Case Study: Usability Testing Increases Sales 34%

Case Study: Usability Testing Increases Sales 34%

How would you like to increase your sales significantly?  We wanted to share some of the things we've learned in the past few months that can do just that.

Slowing Sales Trend 

A clients of ours sells one of the leading audio Bibles on the market, The Listener's Audio Bible.  After seeing double and sometimes triple digit sales growth for 2-3 years, some disturbing trends started developing.  Toward the end of 2006 and beginning of 2007, the sales growth had began to flatten significantly.  We (us and the client) were all asking ourselves "Why are sales slowing" and "What can we do to jumpstart the sales trend."   Here are the results of what we found.  We hope you find them helpful.

Laying a Foundation

Google Analytics
We have been using Google Analytics almost exclusively to give good marketing information.  These are some of the features we are utilizing:

• Ecommerce Reporting is turned on.
• Revenue Tracking allows us to use Google Analytics to see actual sales.  This is powerful for defining where your most profitable traffic is coming from, learning which links on your home page generate the most sales and other important information.
• Campaign Tracking tracks the results of all our email and banner ad campaigns.
• Conversion Goal Tracking gives the ability to see how effective our shopping cart process is by tracking the Abandonment Rate and giving us a "funnel view" of the checkout process.

Compare Industry Standards
We reviewed our 2006 copy of MarketingSherpa's Ecommerce Benchmarking Guide to see how our site compared to industry standards.  One weakness was the higher than average Abandonment Rate of the shopping cart.  This became our mantra for the next 6 months -- lower the abandonment rate, increase traffic and increase the conversion rate and that will increase website sales.

Onsite Visit With the Client
As much as we love working in a virtual environment with email, phone and instant message, there is still no replacement for face-to-face communication.  After our initial research, we flew out to New Jersey to present our findings and to plan our strategy for change.

Set goals
Coming out of our initial research and onsite meeting, we set some key indicator goals for the next year. We're not at liberty to share those goals here, but let's just say they were realistically aggressive and covered the following areas:

• % Abandonment Rate
• % Conversion rate
• $ per Visit
• # Visits
• # Orders
• $ Revenue 

Digging In 

Usability Study: You Don't Know What You Don't Know
We were on a mission to decrease our abandonment rate.  But as designers and developers who had spend hundreds of hours on the site over the years, we knew that when we went through the purchase process, we were likely blind to how actual purchasers went through the site.  So we first had to learn how our users interact with the website.  Where were the points that they got frustrated?  Why did they leave so many carts abandoned?

To find the answers, we conducted a usability study.  Armed with a pre-qualified pool of willing subjects, Morae Usability software and a copy of Steve Krug's masterpiece, Don't Make Me Think, we reviewed and interviewed 5 different people and received some excellent insights.  The humbling part of a usability study is to see people hit a roadblock, and you slap your hand to your forehead and cry, "Duh, why didn't I see that!" Here are some of the key things we changed on the site:

• Implemented a better site search mechanism that listed relevent products near the top of the search results.  We are sold on Google's Custom Business Search Engine (see our recent blog article on CBSE) as a great, economical tool for internal search on your website.
• Made "Listen to Sample" links more prominent.  After watching how people skipped right over the "Listen to Sample," we made this feature more prominent on every product page in the store.
• Made it clear how many steps were in the checkout process, and told them what to expect each step along the way.
• Changed wording on buttons from "Submit" to "Continue Checkout" during the checkout process.
• Made the checkout buttons 25% larger.
• Provided a much clearer way for people to update quantities and remove items from their cart.  This required custom programming to the software package we are using.
• Made our error message much more visible by increasing the font size and setting the color to a more visible red.
• Price Testing. People compare prices (including shipping) to Amazon before buying. We tested dropping our retail price to be closer in line to Amazon. We tested slashing the shipping cost for specific products.  We are still testing other price points to see where the best pricepoint will fall.
• Add ISBN number.  Because people are comparing prices, we now provide the ISBN number to help them confirm that they are comparing the same products. This also helps search engine rankings when people search Google for the ISBN.

There were many other changes that we made on the website.  Some of the changes we made didn't work and we had to reverse them.  But that is the joy of testing, reviewing results, making changes and testing again.

Dropped Google Adwords
In the past, Google Adwords had been the source of significant profit to the site.  We had been seeing the return on investment (ROI) for those sales slowly dwindle over the past twelve  months--even while adjusting the campaigns for better effectiveness.  We made the decision to cut our adwords spending, expecting a drop in revenue growth, but expecting improvement on the revenue per visit and the overall return on investment from the website.

Weekly Tracking
To make sure that the changes were making a positive difference, we implemented a weekly tracking system.  It was a simple spreadsheet that tracked our key indicators. Note: We've removed the actual numbers for privacy reasons.

The most important column was "Comments/Changes."  Anytime we made an update to the website, we could see within a week the impact of that change. We had a conference all with the client once per week to review the results and plan for next weeks changes to the website.

Year to Year Comparison on a Monthly Basis
It is crucial that your numbers are placed in context.  While we had been tracking the monthly stats for years, we weren't getting the full picture for how we were doing in context with the same month the previous year.  We added a row to our monthly key indicators that showed the exact growth that we had compared to the same month the previous year.  This became our plumb-line to see if we were accomplishing the goals we had set earlier in the process.

Escalate Email Marketing
We had been dabbling with our internal, opt-in email list.  In reviewing the stats, we saw that each time we sent out an email blast to our opt-in names, sales increased dramatically. We started testing:

• Frequency: Being cautious of not wanting to overcommunicate to our loyal customers, we started sending email promotions more often--being careful to guage the open rate and unsubscribe rate.  Currently we're seeing that about twice a month is optimum.
• A/B Split: We also upgraded our client to an advanced email marketing platform that gives us the ability to do a/b split testing to test the effectiveness of subject line wording and other elements of email marketing. 

Results: The Proof is in the Pudding

After implementing the majority of the changes and recommendations we've been pleased to see the following results from the past 6 months (compared to the same 6 months last year):

• 34% increase in revenue
• 76% increase in new names
• 11% increase in $ per order
• 9% decrease in cart abandonment

Needless to say, we continue to test and guage results and are looking forward to the holiday buying season with the confidence that the current store is in better shape to grow sales than it was last year and with humility knowing there is always more to learn in this ever-changing world of website design and development.

Write a comment

Name:
Email:
Subject:
Comment: *

Security Code: