April 21, 2003
Case Study

How a Travel Site Raised Sales Conversions 30% With 26 Little Site Design Tweaks

SUMMARY: Are you daunted by the idea of improving your company Web site so more visitors buy? You know that a few tweaks can make a world of difference, but which tweaks exactly?



This Case Study features practical step-by-step details of how 11thHourVacations.com studied their metrics and conducted little "tweak tests" to come up with the 26 small changes that raised conversions by 30%.



One simple change alone resulted in $250,000 more sales this quarter. Especially useful for eretailers.
CHALLENGE
11thHourVacations.com started tapping the online audience back in 1996, selling distressed travel inventory, primarily quality tours and cruise packages. They have done pretty well with it, too. After all, about half of all US Internet users have made a travel purchase online. It is a hot market.

In Spring 2002, CEO Mike Putman took a long, hard look at the website and realized it was time for a change.

He says, "We were very successful at attracting traffic, but our conversion rates were not terribly good. We saw all this traffic but had few sales in comparison."

He set about boosting sales by tackling the conversion leaks in the site.

Here is how he did it.

CAMPAIGN
Putman and his in-house developers did most of the original site design. They are travel merchants, not usability gurus, and did not have the expertise or tools to take the site to a new level. They then contracted NetConversions to help with the usability study.

-> Step 1: Analyze current site performance

In a first phase the team installed custom code on the site to track key aspects of visitor behavior. These covered numerous factors, but fell into five categories:

1. Scrolling behavior: For example, how much of a page is actually seen by a visitor?

2. Field analysis: Looking at how people fill out forms. For example, what fields get filled first, what last, and when do people abandon the form (critical for improving lead capture or shopping cart completions)?

NetConversions' Project Leader, Adrian Chiu makes the point that knowing which page leads to shopper abandonment is only half the story, you also need to know what aspect of that page is causing them to leave.

3. Click density: For example, where are people clicking (whether on white space, a button or link) and which clicks are the ones driving the most success?

4. Link analysis: For example, which links are most popular?

5. Path analysis: How are people moving through the site?

This produced a wealth of data on visitor behavior and the parts of the site contributing most (and least) to sales success.

For example, they found that over half of visitors did not scroll on the homepage, meaning a whole chuck of information was effectively invisible.

Chiu says understanding the importance of screen real estate is vital. It is not just about monitor resolutions and page length. For example, 7% of site visitors resized their browser window at some point in the session.

They also found certain site elements drove conversions much more effectively than others. For example, recommendation-based travel options converted visitors twice as well as other site features.


-> Step 2: Turn data into practical recommendations

The tricky part of the process is turning that huge amount of data into actionable recommendations. The team started by looking for the "low-hanging fruit," using path analysis to see where the major visitor leaks were.

They supplemented this by tagging recorded clicks according to the success criteria (where did non-buyers click, where did buyers click). This showed which links and site areas worked and which did not.

Then they blended in all the other more specific data to gain an understanding of the whole user experience and to draw out a final collection of two to three hundred factors that might be losing the site visitors and conversions.

Chiu stresses the importance of going back to the site to better understand whether a particular user behavior is a real problem or not, and what the causes and solutions might be.

He says, "We'd find a certain percentage of users wouldn't see beyond a certain part of a page. So we'd go to that page to discover what they're not seeing. It might be that a submit button is at the bottom of the page and X% of users aren't scrolling to it. That might mean they're simply not qualified to fill out the registration form or it might mean the form itself may not be engaging and calling the users to action."

Putman was able to discard some of the suggestions based on his more intimate understanding of the business (never assume an outside service has a perfect understanding of your market and operation). Chiu and Putman's teams brainstormed together to come up with a priority list of some 40 key variables to test.


-> Step 3: Test, test and test again

Each variable was tested independently of the others. Visitors to the site were split into two groups. One group saw an original site version, one saw the same site but with one particular change made, such as a different color buy button. Then the team compared the two groups in terms of sales.

This step-by-step approach is vital, says Chiu, because of the so-called rule of thirds: One third of your changes will keep things the same, one third will improve things and one third might make things worse.

If you do not look at each variable on its own, you might find a group of changes producing a net positive impact, but some of those changes might actually be hurting your business.

Putman says, "Some of the things that we thought were going to work didn't, they even had a negative impact."

For example, they tested putting a link on the site's advanced search page, leading to a high-converting Ideal Vacation Finder. It bombed.

Other examples:

- The site's top 11 recommendations page has three lists: Top 11 cruise destinations, top 11 vacation destinations, and most recent 11 email alerts. The team tested the impact of putting the cruise list top, or the vacation list top. They found the top list always got five times the clickthrough of the second list, regardless of which particular travel product it covered.

- On the first page of the booking process, they tested adding an image of a person, a testimonial and a call to action ("Step 1: Fill in your contact info.") The result was an 11-fold increase in completions.

Speaking of bookings, Chiu gave us three quick tips for shopping cart forms:

Tip #1. Take care with form fields which require some kind of formatting. He says, "If you ask for a phone number and you have one big field, do they put in the dash marks or parenthesis, or does it matter? Just a simple little direction helps so much."

Provide an example, or customize the form error message so that the visitor learns exactly how they should be formatting their phone number.

Tip #2. If you require a shipping and billing address, and let visitors check a "Click here if shipping address is the same as the billing address" box, then make sure your script automatically fills out the shipping address fields for the visitor.

If it does not, Chiu says visitors can assume they have to fill out the fields themselves (i.e. more reason for them to give up) or they fail to fill out some fields because they assume the check box takes care of everything.

Tip #3. Get someone who has never seen your site to go through the process. Chiu says that will identify the biggest problems right away.


-> Step 4: Implement changes

Based on the test results, Putman made 26 changes to the website. Some examples:

- He gave the recommendation engine more prominence on the home page and designed specific email campaigns to drive traffic to that high-converting feature.

- He pulled all of the higher margin link paths to the top of the home page, and moved less successful links, like "air only" and "car rental" below the fold.

- He added two abandonment popups to the site. One for visitors who get four or five layers deep into the site before leaving. This popup tells them "you can put this vacation on hold for 24 hours at no cost and no obligation."

A second popup activates if a prospect abandons the actual booking process, and offers them a $25 discount if they will finish the transaction.

RESULTS
The changes implemented increased the site's conversion rates by about 30%. Putman notes they do not have the systems in place to measure the ongoing independent impact of each single change they made, but he is more than happy with the net outcome. (Remember, each change was based on test results anyway.)

He *has* tracked the impact of abandonment pop-ups, and says they've added $250,000 in revenues over the last four months.

Most importantly, he says there is no question that the extra money coming in justified the cost of the project. In fact, he is planning to repeat the process in the near future.

With all the travel industry problems caused by Iraq, SARS etc., it is welcomed timing for Putman. He adds, "Business is so contrary to what should be happening. People ask me all the time how business is, and I say, for example, we did more in Q1 this year than we did in the first 7 months of 2002."

http://www.11thhourvacations.com
http://www.netconversions.com

Improve Your Marketing

Join our thousands of weekly case study readers.

Enter your email below to receive MarketingSherpa news, updates, and promotions:

Note: Already a subscriber? Want to add a subscription?
Click Here to Manage Subscriptions