On Digital Marketing you must have a “mantra”: Every webpage could be improved. Always. This mantra should be used for every Marketing tool.
This is the premise that lead to always be testing. The A/B Testing is one of the most reliable evaluations, because it goes by comparison in real time. The information is being collecting immediately controlling if the objectives are accomplished or not.
That information collected is used to improve the money you earn.
There are websites with good performance. They have traffic and many of their visitors do what they are expected to do:
- To subscribe to the newsfeed
- To download a free trial of a software
- To know more about the services
- To ask a meeting using a Contact Form
- To buy any product
Is that enough? Of course not. The good percentages can turn into great ones and those into extraordinaire. Testing, analyzing and re-testing.
There is percentage of visitors that don`t meets the objective of your page. But they could do it! There are so many factors that influence the visitors.
How do you know what to do?
Every department on an enterprise will have an opinion about what it is important to improve:
- From Social Networks they will say you need more interaction with your visitors and clients
- The Sales team would suggests that you must put more products and do more promotions
- The Web Developers could say there is a need of redesign the architecture and the aesthetics of the site
- Yourself could have a an idea based on which area you prefer, or you have been told it is important or because you regard as much important some areas
However, the most reliable requirements source are your own clients. They know what they want, on the way and at the precise moment they need.
Too much bounce on your website? Maybe your promise of information or the solution is not so clear to understand. Maybe your Pay Per Click (PPC) Ads are not clear. Maybe…
The only way to know is to test. The changes made on A/B Testing will influence the behavior of your visitors.
The A/B Testing requires information
Before starting up an A/B Testing, it is necessary to do a detailed study of the page. If you evaluate without proper objective data, you could have a modification on your visitor behaviors. But you won’t know why at that will be something completely random.
On Digital Marketing you don’t go tossing dices. The tests and tools used are chosen according the collected statistics. Following the characteristics of:
- Your business
- Your clients
- Your products or services
And, above all, the information gathered must go hand to hand with the objectives of your website. So, to do tests without the proper data is to let go opportunities.
This initial information gives you some clues:
- Is the expected action too difficult?
- Or maybe it is so boring to accomplish?
- The content could weigh some types of people down?
- The design is not attractive?
- The buttons really looks like buttons?
The information can be collected by Web Analytics, detecting clicks, the time taken- to do an action, among other things. You can interview some clients too, but is more slow and costly.
On A/B Testing you create two versions con some differences, you are testing design or content against behaviors. Subjective aspects such how nice it could be are not tested. They don’t provide significant data.
The aesthetics are not an evaluation criteria easily identifiable. It is very varied and respond to a very specific characteristics of any person and society. Being so different makes it a not secure predictor.
To do a change just to make it look “nice” is a useless effort. Unless you have the necessary data that back that decision: colors, dimensions, diagrams to increase sales. In that case you can improve your site “only to look nice”.
This statement support that every change on A/B Testing must be relevant. Not only on clients matter but on what you are looking for. No more downloads? The forms are left half filled? The shopping carts are filled with products but there is no sales?
Above all, you should ask yourself: which are the objectives of your webpage how that change will help to accomplish them? When you could answer that, you can proceed with the test.
The page that was “finished”, that don`t required more work on it have become a moving thing. The A/B Testing is not only just one test. It can be done as many times as it is needed. Every time providing with new information.
At the same time you optimize your site, your competitor optimize theirs as well. Also, your clients could have new needs and different desires. That is why a webpage will never be “finished”.
- Every website can be improved
- The information comes from your visitors, they really know what they want!
- The test should be done based on data gathered previously
- “Beauty” isn’t a good testing element
- The relevance (how it is related to your objectives or to your clients) is the test criteria
- When you do an A/B Testing, you need a hypothesis, a control group, to calculate the necessary sample and how to validate it