PORTFOLIO

Global eCommerce Cart Optimization
The $13 Million Dollar Idea
Background & Initial Analysis

I joined Symantec when they had just launched their eCommerce platform in-house. It was an exciting time as now we had access to a multitude of data that we could research for revenue opportunities.

 

While one team looked at optimizing landing pages for revenue conversions, I worked with the Business Intelligence team to search for additional opportunities.  One such opportunity was the shopping cart experience.  Together, we analyzed baseline data (comparing Quarter over Quarter performance) and built a sensitivity model:

$1Million 
Site Optimization Idea?!

All those numbers and charts are enough to make any designer's head spin, but together with my wonderful analytics partners, we identified the Cart-to-Billing pages in the eCommerce flow to be a revenue optimization opportunity.  Essentially, if we could improve the customer experience and conversion from the Cart to the Billing pages by 1.5%, this could possibly be a $1 Million Dollar idea!

Baseline Visual Analysis

From a design standpoint, I had often thought the shopping cart experience could use a little bit of love in the Customer Experience and Look/Feel department.  Now that I had some hard analytics to justify officially reviewing the design, I could get to work!

 

Here are the Cart and Billing pages that I had to start with:

Control: Cart Page Control: Billing Page

I believe even non-designers can see the opportunity here.

 

As excited as I was to work on this, I had to be conservative in my initial design approach. The eCommerce shopping cart experience is a high-traffic website flow, and many eyes and departments were watching everything on these pages.

Hypotheses

My general hypothesis about the eCommerce Cart and Billing pages was that the User Experience was confusing to our customers.

 

Specifically, on the Cart page, it is not obviously clear what the customer should do to advance their experience forward. Below are some things I noted about the original Cart page:

Multiple Navigation Headers
Are these relevant in the checkout experience or do they contribute to customer exit rate? Cross-Sell Area
What is the take rate on the cross-sell?
Does this area detract from conversion?
Does it confuse customers to see the button in this area?
Call-to-Action Button
Does it stand out as a design element? 
Does it need to be orange?
Are the lock icon and button message assuraning and compelling to our customers?

With the Billing Page, it is also something of a confusing experience. Below are a few things that I noted that could be potentially contributing to customer fallout rates:

Multiple Navigation Headers
Are these relevant in the checkout experience or do they contribute to customer exit rate? Account Login
These fields are for returning customers to enter in their username and password, yet are not clearly labelled or communicate any obvious value to customers Space between Form Fields
Legibility could possibly be more clear if there was more space between this cluster of fields. As a designer I feel some anxiety when I look at all these fields clumped so closely together.
Purchase Total
This seems like one of the most important things a customer would care about on the page, but design-wise is lost in the shuffle. Call-to-Action Button
Does it stand out as a design element? 
Does it need to be orange?
Are the lock icon and button message assuring and compelling to our customers?

There were many, many more things that I noticed, but I narrowed it down to three hypotheses that I would test:

 

  • If we optimize the Call-to-Action (CTA) elements (i.e. button size and message) the close rate on the cart page will increase.
  • If we clarify the information on the Billing Page fields, close rate on the billing page will increase.
  • If we remove the top navigation on the Cart and Billing pages, close rate on the cart and billing pages will increase due to reduced exit rate.
Test Designs & Results
Call-to-Action Button Multivariate Test
on Cart & Billing Pages

Hypothesis: If we optimize the Call-to-Action (CTA) elements (i.e. button size and message) the close rate on the cart page will increase.

 

In designing this test, I leveraged my experience from the CNET/Download.com Display Ad Optimization project. The main learning I had there was that message, size, and color on the CTA button did affect conversion. In layman's terms, "Make the button bigger!"

 

This test was set up as a Multivariate (MVT) Test, which meant that we could create 9 different test experiences that would simulate 81 experiences and predict a winner. This was the first historical MVT test to run on Symantec's brand-spanking-new eCommerce platform.

Final Result: Blue Buttons Win $500k/Q

With so many combinations, it was surprising that Blue was the preferred button to push, and increased conversion by 9% over the original experience!  An analysis by the Business Intelligence team confirmed that this earned $500k/Quarter for eCommerce.

Billing Page Account Messaging Test

Hypothesis: If we clarify the information on the Billing Page fields, close rate on the billing page will increase.

 

For the design of this test, I requested was that we separate out the fields for the Account Login and create legally approved text to explain the value to the customer.


I worked with the Program Management team in designing this one. They were instrumental in providing the data for the business case.

Final Result: Clarified Fields & Copy Win $500k/Q

It almost seems like a no-brainer that this test would be a winner, but you never know until you try! There were actually two test experiences, both of which yielded a positive increase in close rate (+1.4% and +3.64%) over the original experience. An analysis by the Business Intelligence team confirmed that this earned $500k/Quarter for eCommerce.

Navigation Removal Test
on Cart & Billing Pages

Hypothesis: If we remove the top navigation on the Cart and Billing pages, close rate on the cart and billing pages will increase due to reduced exit rate.

 

This test was designed to address a common practice heard in eCommerce, which is to remove any excess navigation in order to reduce the exit rate on the cart flow.  It ran as a test on the Brazil store.

Final Result: Removing Navigation Didn't Matter

The test results were flat for this one. It wasn't compelling enough to re-run the test in any market other than Brazil at the time.

Final Results

These series of tests ran for one Quarter.  They brought huge learnings to the eCommerce team, in both test ideas as well as platform performance.  The yield for the first quarter was +$1 Million Quarter over Quarter, and resulted in changing the buttons on the eCommerce platform to change from orange to blue for the US and UK markets.

 

The following quarter, I collaborated with the eCommerce team in Europe to design a Global Cart Optimization program. Together, we continued to iterate tests in the UK and US markets and furthered our learnings and earnings. The yield of this program was +$13 Million Year over Year (again, confirmed by our friends in the Business Intelligence team).

 

As a designer, I never imagined anything I worked on would be worth so much. It was an amazing experience!

THANK YOU

This project was a tremendous global cross-functional effort. I would like to give a big "THANK YOU" and acknowledge all the parties involved:

 

  • Mark Wynne
  • Mervyn Alamgir
  • John Barrington
  • Alan Barry
  • Jen Dante
  • David Denham
  • Vishnupriya Devabhaktuni
  • Suzie Farrell
  • Justin Gray
  • Brandon Isom
  • Natalie Kini
  • Danka Lochowicz
  • Peter McRae
  • Patrick Mellinger
  • Patrick Ng
  • Sandeep Rao
  • Zunaira Qureshi
  • Andrew Stepner
  • Oleg Zhukhovitskiy

 

(If I missed anyone, please send me an email and I will update this list!)

Logos used in the portfolio area are © & ™ their respective companies.

All other content and artwork on this website is © & ™ 1999 - 2015 P.M.B.Q. Studios.