At 2Checkout, we’re constantly thinking about conversion rate optimization (CRO). We not only come up with new ways to improve conversion on our own site, we also build new ways for our customers to test and optimize their own conversion rates. Recently, we hosted an in-depth webinar on the topic featuring our own Shannon MacLeod, Director of Optimization. She shared a number of useful tips for planning tests and optimizing your conversion rate that we’ll summarize in this blog post. For more tips and case studies on CRO, you’ll want to check out the full webinar!
What Can You Optimize?
Here’s the good news: you can optimize almost anything! Whether we’re talking about your landing pages, home pages, product pages and content that’s present across your site, to your shopping carts, purchase flow elements, form fields, and cross-sell or upsell offers, it doesn’t matter—you can optimize it all. But be sure to keep a level head—you shouldn’t try to optimize everything at once. Instead, go through the optimization cycle to ensure successful testing and optimization.
What’s the Optimization Cycle?
The optimization cycle starts with analyze, then proceeds through strategize, test and finally optimize. When you analyze, you’ll identify a success metric—conversion? downloads? revenue per visitor?—and do a little research on what might improve it. Once you have a hypothesis about some factors that could affect conversion, you can strategize effective ways to test those factors and create a focused test plan. After running a well-planned test, you’ll take a look at the results and, finally, optimize by implementing the winner.
An Important Tip for Testing: Split the Difference
A key element of successful CRO is doing split testing instead of before/after testing. That is, split your website traffic between a control page and a page with the update you want to test, rather than making a change and comparing traffic before and after. Because your website traffic and conversion rates are constantly varying due to a variety of unpredictable factors, before/after test results may be skewed due to inconsistencies in traffic volume, traffic quality and other reasons. Splitting your test will ensure that differences are related to what you’re testing, not traffic quality.
A Few More Testing Best Practices
Beyond choosing to split test instead of before/after test, what can you do to ensure a successful CRO test? Here are five key elements to effective testing:
- Test one hypothesis
- For at least seven days
- Track at least 100 conversions per variation
- Reach 95% statistical significance
- Remember that 2 out of 10 tests are neutral or inconclusive
By keeping your testing focused, running tests for long enough (both in terms of time and conversions) and aiming for statistical significance, you can run successful tests. And remember, even if you test well, your results may not tell you much. That’s okay! Start the optimization cycle over, with more analysis, and test again.
Purchase Flow Items to Test
When it comes to conversion rate optimization, there are several key elements that you can focus on when testing. These include cart flow, form fields, risk reduction elements (like money-back guarantees) and cross-sell/upsell opportunities. It’s also smart to test and optimize your free trial signup if you offer one. Again, don’t go too wild and test a ton of items simultaneously: it’s best to run focused tests of one element at a time so you can tell what’s really influencing your results.
Now that you’re beginning to understand the optimization cycle and how to conduct successful, focused split tests, it’s time for you to go forth and test! If you still need additional information or inspiration, check out the full Boost Online Revenue through Conversion Optimization webinar for more details and some insightful case studies showing double-digit increases in conversion. Happy testing!