Conversion: All About That Iteration
Aside from the very specific problems laid out in my other case studies, the general problem of a growth marketer is growth. As growth hackers, experimentation and iteration are our meat and potatoes. Often multiple tests are running at once so we can look at interaction effects in addition to main effects. Here's a sample of some of the small-scale tweaks and hacks that I've implemented.
1. Communicating value proposition. When you work with a long funnel, there's going to be attrition at each step. When your product is a long funnel that requires the user to do a not insignificant amount of work, you're going to see even more attrition. This makes communicating the value proposition of your product that much more important. I ran several tests to strategically communicate value proposition in the app, increasing conversion by 1-3% to the next step. The general conclusion was that the best value prop messaging is visual, brief in text, near the CTA, not buried within a link or pop-up, and not too large as to disrupt the design.
2. Clarifying the UX. We’d heard from users that there was some confusion as to what happened when one pressed “Get started” on one of the pages in our funnel. The hypothesis was that "get started" might be too daunting of a CTA, or create a false impression that there was a tedious process ahead. By changing the button text to “Next,” I was able to improve conversion to the next step by 5% (z = -6.05, p <<< 0.001).
3. Getting emails opened. Sometimes you get to reach for the low-hanging fruit. One of our triggered emails was seeing good conversion, but I suspected that it could be much better. By simply changing the “From” description from the company name to a person's name, I was able to increase our open rate by 39% in the first week. The novelty of the change eventually wore off, but the open rate of the test version stayed consistently 25% better than the original version thereafter.
4. Getting clicks from emails. A similar low-hanging fruit. Another of our triggered emails had a simple call to action: click here and [x] will happen. I tested an alternate version with a button class hyperlink and immediately saw a 10% increase in conversion.