Speed Up Your Decision Making With Split Testing

March 5, 2019

Speed is paramount in business. People are always looking for the newer, better thing. Smartphone manufacturers who release a new phone every 2 years won’t make nearly as much as companies like Apple or Samsung who release a new phone every year.

It’s all about iteration, constant improvement, and doing it quickly. If you don’t move quickly, your competition inevitably will. And they’ll win over the customers you could’ve had.

Most businesses intuitively understand this, but are unable to put it into practice. The reasons for this are many, but they mainly boil down to fear of failure. Fear that their decision will backfire on them in some way.

The obvious solution here is simply to try anyways, accept your wins and losses, and move forward, but I’d need a motivational speaker to convince certain people of that. For others, the politics in their business are simply too tense, and they cannot afford to take a loss.

So here’s another solution: Reframe your decision as an experiment instead. Here’s where Split Testing comes in…

Split Testing is the act of splitting your web traffic to multiple versions of your site, with each version testing a different strategy. If you were trying to improve newsletter signups, version 1 might test placing a newsletter signup above the navbar, whereas version 2 might place it in the sidebar. You can test different things like copy, design, layout, anything you want—the key is to make sure you can measure the results. Visitors to your website will be randomly selected to see one of the versions, without ever knowing that they’re part of this experiment. Over time, you track particular metrics to see which version of your strategy garners more results, and stick with the winner.

Split Tests are easy to set up and perform, and you can test as many versions as you want simultaneously.

With Split Testing, you no longer need to debate endlessly over which strategy to choose. If the decision comes down to a stalemate between two options, simply test both and see which one is the correct answer. Or, if your business typically runs on a certain strategy, and somebody wants to try a new one, perform a split test with a control group and an experimental group.

Being able to test multiple options removes the fear of choosing one and always wondering if the grass would’ve been greener with the other option. It also prevents getting stuck in “analysis paralysis”, where the tough choice between multiple options causes overwhelm, and leads to slow decision-making or no decision being made at all.

There’s no need to waste with excessively hypothesizing when you can run an experiment that directly proves or disproves your hypothesis.