A/B testing is one of those technical terms that automatically elevates you to the category of “ninja” if you are any room talking about how you were using it. It’s one of the most powerful tools to help optimize your website conversion experience.
WHAT IS A/B TESTING?
On a very basic level, A/B testing iis taking one element on the webpage and changing it so that you have two versions of the same page. And then present the one of the two different pages to website visitors as they come to your site. After you have enough visitors try each version of the page, you can compare the results – so that you could measure consumer behavior and determine which one helps you reach your business objectives quicker.
TESTS THAT YOU’LL LOVE
You would be amazed at how changing colors on a button, or even one word in a headline can impact if someone immediately clicks back to abandon your website or stays longer and actually does what you want them to do on the page.
In my experience, A/B Testing has allowed me to increase click through rates for product sales or donations or newsletter sign-ups by up to two digit percentage gains.
Given that generically 90% of the people who visit your site simply don’t do anything and leave and never come back, being able to optimize and increase the engagement on any given Paige is helping to increase the value of all the efforts you have been putting in to get people to your website. Traffic can be expensive. This is a strategic way to lower the cost of each incremental visitor to your site.
Just think of 1000 website visitors to your page. .5% click through on your call to action. That’s 5 people converting on your page every day.
If you can optimize your page so that the .5% click through rate goes up to 10.5% click through rate, the impact is huge! That is 105 conversion actions on the page now.
That is with NO change in website traffic.
ONLY FOR SUPER NERDS IN THE PAST
In the past, it was ninja-evel work becaus everything had to have been done manually. You needed to design to separate HTML pages. Have them installed on your site. Run a script to distribute the traffic randomly between the two pages as visitors come in. Record data analytics on whatever behavior you want. Have the reporting pull together in a way that makes sense so that you can make decisions. Review it, make the decision, so that you could make the change to the Resort webpage, re-upload it, etc.
Whew. That is a lot of work. And it is pretty much impossible if you were not technical.
TOOLS YOU CAN USE
Then tools came on the market to make the process easier, and we have things such as optimizely.com, kissmetrics.com, and unbounce.com.
These paid solutions are great website optimization tools to look into if you were serious about increasing the effectiveness of your website.
Google Analytics has google experiments, but it is not that simple to implement.
GOOGLE TO THE RESCUE
But this month, Google finally released its A/B testing optimization tool, which used to be reserved for enterprise customers, to the general public and it is free!
Google Optimize is one of the top five things that I think all people who are in charge of websites need to be looking into in the next 30 days if you have not considered A/B Testing at all.
The way it works is that you install Google analytics. If you don’t have a Google analytics installed and ready, shame on you. Stop reading this post, and go install it right now.
Google Optimize gives you one line of code to insert into the Google analytics header code snippet.
Next you’ll download the chrome extension for Google optimize. This will allow you to view your webpage and click and drag change different elements on the page that you were going to test.
Once you create the second “B” version of the page, you’ll set up what I’ll come you were trying to test. You can test click through rates, how long they stay on a page, etc. etc.
MAKE DECISIONS BASED ON DATA
Then you simply wait for enough traffic to hit the page that you have set up the experiment for. After several hundred visitors, you can look at the results and Google Up the mice will even give you a percentage probability based o statistical extrapolation of a confidence for any given page version to be the winner.
Once you have a winner, you can kill the losing page and set up a brand-new experiment to try and beat your new optimized page even further. This can be a never ending game to tweak and improve any given page on your website.
Of course, there comes a point where your improvements will plateau out, and you will see your metrics stabilize. That’s when you can start to look at other pages — or consider how to radically change layouts and presentation of content.
Please promise me that you’ll get started today: go sign up at http://optimize.google.com
And let me know about your first experiment!
