I’m a big fan of testing. If you’re not testing, you’re missing out on opportunities to improve the performance of your email program.
But even if you are testing, you may be missing out.
The key is the type of testing that you’re doing – is it soft or is it strategic?
Not sure what the difference is? Here’s an example to illustrate.
Background: The client is a membership organization; two key calls to action on its website and in its email messages are to “join” and “renew.”
Goal: To boost the number of website visitors that either a) join or b) renew.
Landing page: Calls to action with the combined “join or renew” message appear throughout the website and in email messages sent. When visitors click on them, they hit a landing page with standard navigation and banner advertising as well as both “join” and “renew” options and then choose the appropriate one to begin the process. On this landing page, both options are text links (as opposed to buttons).
Hypothesis: If the “join” and “renew” links on the landing page were buttons instead of text links more people would click on them.
Test: Four versions of the landing page:
- Control: “Join” and “renew” are both text links
- Test 1: “Join” is a button and “renew” remains a text link
- Test 2: “Join” remains a test link and “renew” is a button
- Test 3: Both “join” and “renew” are buttons
This is an example of what I would call a soft test. It has the potential to increase performance, which is good, but it’s driven by a valid but soft best practice concept (“buttons are more prominent than text links and will drive more people to click”), rather than by a strategic, quantitative analysis of visitor behavior. This test was actually being done on a client’s website when I started working with them.
So what would a strategic test for this situation look like?
A strategic test would utilize web analytics from the landing page to find out what people who don’t click on “join” or “renew” (the desired actions on this page) are doing instead. By finding out what’s taking people away from the task at hand (joining or renewing), we would be able to test closing these exits to improve performance.
Taking the strategic approach a step further, it would be good to look at the entire join/renew flow, from the landing page through the completion of the task. This analysis might find that the key issue that needs to be addressed doesn’t lie on the landing page – it may be that a significant number of people are clicking on “join” and “renew” on the landing page but that they’re dropping out of the multi-page process to achieve these goals before completion. Once again, identifying where people are exiting the process and closing these “leaks” should cause performance (read: conversion rate) to improve.
The key difference between a soft test and a strategic test is the quantitative data. The best marketing relies on quantitative analysis to inform test creative and other decisions. It’s not easy. There are folks who are great at analyzing the data but have no idea how to translate what they learn into creative changes to boost performance. There are other folks who are genius with making attractive creative, but who aren’t comfortable with the quantitative analysis required to do strategic testing.
Remember that old saying about working smarter, not harder? That’s what strategic testing allows you to do. While I can’t guarantee that a strategic test will always boost performance more than a soft test, this is usually the case. And it’s not that you can’t include elements of a soft test in your strategic testing; you can – just be sure you have a sound hypothesis based on quantitative analysis driving your efforts.
So, was your last online marketing test soft or strategic? If you answered soft, give a strategic testing approach a try and let me know how it goes!
Until next time,