your decision. An ad curve is simply a mathematical way to estimate what might happen if you employ different advertising strategies.
The most common ad curve is square root. Here’s what you do: Recall our email example, where you increase from one contact per week to two. Take the ratio (2/1 = 2.00) and raise it to the power 0.50 (0.50 = the square root of a number). In your case, (2.00 ^ 0.50) = 1.41.
The ad curve suggests that the demand you get from email marketing will increase by a factor of 1.41 if you double the number of contacts.
The same strategy works for the catalog example. If you go from 124 pages to 104 pages, demand changes by ((104 / 124) ^ 0.50) = 0.916. In other words, a 16 percent drop in pages yields an 8 percent drop in demand.
When you don’t have time to test, or you simply don’t want to test a strategy, give ad curves a try. While not nearly as accurate as conducting a live test that proves what the impact on demand will be, ad curves represent a credible alternative to simply executing a strategy without understanding the consequences.