Home Business What is experimentation and why are so many doing it wrong?

What is experimentation and why are so many doing it wrong?

by gbaf mag

 By Jil Maassen By, Lead Strategy Consultant at Optimizely

This year online offerings have taken on a new focus for businesses, and the financial services sector is no exception. With in-store interactions limited, customers have been flocking to online facilities instead. In fact, 71 per cent of consumers globally are now using digital banking channels on a weekly basis. This has meant that optimising customer experience online is now much more important to a business’ success. Financial institutes have a real need to maximise the ROI of their digital offerings, by making sure that their web pages and apps run as efficiently as possible, so they can maximise conversions and satisfactorily complete customer journeys.

To do this, they need to embrace experimentation. Too many businesses use gut feeling, or individual preferences, to decide the structure of web pages. This approach is wildly ineffective. Experimentation allows a much greater understanding of how changes to an online system will affect user behaviour. This helps take the risk out of change, allowing for positive improvements in design, that have been tested first. This ensures that customers are moving through the website or app in the way the business wants them to – and ultimately down the sales funnel.

Testing out the options for experimentation

A very common method of experimentation is known as A/B testing. Also known as split testing, this method of website optimisation looks at the conversion rates of two versions of a page — version A and version B — comparing them to one another using live traffic. Site visitors are bucketed into one version or the other. By tracking the way visitors interact with the page they are shown — the videos they watch, the buttons they click, or whether or not they sign up for a newsletter — you can determine which version of the page is most effective. This is a popular method for experimentation as it allows changes to be based on actual user feedback.

As an example, take the addition of a click-to-buy button on a web page. Here you would test half the group with a version of the web page that includes the addition of the click-to-buy button and the other half would see the page without (the original version). You can then determine which variation gets the better engagement or response rate.

This testing is very common; however in many cases it is limited when it comes to meeting the ever-changing needs of those within the financial sector. This is because customers now expect a seamless online banking experience, with the ability to manage all aspects of their accounts online. Sticking to only A/B testing limits a business’s ability to provide a competitive offering of this experience and leaves massive opportunities and potential revenue on the table. Instead, a blend of both A/B testing and multi-variation testing, where you experiment with multiple combinations of web page elements, simultaneously, can help drive success.

Dispelling the myths behind A/B testing

Before we look at the benefits of this blended approach, it’s important to understand why so many businesses have tied their fate to A/B testing alone. Often, the answer is simply lethargy. Teams are comfortable with A/B testing and don’t want to think beyond the obvious.

One of the common excuses given is lack of resources to implement more in-depth experimentation, however in many cases this is an inaccurate perception, as for most simple test cases, testing more than one variation isn’t much more effort than just testing the one. Even more commonly however it is because A/B testing is being used in the wrong way. It is being used to validate the known. For example, someone in the business may have said: “let’s try changing our landing page colour to blue because our competitor uses blue”. The test is then run using the original colour vs the change to blue. The users in the test don’t like it, so they stick with the old version. However, this approach means they have missed a huge opportunity to try other colours or even other changes, like a position or text change. If they would have tried variations like that, there might have been massive hidden uplifts.

If you are running multi-variation testing, you are trying out more things, thinking outside of the box and taking more risks by going with bigger changes. At the same time, more variations means a higher chance of getting it right. Just like in horse racing, you don’t just bet on one horse, you bet on several horses to increase your chances of winning.

As the usage of online services has grown, along with a greater number of demographics using them, the case for a more powerful testing process has become stronger. As online services are now a core part of a business offering in the financial industry, multi-variation testing unlocks more opportunity to deliver the best possible product.

Finding the right experimentation fit for financial services

As online services have become more important, they have also become more complex. From customer service through to sales, a business’s digital offering will encapsulate the needs of many different functions. This means that there will be, on any single web page, multiple different elements, each pointing the customer in a different direction. In this case, simple A/B testing is limited in what it can tell you. While you can find out which variations of an element are more effective, you can’t tell how different variations, of different elements, might affect the customer journey when changed together.

This is where multi-variation testing can offer more opportunities for digital experience optimisation. For example, a business can set up multiple different versions of a web page, each with a different blend of web elements. This not only allows visibility on which version delivers the best conversion rate but also how the individual elements work together to do so. For example, if a retail bank is looking to implement a customer service chatbot on their landing page, it would be very useful to understand how changes in other web page elements, such as a sales popups, might affect the customer’s desire to interact with it. Using multi-variation testing, a business can discover the best possible variation of webpage assets, whether that be headers, colours or pop-up buttons, that encourages visitors to use the chatbot. Then they can adapt the page accordingly to make sure that they are driving the desired customer outcome.

A/B testing is great for web pages with smaller traffic numbers, or for introducing the concept of optimisation to an organisation, however the real businesses value lies in experiments with four, five or even six variables, along with the testing of several different versions of different elements simultaneously, through multivariate testing. This type of experimentation is extremely important for businesses in the financial sector, as the ever-increasing scale of competition online means that identifying the right combination of elements which will have the most impact is hugely important in setting a business apart from its rivals.

The best of both

It’s important to remember that both A/B testing and multi-variation testing have strengths and weaknesses, and neither one should be conducted in a vacuum. The issue is that many businesses often focus only on A/B testing when in many cases multi-variation testing is the better option. Together they are two powerful optimisation methods that complement one another very well. A business that runs successful experiments of this nature can optimise both large-scale change, while also driving continuous improvement in their online offering. Moving into the new year, it will be those financial services businesses that have the ability to build engaging web pages and apps which stand above the competition and maximise the positive effects of customer interaction.

You may also like

This website uses cookies to improve your experience. We'll assume you're ok with this, but you can opt-out if you wish. Accept Read More