A person’s perception is not related to colours  but colours implicitly influence perception through placebo and nosebo effects. Although every person has his/her own feelings related to colours, it is seen that, in general certain colours induce certain moods in people. Amalgamating this psychology of colours to design our A/B testing scenarios, we would have a more pronounced effect on our results when it comes to the design of user interfaces or webpages. In this article we present a four-point methodology to design better A/B testing scenarios, leveraging colour psychology.
A/B testing, also known as hypothesis testing, is the process of running an experiment simultaneously on two different versions of the same product against a user base to check which version performs better or produces a better conversion (in terms of numbers). The differences between the two versions are usually a single parameter but it is also possible to perform the experiment by changing a range of parameters. This is termed a multivariate testing. 
This is a proven method to figure out which strategy, hypothesis or assumption is best to improve our baselines. We can statistically prove which version is better and use the better version for our future promotions or business progress. A basic outline of the process can be seen in the figure below.
Our lives are closely knit with the colours around us. We create our impressions and associations of culture, emotions and personal feelings from our encounters with the colours that we see on a day-to-day basis.  Most colours have an underlying emotional meaning and association with a person’s common moods, often provoking reactions.
For example, warm colours such as maroon, red etc. represent comfort and strength, pink is generally associated with femininity, orange excites energy and creativity, yellow induces a sense of authority, warmth and intelligence, green in general implies greenery and environmental friendly activities, blue is associated with trust and security, purple represents luxury and nobility, brown produces a feeling of comfort, black and white are associated with simplicity and sophistication etc. 
The basis of any A/B system-testing scenario is to change one parameter in a version of website or software while keeping the rest of the items exactly identical and test both the versions against a user base.  The aim is to identify the changes that can be made to increase the efficacy of our desired result. Primarily, when it comes to a user interface of a software or website, visual and auditory influences determine the user actions and subconsciously direct the user to perform certain desired actions.
Of the two, visual effects tend to have a greater effect on us, as vision is the most dominant sense organ in humans. So, empathising the psychology of colours in our software user interface or websites and designing our A/B testing scenario with this in mind can hypothetically produce much better results.
This can be done with the 4 point methodology below:
Identification of the business needs
The first step when creating any product (in our context, software) depends on the business need. It is essential that any the software tackles a particular business problem. Once the needs are identified and we are clear on the goals we wish to achieve, we can proceed along the process of designing and development of the software. Although this forms the primary requirement and functionality, the success and achievement of the end goal depends on tackling the non-functional attributes as well. There are various non-functional requirements, but in this article we plan to tackle the usability, ease of access and the user attraction and retention factor through colour manipulation.
To integrate the functionality of the product along with the end goal of directing the user to perform the desired actions as intended by the business, we need to concentrate on the visual appearance, the user interface or look and feel of the software.
Proactive identification of points of interest
In a software user interface, there will certain key areas that are to be highlighted to attract the user and have them perform certain actions. This is primarily goal oriented. Say for example we have a login page; the login button is usually placed right below the text boxes where one would have to enter the user name and password. This button would generally be highlighted so that the user’s attention is directed towards that button. This subconsciously prompts the user to click the button thus taking the user along a flow of actions.
Similarly each page will/should have such key points, so that the user takes a directional path while using the software. These key points of interest depend on the business need and as a developer/owner of the software we need to have them highlighted, so as to extract maximum usage from the users/customers.
Colour psychology driven scenario design
Research has shown that people associate colour with feelings. Exploiting this information we can trigger and stimulate selective moods with each and every colour. Conversely the wrong colours may induce inappropriate feelings and negate the underlying business need. For example if we use a colour such as red for an environmental and eco-friendly organisation’s web page, it would potentially induce moods of destruction and harm, rendering the organisation untrustworthy and unprofessional.
Keeping this as a baseline, we can design our “a/b test scenario”, wherein we highlight the points of interest which were identified earlier and thereby we optimise our user interface or websites to obtain maximise usage, customer retention etc.
Data driven analysis triggering reactive modification and optimisation
Now that we have identified our points of interest and have taken an educated guess of which colour scheme we are to follow, we can proceed to validate our hypothesis using the “A/B Test Methodology”. I.e. we will test our two versions of the user interface, the original and the modified colour scheme version simultaneously against a user base and based on the results obtained, we will decide how we are to proceed.
If the newer version produces better results, we will then make the latest version of software/user interface our final release version and in case the converse happens and the new version produces comparatively poorer results, we can revert back to the baseline version of the user interface.
The impact of colour in action
Studies on impact of optimising colour usage in marketing campaigns have shown that in some cases there has been a staggering 80% increase on brand recognition, a 40% improvement in readership and 73% increase in comprehension. It has been seen that 85% of people decide to buy a product due to its colour. 
Paras Chopra, the founder of Visual Web Optimiser, has observed that there was a 60% increase in downloads count when the colour of a button on a web page was changed . Although the text of the button was also changed along with the colour, the change in the colour from blue to red had a valuable role to play in the end result.
Colour plays vital role in influencing once mood or feelings implicitly through emotional and cultural associations. Understanding the psychology of colours when crafting our A/B test scenarios will help us in their design, which in turn can make a big difference in the outcome and effectiveness of our test’s intent.
- Cherry, K. (n/a). Color Psychology. Retrieved March 19, 2014, from About.com: http://psychology.about.com/od/sensationandperception/a/colorpsych.htm
- Macdonald, M. (2014, February 20). Shopify. Retrieved March 20, 2014, from The Beginner’s Guide To Simple A/B Testing: http://www.shopify.com/blog/12385217-the-beginners-guide-to-simple-a-b-testing#axzz2wUcjZpgg
- Team Wix. (2013, June 25). Wix. Retrieved March 20, 2014, from Color Web Design : http://www.wix.com/blog/2013/06/color-web-design/
- Bhangu, N. (2012, July 24). vandelaydesign. Retrieved March 19, 2014, from The Psychology of Color in Web Design: http://vandelaydesign.com/blog/design/the-psychology-of-color-in-web-design/
- Chopra, P. (2010, June 24). Smashingmagazine. Retrieved March 19, 2014, from The ultimate guide to a-b testing: http://www.smashingmagazine.com/2010/06/24/the-ultimate-guide-to-a-b-testing/
- Dyer, P. (2013, April 22). Social Media Today. Retrieved March 2014, 2014, from The Role Of Color in Marketing: http://socialmediatoday.com/pamdyer/1401166/role-color-marketing-infographics
- Chopra, P. (2010, November 24). SmashingMagazine. Retrieved March 21, 2014, from Multivariate Testing in Action: Five Simple Steps to Increase Conversion Rates: http://www.smashingmagazine.com/2010/11/24/multivariate-testing-in-action-five-simple-steps-to-increase-conversion-rates/
About the authors
Vikas Chakravarthy Srinivas has a master’s degree in Computational Software Techniques in Engineering from Cranfield University (UK). He also has around 3 years of performance testing experience working with reputed clients like Fidelity Investments and Lloyds Bank. Apart from constantly trying to innovate and trynew things, he enjoys food, travel, singing, dancing and cooking.
With over 14 years of experience in performance testing and engineering Yoganand Lakshmanan is currently employed with HCL technologies as Sr. Technical Architect in Performance engineering, with tremendous experience in fine-tuning the performance of applications built on .NET and Open Source technologies spanning across various domains. Apart from technology, Yoganand has constructed and presented solutions to customers thereby helping to win clients in the performance testing and engineering space.