Design concepts don’t always turn out in the way you’d like.

The layout looks great. The color options seem excellent. The CTA is both clear and clever.

But…

The system isn’t working. It’s all wrong. A little bit. You’re not sure But something has to be done.

Despite the best intentions of everyone, and despite all the hours spent on studies and analysis it’s not always working out the way you’d like them to.

This is the reason why constant testing is essential. This is not a simple hail and prays plan.

The best part is that your testing doesn’t have to be complicated and time-consuming.

Here’s how you can set up a split test within Google Analytics in just a couple of minutes.

What are these Google Analytic Content Experiments?

Let’s say that your online shop sells Pug greeting cards. (That’s a thing to be aware of.)

It is obvious they are supposed to be able to sell their own.

Let’s suspend our disbelief for a second and speculate that the sales aren’t as high due to the difficulty in getting customers to these product pages initially.

Your homepage isn’t an endpoint It’s a point of departure.

The peeps are here, they browse around and then go to a different site.

A lot of times, that’s your product or service pages. Most often, it’s the About Page.

Whatever the case, the aim is to put them into the channel or pathway as swiftly as is possible, (a) helping them to find what they are looking for, and (b) making them closer to initiating an event that triggers conversion.

The magic happens when you create the landing page, where the two factors such as a person’s curiosity and your business’s goal merge and create the form of a stunning orchestra.

Let’s try some homepage designs to determine which one does best in getting new customers to your top-selling products.

One includes an embedded video and the other doesn’t. One video is sweet and short but the other is lengthy and specific. One of them has a GIF but the other one doesn’t.

The new traffic that is incoming is split into these different variations of the page which allows you to observe and evaluate the number of people who have completed the desired action until you are able to confidently declare the winner.

(It’s likely to turn out to be one that features the video.)

A simple split test such as this is landing page optimization 101, in which you determine the specific variables on your landing page which produce the best outcomes for your visitors and then multiply them throughout your site.

Google Analytics comes with a basic feature called content experiments. It lets you compare different variations on a page, divide the traffic according to these variations, and receive email updates about the trends in your results and whether you’ll achieve your goal or not.

However… it’s not an easy A/B test. Let’s look at why, and why it’s actually a positive thing.

How Content Experiments can be More Effective than A/B Tests Using Traditional Methods

The typical A/B test picks the particular element of the page, for example, that headline. The test then alters only one tiny element in the new variations of the page.

The web is packed with stories were changing the color of the button led to an increase of 37,596% in CTR* due to people preferring buttons that are green instead of blue ones.

(*That’s an invented number.)

There are some issues with the classic A/B test, however.

The first thing to note is that small changes are often regressed back to the average. Therefore, while you may notice some small changes as you begin to run an experiment, even small variations typically only translate to small changes.

A significant Update on Content Experiments

Google Analytics and experiment resources of the Management API. Starting on August 7, 2019, the users won’t be in a position to create or begin the creation or start of Content Experiments (CX). Current experiments are allowed to finish. The completed experiments will be visible in Google Analytics experiment reports.

If you’re working on any projects that are in the process, you should consider moving your experiments into Google Optimize. It’s completely free, has an integrated connection to Google Analytics, and has all the features listed below: