The Shots You Get to Take

At Grouparoo, we have been interviewing a lot of marketers. The overall learning is that it's a hard job. The biggest reason is that they need data to make their campaigns work and do not have the means to get that data. Basically, they need Engineers to prioritize writing code to get the data into the tool they are using. That rarely happens.

Let's say your job is to move the retention number up and to the right. This means that you need to find ways to get people coming back to your site and buying stuff more frequently. I bet you have lots of ideas.

The best way to achieve your goal is to test out an idea and see how it goes. Maybe you'll get a 2% improvement. Then, you can try another one. Layer it on! The dessert of success looks more like baklava than a layered cake.

Around and around the learning loop

In a world where you are not sure what will work and you are looking for many of these small wins, the winners are the ones that go around the build/measure/learn loop the most times. If your test requires new data (and most of the good ones do), getting that data becomes the gating factor in how many times you can go around the loop.

Maybe you need the date and largest item of each customer's most recent purchase to follow up and keep them engaged. The question now is: how long is that going to take to obtain? It is not unusual to hear that it could be take months for that "data pull" to be a priority, assuming it ever happens at all. What can you do to make it clear this is important?

Among the challenges at play here is that this specific test might not even be that great of an idea. You will never get prioritized attention if you have convince someone that each of your crazy ideas will move the needle in a big way because it probably won't.

A great track record might show that 1 in 5 tests yields that 2% change, so we have to find a way to shift the conversation away from your specific idea and towards maximizing the number of shots you get to take.

Math and spreadsheets to the rescue

Fortunately, you have some of the greatest forces known the modern humans at your disposal: compound interest and spreadsheets. In this Google Sheet, I took some fairly simple things about the current situation and ran a simulation, projecting them out by cohort.

The punchline

In an attempt at brevity, I will skip talking about how fun it was to make that spreadsheet. Let's just get straight to the point.

In 2 years at this business, doing 4 tests a month instead of 1 moves the retention rate 146% instead of 110%, resulting in more than 20% additional revenue. Also, because compounding interest is a powerful force, Year 3 would be even better.