What is conversion rate optimisation?

If you are in e-commerce you will be familiar with the conversion funnel, which is the journey that a customer will take through your website before converting to a sale. In general the fewer steps you include in your conversion funnel, the higher the final conversion rate.

For example, an ad on a search engine may have a 10% conversion rate. An email newsletter typically converts at 1% while a followup after a sale may convert at 5%. These numbers tend to increase dramatically if you pick the right time and day, or target a customer just after interacting with them.

A badly designed customer journey can result in huge losses due to missed conversions. Often it’s possible to use data science and analytics to identify the customers who are failing to convert and tailor a strategy to boost the conversion rate.

One common mistake we have found is that many companies, for the sake of gathering as much data as possible on their customers, will force them through one or more pages of long multi-field web forms. Did you know that removing a single field from a form can boost conversions by 10%? We can design solutions for your business that involve using machine learning to either pre-fill or eliminate some of those fields.

Why should I use AI for conversion rate optimisation?

If you simply remove elements from a form, without introducing AI, then you may improve your conversion rate but at the expense of the quality of the data that you are gathering from your customers.

On the other hand, AI can not only fill in missing data, but can often produce better quality data than data obtained from users directly. This is because users tend to be messy and impatient when filling out long forms, especially when the design is bad, and are likely to just put the bare minimum of information in each field. On the other hand, an AI is built on the wealth of information harvested from all past customers and can often produce cleaner values.

An improved conversion rate can deliver improved customer insights, a better ROI in your website, a more scalable business, and a better user experience. Furthermore, it can improve the trust your customers have in your business.

Examples of AI in conversion rate optimisation

For example a small-ads site could use computer vision to identify the item being sold, and pre-fill the description field. A job website can scan your CV and pre-fill every other field, simplifying signup dramatically. A dynamic pricing model can be used to pre-fill the desired price. Smart use of machine learning like this can improve a customer’s experience and deliver a huge boost to conversion rates.

Measuring the outcome of AI conversion rate optimisation

The simplest way to quantify conversion rate optimisation initiatives is to run an A/B test and monitor the conversion rates in both arms of the test.

A/B testing is not unique to conversion rates, and in fact the same techniques are applied in clinical trials. You split your website into two versions. On one version, the original, the user must fill out all elements of the form as before. On the other version, AI is used to fill out some of the data, and the user has fewer elements to fill out.

You will need to ensure that you run the A/B test for long enough to gather a statistically significant amount of data. We have found that in general each form element removed can deliver an improvement of 7%-15% in conversion rate. This effect compounds if AI can be used to infer a large amount of user data.