First, let's define what we mean by A/B split testing. A/B split is the comparison of two components in a digital campaign to improve conversion rates. Half of your test base will receive one variation, the other half will receive another. The winning result will then be used for the full campaign base.
We A/B split test to improve conversions for a campaign, but which conversions do you want to improve? The secret to effectiveness is to choose one at a time. Testing multiple variables will not show you what increased your metrics.
Some components to test should include:
- Subject line or time of day, with the conversion being an email open;
- Copy, links and HTML vs. text, with a click-through being the conversion;
- The “from” label and email address, using email deliverability to measure success.
After identifying the component you'd like to test, here are some additional testing tips to improve effectiveness:
Majda Anwar is a revenue engineer for marketing agency Pedowitz Group (www.pedowitzgroup.com). She can be reached at email@example.com.
- Use a large enough test base. Your test pool should have at least 1,000 records to split evenly for conclusive results.
- Test one audience at a time. Send to a single industry or job level (C-suite, for example), or a single buy-cycle phase. Your campaigns should be segmented for best results at all times. Remember: right message to the right audience at the right time.
- Be conscious of your channel. A well-tested trade show subject line may not work for a newsletter send. Test across channels just as you would test across your audiences.
- Isolate the variable. Test one component at a time. Testing multiple components simultaneously does not allow you to understand what made your email successful. Was it your subject line, your copy, your graphics or your timing?
- Be consistent. Test for one conversion at each campaign. Every new campaign is an opportunity to test.