Case Background :
This case study presents how we discovered a variant of product landing page that converts 45% better than its predecessor variant!
The website is built on Shopify, but they have a separate landing page which is built on hard core html code base.
The business is in the niche of CBD based Pet Food chewable bites.
The traffic acquisition channels are primarily social media.
The website uses Google Analytics + Google Optimize combination for analysis and Optimization.
We additionally installed Microsoft Clarity on the site for sessions recordings and user interaction heat maps.
The landing page receives on an average 50,000+ every month and prior to experiment was converting under 3%.
We use this Google Data Studio dashboard to do a Quatitative analysis of the site :
The link is publicly accessbile, so feel free to use it 😉
Here is what we found :
Qualitative Analysis :
For qualitative analysis we have curated a checklist of over 60 parameters that we have developed post exhaustive experimentation across 200+ projects. We have found that only by implementing the guidelines from this qualitative audit, organizations can expect at least 30% bump in their conversion rates with statistical significance.
Following presents excerpts from out audit deck which is usually over 100+ pages :
Following were the key discoveries from qualitative audit :
Soution Delivered :
Since the site has limited traffic, taking the route of experimenting one change at a time would take eons. We rather suggested the company to completely revamp the page and drive a split test between the old and the new variants.
The team at DataVinci then went on drawing board to design a new variant from scratch.
Following is the sequence of steps taken :
Leveraging Google Optimize, the team launched a split test to statistically compare the performance between the existing lander and our new lander. It took Google Optimize 25 days to declare the variant a winner with 96% probability.
The test was ran for 25 Days to ensure statistical significance.
The challenger created by DataVinci outperformed default by 45%
Variant declared winner by Google Optimize with 96% probability
Closing Notes :
Fundamentally, when we call something an experiment – we cannot be certain of the outcome. Hence, with A/B tests or split tests as well – we must accept that there can be a possibility that the test will not produce a high converting variant.
But, if you want to ensure a consistently high performing digital asset, then testing and optimizing MUST be in the culture of your organization. Having said that, here are few things to consider to improve chances of success :
- Work with an experienced team. A good team with experience develops decent intuition on what will work and what will not. Saving considerable time as testing can be long sometimes
- Page speed performance is a low hanging fruit
- Work on your offer first and then on your website, no amount of CRO can cover up for a bad offer
- Decide your tests based on your traffic volume, don’t do minor tests with low volumes, it will take you ages to disover a better variant
- Design your page copy based on your key traffic channels. Someone coming through search is perhaps already aware of your product category, while social traffic needs more education
We hope you found this case useful. If you want to discuss any Analytics or CRO project with us, you can use the link below :