Achieving the perfect conversion model on your website is no easy feat. At Hubnest, we believe the most accurate results come from testing in the form of trial and error, and prototyping. We understand the web but we don’t pretend to understand our client’s business better than them. We base our approaches on data, and the best data to measure conversion models is with A/B testing.
A/B testing is simply performing an experiment with one component of your website in order to achieve a higher conversion rate. Using different versions (A and B, or more) of the same component, A/B testing presents Page A and Page B to a controlled, similar audience, with the objective being to see whether or not the experimental variable (B) receives a higher conversion than the controlled variable (A).
We built an 30-day A/B testing environment for our client, Get-A-Rate. Two controlled environments with identical content but each offered a different user flow in their loan application call-to-action. Through our back-end tracking system and logic of how answers are provided, we served the most efficient and intuitive flow.
We tracked the fall-off points within the application process and identified the user type, device used, time-on-site to understand why this was occurring.
A Zip Code is requested in the beginning and an Email Address as a second-to-last step. There is no drop-off in users, until the last step where a phone number is requested, but at this point we already have enough data for Get-a-Rate to be in touch. After several adjustments throughout the 30 day period, the final results of the tests were dramatic. The best model converted at a staggering 28%, from their past conversion score of 4%.
If you’re asking yourself, why should you use A/B testing, the answer is simple: unless you are seeing a 100% conversion rate, there is always room for improvement to get the most of your website traffic. If conversion directly impacts your business, the website is no longer just about performance, it’s about business goals. Get experimenting today, and contact us if you need assistance.