When was the last time you did computer-driven improvements to your website? Almost all dream of multiple conversions, but often it stops with the dream. Here comes conversion optimisation Enter.
Conversion optimisation, also known as CRO, is a strategic approach and approach to increasing conversions and improving results.
What is conversion optimisation?
Conversion optimisation is a scientific process to make the most of people implementing a desired action of value. This is often done by affecting psychological factors that have a direct effect on consumer behaviour.
It is also a continuous process that consists of testing, error and learning. It always starts with analysis, which leads to some hypotheses, which are then tested and either confirmed or rejected. In summary, conversion optimisation is about increasing the number of sign-ups, sales, downloads or other important actions that are defined as conversions for you or your company.
This short video is very clear what conversion optimization is:
When is it appropriate to start with conversion optimisation?
A good rule is that you have to have significant amounts of data of what you want to test. If you have less than 1000 conversions per month, you are in what is referred to as the ‘risk phase’. Here you will strive to have enough data to get results you can trust, and if you have too low volume it is often better to only implement changes and monitor afterwards.
As you approach over 10,000 conversions per month, it is more appropriate to look at more automated solutions for optimization. It's important to remember that the more conversions you have, the more accurately you get from tests.
Advertising in digital channels is getting more expensive and more expensive. It is then important not only to drive traffic to the website but also to make the most of traffic that is already there. Many people are too focused on driving traffic to the side versus optimising the side.
What are the elements that I can start testing with?
Everything can be tested, so the most important thing is to choose where to start. What to start with depends on your goals and your current situation. The analysis done in advance of an A/B test will always look for potential problem areas and start with the largest ones.
A/B-testing (also known as split testing) is a randomised experiment with two variants, A and B. A/B testing is a method of evaluating the effect of one or more variables. This is done by showing different versions of a website or app to different users. The version that converts best after reaching the right size of the selection and the test for a long enough time will be the version that is implemented.
Here are some universal suggestions about which elements you can start testing already today:
- Call-to-Action (to increase CTR, not purchase)
- Layout
- Text
- Graphics
A/B test your Call to Action buttons
You can test this by
- Switching the words you use
- Toggle button position
- Change layout as size, design or colour of a button
Note that A/B testing of buttons usually only results in a decision when the original buttons are a friction point in the customer journey. If the buttons have low contrast to the background and are hard to see, then it makes sense to just fix the buttons and not spend time testing A/B. One way to see if a button is intuitive is that a user can understand what happens on the next page without clicking on the button. This is usually not a high value thing to test.
Read here about how you might be encouraged puts your CTA buttons
See this article for more inspiration
A/B Test Your Headings
The idea of a headline is that it should capture the customer's attention. If you use digital advertising, the titles should also match or reinforce the advertising message.
To test headings, you can:
- Test the words and order of them
- Test General Message
- Test different sizes, colours and fonts
- Test personalisation and various messages against different target groups.
A/B test your text
The text you use is important and is a great deal about how you choose to communicate your message.
For example, to test text A/B:
- Words - What and How
- Length and Formatting - Text Structure
Frequently Asked Questions About Conversion Optimization
How do we figure out the conversion rate?
Here are three formulas for conversion rates you can use: Conversion Frequency = number of conversions/number of sessions * 100 Conversion Frequency = number of conversions/number of unique visitors * 100 Conversion Frequency = number of conversions/number of potential customers * 100 All three above formulae are valid formulae. The best way to calculate the conversion rate depends on what you define as conversion and how you measure traffic.
When is it advisable to avoid A/B Testing?
In intense sales periods that go beyond shorter times, there may be a high risk of A/B testing. This is for the simple reason that you can lose a lot of turnover on sending 50% to a suboptimal version in these periods. In cases such as Black Friday, Christmas trade and other short selling periods, it is often lurking to not test A/B, possibly testing through a bandit algorithm if you have enough volume to do so. Read more about this Here.
How does CRO differ from UX?
There are many different views on this, because areas with optimisation of conversion rates and user experience overlap. Conversion optimisation is a process that is about improving your existing website through ongoing A/B testing and optimisation (more focus on business goals). UX is the process of planning and examining the best way to make a significant one-off change (more focus on customer objectives).
Do I need developers to do A/B testing?
There are two methods of implementing changes on a website. Either HTML/ CSS/ JavaScript can be used to make the changes, or you can use a visual drag and drop builder (WYSIWYG editor). It is recommended to use developers to design the test as visual solutions can more easily lead to misdesign on certain devices. Among tools available are: Optimize, GetSiteControl, Sitegaines and Google Optimize which allows you to make changes to your website without development. All you need is JavaScript code on each page of your website, which can be implemented through Google Tag Manager.
What should I think about before I start with the CRO?
Before you start optimising, you must have the following things in your head: – Feel and be confident about your USPs – Know and understand your target group – Define clear KPIs – Have a solution for mobile version – Know who your competitors are – Check the Web Analysis Tool – The hypothesis may be correct but the design of the test may be incorrect – Small changes are more difficult to see the effect of – Try major drastic changes, but be prepared to be more difficult to analyse and isolate effects with several changes Ask yourself the following questions too? – Do I have enough data to start and test, or am I in the risk phase? Do I have a good enough data base to design good hypothesis? - Are the hypotheses good and testable? – Is there at least one clearly defined measure per hypothesis? – What do you learn if the hypothesis is wrong?
Can I use A/B-test on websites if there is little traffic volume?
There is usually no point because it is time-consuming to analyse both before and after the test. Here it is better to only implement and monitor afterwards. At least 1000 conversions are critical.2 If you are in border countries with respect to conversion volumes, you can test, but it is important to remember that radical changes will be easier to measure. Changing the color of a button is likely to never be able to calculate any difference.
How much does CRO cost?
The costs associated with the CRO are mainly the time it takes to analyse, design hypotheses and tests, and then analyse the results. It is a time-consuming process for most enterprises, and is often the reason why the method is downsetting.
What are the best tools for A/B testing?
Good tools for A/B testing include Optimize and VWO. Google Optimize can also be used, but has some limitations.