July 22, 2010 5:18 pm
One of the biggest challenges I face when starting a project with a new client is getting a green light for a user-testing budget. It’s not a matter of whether or not the client thinks the data would be valuable - most of you understand the importance of user-centered design now that it’s become common practice. But what alternatives do we have when necessity isn’t enough to overcome budgetary hurdles? Enter the champion: A/B testing.
User experience (UX) professionals have been using the A/B testing model for ages, and in fact, this approach originated long before the existence of the internet. I’d even bet a big bag of money that the invention of the wheel involved A/B testing methodologies. For those of you who aren’t yet in the know, A/B testing is a method that uses 2 or more live approaches to achieve the same goal over a period of time while gathering data about the success of each approach.
Sounds easy enough, right? Well, it is! However, we all want to make the most of our A/B testing efforts. Here are 5 ways to do just that:
1. Know what you’re measuring.
This seems like a no-brainer, but there’s a little more to consider than “we need more sales,” to make your online endeavor successful.
• Focus on one page or function at a time, not your entire website.
• Using business goals you’ve defined via user research and stakeholder interviews, make a list of exactly what you need to achieve and make that list as detailed as possible. Common desired results are things like decreasing bounce rates, increasing customer conversions, usage of a new tool or widget — you get the idea.
• Define numerical or statistical benchmarks and be specific - data is far less valuable when you don’t start with real numbers.
• If you’re testing a new version of an old page or function, record your baseline control sample data. You need to have detailed information on the performance of your old page to help inform the goals you have for your new page. Plus, your old data is still extremely important even if you don’t reach your defined benchmarks for success. Perhaps your new page still performed better than your old one —or (*GASP!*) — your new page could actually be less successful than your old one and you’d never know!
2. Find out what your competitors are doing.
Ever heard of a website called compete.com? Well, you have now! Websites such as compete.com and gomez.com are fantastic resources for seeing how your site measures up against other sites within your industry. Much of the data can be had for free, but for a nominal fee you can get more detailed reports. Whether you stick with the free information or spring for the details, the numbers you get from these sites are extremely valuable for setting benchmarks, especially if you’re new to your marketplace.
3. Make more landing pages.
If your goal is to drive more traffic and increase sales by launching a new product or service, effective landing pages should be an integral part of your strategy. Testing more than one landing page won’t just help you determine what design elements, content and action items are most effective — when executed properly (using different content, page titles, etc.) you’ll uncover a bounty of information about your SEM strategy.
Designing a ton of landing pages might sound expensive, but it really doesn’t have to be. Reusing a design template and simply rewriting the content and swapping out a few key elements is a cheap and effective approach for this tactic.
Need it to be even easier? Try using a service such as unbounce.com that will allow you to create multiple landing pages that look clean and require no HTML knowledge at all. Unbounce has a subscription pricing structure based on site traffic and branding requirements that range from $0-$250 per month. Added bonus? They have an easy-to-use A/B testing interface feature that tracks data and tells you which page version is the best performer.
4. Make friends with Google Analytics.
Most people are familiar with and using Google Analytics to track their site statistics by now, and if you aren’t, you should be! It’s free, it’s intuitive, and it’s robust. Google Analytics tracks basic data such as visitors, referrals, and bounce rates just as any web statistics tool does, free or not, but with additional in-depth functions such as advanced segmentation, custom variables, reports and alert capabilities as well as a continuously growing library of add-on applications, you can’t afford not to use this tool. If you’re already using Google Analytics but haven’t yet explored the depths of its capabilities, make the effort to delve deeper. The return you’ll get from becoming a Google Analytics expert will far outweigh the time and energy you’ll put into familiarizing yourself with the intricacies of the tools.
5. Educate yourself.
A/B testing can be as basic or as detailed as you want it to be. The more knowledge the better applies to every business decision we make, and A/B testing is no exception. Here are some resources to get you started:
• Always be Testing : The Complete Guide to Google Website Optimizer
• Landing Page Optimization: The Definitive Guide to Testing and Tuning for Conversions
• Web Analytics 2.0: The Art of Online Accountability and Science of Customer Centricity
Now that there are so many accessible tools and resources for testing web concepts and measuring success, even the tiniest of shoestring budgets can afford to conduct A/B testing and collect meaningful, valuable data that will ultimately ensure online marketing efforts have higher returns than you ever thought possible. All that data for such a small investment - you just can’t put a price tag on that — but if you decided to try, you can now measure its online efficacy for practically nothing.