PubCon – Multivariate Testing, Conversion Tweaking
- Gillian Muessig
- Tom Leung, Product Manager, Google
- Glenn Alsup, President, Viewmark
- Philippe Lang, General Manager, SMB Solutions, LivePerson
- Rand Fishkin, CEO, SEOMoz
First speaker, Glenn Alsup, the president of Viewmark starts discussing about qualitative vs. quantitative data.
- Reject the thought of social sciences being studied like natural or physical sciences
- Qualitative researchers (Ql R) believe human behavior is always bound to the context in which it occurs
- Human behavior is personal and subjective
Quantitative researches: Believe that both natural and social sciences are testable and confirmable theories.
Glenn says Ql R is Gods Fuel:
Qualitative Researchers look at
- Usability Issues
- Effects on Goals
- Lessons Learned
How is it effective?
- Look at it as a sales funnel
- Identify all your drivers – Offline, Online, Explicit and Extract
- Take the data and group attributes based on filters
Stages in the funnel
- Weight them by the value of that event
- Downloading a white paper is more valuable than looking at a banner
- Identify when the action took place
- When you convert, those drivers are calculated in different ways
- You want to increase conversions by identifying the right mix of marketing drivers
Next up is Tom Leung, Product Manager, Google who will discuss Google’s Web site optimizer.
Tom asks the audience what they’d do after visitor lands on site and what will they do to convert them? Apparently, Driving traffic is just the beginning.
- Invest in SEO and SEM resources for 100 percent of visitors
- Your pages may lose more than half their visitors in seconds
- A majority of those who stay do not convert
Steps to take:
- Evolve with continued improvement
- Drive appropriate traffic to your site
- Measure & analyze site activity
- Test changes and implement winners
- Repeat steps 1-3 until conversation rate is 100 percent
How does testing work?
When visitors reach your site, they see a random version of your site and the testing tool will tell you what percentage of users converted based on what version of your page.
Tom states the example of Picasa. The homepage has many images and uses the word ‘free’ a lot. The second page is much cleaner and offers a trial button. Also, it says a clear value of 30% conversion.
How to set up a test?
- The control on top showss you that someone is trying to load the page
- The tracking script on the bottom tells you that they saw the page, and then you have another code on the conversion page which tells you they converted and what version of the page they were looking at
If you perform a test, each version of the page has a unique sticker for you to identify it by. After the test runs for awhile, Google will populate reports for you.
Best testing practices:
- Test a small number of variations
- Test big changes. If you don’t notice the difference between 2 combos within 8 seconds, visitors probably won’t either
- Consider early indicators if you don’t have enough conversions
- Don’t jump to conclusions. Tom says that less than 2 weeks is no good, focus on absolute conversion difference, don’t get too excited by sliver of green.
Additional Testing Ideas:
- Conversion Cocktail
- Trust seals
- Which testimonials
- Inspirational or fact-based pitch
- YouTube video
- Navigation bar
Next up is Philippe Lang, the General Manager of SMB Solutions, LivePerson.
Philippe suggests using a lot of information from analytics to know more about the visitors who come to your site. There’s a catch, as valuable as the data is, it can be tough to understand. Talk to your customers and ask them for their honest feedback.
- Points to watch out for when you monitor your Web site
- Keep an eye on visitor referrer information such as
- Search engine used
- Marketing campaign
- Navigation behavior (page views, shopping cart content, and shopping cart abandonment)
Combine both real time monitoring and feedback data by involving your customers into a chat. Keep a watch on the patterns or trends.
Ultimate Feedback Look
- Begin by monitoring the visitors who come to your site
- Identify visitors who have unexpected behaviors
- Involve visitors by listening
- Translate customer feedback into action
- Refine and optimize your Web site
Philippe has 2 case studies.
1: Online University degree
Challenge: High abandonment rate in the online inquiry form
Reason: Customers did not appreciate the question about age of applicant. This led to customers leaving the form
Solution: Question deleted
Result: 20% increase
2: Outdoor gear retailer
Challenge: Sunglasses faced low conversion rates
Reason: Customers confused over size chart
Solution: Size chart redesigned
Result: 25% increase in conversion rates
If you know what’s wrong, you can fix it! Testing is great as it
- Heightens conversion rates
- Improves the sales process
- Increases average order value
- Forms relationships with customers
Finally, it’s Rand Fishkin, CEO, SEOMoz. He starts by mentioning the landing page contest his website had where readers could submit landing pages. After scouring through many submissions, the ended up testing 10. after going through all of them, one finds out that the page where you had to scroll down to get to the bottom of the page had the best conversion scores.
After the competition, he and his team tweaked the winner’s landing page to convert even better.
- If you think it will work, it probably won’t
- Landing page design is not universal
- Testing is the only way to improve
- Doubling your conversion rate is far easier than doubling your traffic
This was followed by a Q & A round where some of the most exciting things said were:-
Rand:– "conversion testing is a legal form of cloaking…"