Top 5 tips for effective landing page A/B testing

A/B testing is one of the best ways to raise your landing page conversion rates and overall online marketing ROI.  A/B testing or split testing is a testing method where a control landing page is compared to a variety of test pages.  Testing landing page experiences against one another with A/B testing offers the opportunity to create and test wildly different things – think in terms of testing apples and oranges.  

Here are five tips to help you effectively plan and run landing page A/B tests:

1) Know your testing type

Just like how investors have varying degrees of risk tolerance, marketers and organizations tend to have predispositions or cultural norms that affect testing.  Do you just want results, or do you want to know exactly where those results are coming from?  The upside of testing two widely different landing pages is that you can find big winners, but you can also find big losers.  If the two pages are completely different however, you won’t know for sure why a page won or lost.   There isn’t a right or wrong answer, but deciding whether you want that knowledge or just the results will affect your testing plans.

2) Create a test plan  

The best results and insights come from tests that are carefully planned and hypothesized.  This doesn’t mean that you should spend forever planning out a test; it just means you should fully understand what you’re testing and, even more importantly, why you’re testing it.  Getting a plan in place that defines objectives, traffic, baseline results and your next steps will help keep your testing program on target even when results get crazy.

3) Statistical confidence

Run tests until they reach statistical confidence.  Even though you may have what looks like a clear winner right out of the gate, wait until a test reaches statistical confidence before declaring any victors. Landing page testing is typically tested to somewhere between 80-99% confidence, depending upon a variety of factors. With LiveBall, ion’s landing page management platform, you can choose your level of confidence and let LiveBall automatically declare a winner, or you can evaluate real-time testing gauges to declare winners yourself.    

4) Be bold

Beware of the never-ending A/B test! Alternatives that are too similar may not deliver a statistically significant result within a reasonable period of time, sending your testing into a state of paralysis. When you find yourself in a test wave that appears to be a statistical draw, call it as such and move on.  Every time you form new test idea, take a step back to look at the alternatives and ask yourself whether you are testing something of significance.

5) Innovation then iteration

A/B testing is a powerful method for increasing conversion rates, but you can get even better results if you plan a test with two phases: innovation then iteration. Start by testing bold, innovative landing pages with A/B testing and then refine the winning page with multivariate testing (MVT).  Think of A/B testing as a way to find diamonds, and MVT as a way to polish them.

Using these five tips, you’ll be ready to plan effective, valuable landing page tests. Happy testing!