5 Tips to Help You Get Started with Google Optimize

Written by Conor Ross

Conor specialises in data visualisation, designing and creating bespoke data dashboards and automated reports with Google Data Studio. With a background in Psychology and Research Methods, he also works on projects focused on improving customer experience and website usability through A/B Testing and Personalisation in Google Optimize.
July 19, 2018

One of the most important aspects of any website is the user experience. While it’s great to understand how consumers reach your website, it’s equally as important to understand how they behave when they get there. This is where optimisation testing with Google Optimize comes in!

Optimisation testing is a reliable way of discovering what works well on your website, and where your website can be improved. Identifying the best user experiences is beneficial for both you and your customers.

Testing typically involves changing elements of your webpage, presenting the different variations to an equal proportion of users, and seeing which variation performs best. At GlowMetrics, we like to use Google Optimize for running optimisation tests, as it integrates seamlessly with Google Analytics and Google AdWords, but the principles of testing remain the same regardless of your preferred software.

User Experience, or UX, Testing can be a daunting prospect, but with Google Optimize it doesn’t need to be. There’s no need to worry about statistical tests or algorithms, as the complicated mathematics are done for you, so you can focus on the most important observations and results. While the usability of Google Optimize itself is, well… optimised, we have come up with some key tips to keep in mind when running UX Optimisation tests to help you along the way.

  1. Keep It Simple

For the best results, keep your tests simple in design. Changing a single element on a webpage leads to a high level of certainty that any observed effects are a result of that variation. If you design a complex experiment with multiple simultaneous changes, it may be difficult to identify which changes are driving effects. Small tests lead to big insights.

  1. Maintain Control

It’s important to maintain control over your tests. For example, if the variation you have made doesn’t affect mobile users, make sure they are not included in the experiment sample. It is also advisable to avoid testing over unusual periods where possible, as non-typical website use reduces the validity of your results. Finally, refrain from making any significant alterations to the webpage until the conclusion of the test. Be mindful of any external factors that might impact your test, and therefore influence your results.

  1. Let It Run

Seeing your test launch and observing the real-time data can be exciting, but be careful about forming interpretations too soon. It is recommended that an optimisation test runs for no less than 14 days, although this can vary greatly from test to test. For example, testing a webpage that appears later in the customer journey is likely to consist of considerably fewer sessions a day in comparison to testing your homepage. Decide upon an experiment duration before you begin, and try to stick to it.

  1. Focus Your Goals

Optimisation tests should not only seek to identify the best user experiences, but should also improve and optimise your own goal conversion rates. Design your tests and objectives based on measurable goals that coincide with the greater objectives of your business.

  1. Embrace the Insights

Unfortunately, it is all but guaranteed that experiments will not always produce the results you expected or hoped for. But don’t be afraid of “negative” results! If done correctly, your test has still produced valid data that can help with analytical insights, and the knowledge you gain through the testing process is still extremely useful. Understanding what doesn’t benefit user experience is every bit as important as what does.


Get the Most from Google Optimize
Website Optimisation is an ongoing process. Each experiment will be influenced by your previous experiences and expertise. Remember these five tips when running UX tests, and you will soon be designing insightful and reliable experiments. Stay calm, and keep testing! If you’d like to know more information about UX Optimisation testing, check out our dedicated User Experience Testing page, or Contact Us.


Enjoyed reading this article? Find more like it below…