How I got from 58% to 38% Bounce Rate
I can recall the day I visited a hospital and wandered in its corridors for forty good minutes to find the clinic I was suppose to reach thirty minutes earlier. Tucked into the crook of my arm was the book “The Design of Everyday Things” by Don Norman that gave me the confidence to make my point loud and clear in front of the facility management of the hospital that day ; the direction signs, (signifiers as Norman calls it) were either missing or not up to date.
Ever since I started working in digital domain, one thing that fascinated me the most was the digital footprint and the sophisticated tools available these days that enable tracking of customer moves so one can plan by analyzing this trail of data that customer leaves behind. And as I started working on my first digital product from scratch, I couldn’t wait to apply everything I learnt over the period of time.
It had been 20 days since the product was live but the bounce rate was soaring at 55%.
When analyzing data for improvement and optimization, you got to segment the data to channelize your efforts or else as Easil stated,
“Its like going fishing with a huge net, sure you will catch a lot of fish. So much so that you won’t notice the big fish flopping around on the deck until it manages to flip itself over the side.”
To sort the 55% bounce rate, I individually tracked behavior of users acquired from different channels (Display, Facebook, SMS and Direct), desktop and mobile through Google Analytics. This helped me identify that desktop users bounce rate was 10% lower than mobile users. To understand why users on desktop were happier, I used heat-maps. Sumo and CrazyEgg are few of the best tools out there to visually analyze users digital footprint.
The snapshot is of the desktop view. The users were interacting most with the elements (Signifiers) in the left panel to reach their goals, hence lower bounce rate and longer session time compared to mobile where these elements were hidden inside the menu.
The data points acquired through user’s behavior on desktop helped me form the hypothesis and MCS (Minimum criteria for success) for the first design experiment.
If we bring the most clickable elements above the fold on mobile, we believe it would help users reach their goal because of instant visual interaction and the metric bounce rate would reduce by 10%
It was a ‘low risk’ and ‘low difficulty’ experiment, but every experiment comes has its cost. I broke down the cost as development, design, advertising and time cost. You need to ask yourself if the benefits would outweigh the cost in terms of improved metrics, increased revenue or new learning.
Two variants were designed and made live along with the original mobile landing page for A/B testing and the bounce rate metric was tracked for 10 days while keeping all the other factors constant that included, the channel of the traffic, the content and it’s sequence on the landing page, the creatives and copy used to drive the traffic on the page.
The hypothesis was proven true with over 15% reduction in bounce rate and improved session time.
Coming back to the design principles from where I started off, people search for clues and guidance that can help them determine and understand , the correct design and placement of signs in the digital store are now like employees in an apparel outlet or in Apple’s Fifth Avenue store who help you find the perfect dress or phone. It is your job as a product manager or designer to provide those clues that can help customers reach their goals.
I especially like to thank