Disclaimer: The coverage is brought to you in real time, using a custom live blogging tool. Feel free to ask questions or leave comments for inclusion into the live coverage. During the live event, live notes will auto-scroll with newest entries at top. After the session is complete the archive version will have the oldest entries at the top. We ask you to please excuse any typos, as these are live notes.
Stay tuned! Live updates will begin momentarily.
Conversion Optimization Science
Testing alternative landing page content can dramatically increase conversion rates. But how do you come up with ideas for what to test? Should you go with straightforward A/B testing, or invest the time and resources to conduct more complex multivariate tests? How do you know when to stop? Or should testing and tuning be an ongoing optimization process? Our scientists will share their insights on these and other questions to help you get the most from your landing pages.
Moderator: Chris Sherman, Executive Editor, Search Engine Land
Q&A Moderator: Angie Schottmuller, Founder, Interactive Artisan
Scott Brinker, President & CTO, ion interactive, inc.
Sandra Niehaus, VP UX & Creative Director, Closed Loop Marketing
Khalid Saleh, President, Invesp
2:35:57 PM Keri Morgret: We're starting off with scott showing us a massive flow chart that he's thankfully printed out for us. He says it should also be available at http://searchengineland.com/a-post-click-marketing-heuristic-61806.
Starts thinking about the segment first. Questions he wants to ask:
Who are these respondents? How well can we get a sense of their identity?
What "job" are they looking to "hire" a product or service to do?
2:37:20 PM Keri Morgret: For a complex sale, who are the people that might be participating in the decision process?
2:37:50 PM Keri Morgret: A real example is looking at the query [data storage SOX compliance]
2:38:28 PM Keri Morgret: Who is looking for this? Probably someone in technical, administrative aspect of the business. They're probably from a company that is public or thinking about going public (because of SOX in query).
2:40:16 PM Keri Morgret: He's talking about Milkshake Marketing. Link at http://hbswk.hbs.edu/item/6496.html for this topic.
2:40:49 PM Keri Morgret: Showing iRobot ad that was non-traditional for Father's Day.
2:40:55 PM Keri Morgret: Looking at Goal next.
What do these respondent want or expect in this context? What do they expect when they click?
After that immediate expectation is fulfilled, what is the next logical step?
He's not a fan of a cookie-cutter landing page.
Ways your landing page could be arranged:
A single page with an immediate offer.
A multi-step conversion path to guide and segment.
A microsite to encourage focused learning and exploration. They are back and better than before, no longer flash junk. [this is example shown in image uploaded]
2:44:25 PM Keri Morgret: Segmentation: Axis of Choice. How do we give them choice? Showing example of New England Journal of Medicine.
2:45:18 PM Keri Morgret: Here's another example of segmenting on the landing page so you can give targeted information to your visitor.
2:45:40 PM Keri Morgret: And this ultimately comes down to content.
Want to try to reach people on three levels:
Rational and logical
Emotional and intuitive
Socialâ€”including social proof
2:47:10 PM Keri Morgret: Enjoying this session as he's showing a lot of real examples of what he's talking about.
2:47:32 PM Keri Morgret: He's not going to spend much time talking about the traditional things in conversion optimization, given our limits on time.
2:48:52 PM Keri Morgret: Scott mentions post that Sandra made about optimizing the thank you page. http://searchengineland.com/conversion-optimized-touch-points-the-thank-you-page-44704
2:49:31 PM Keri Morgret: Conversion optimization is working with next steps. Not just looking at a single conversion.
2:50:04 PM Keri Morgret: He acknowledges he's skipping all the parts about all of the testing, but he feels that people jump too quickly into the testing mode.
2:50:44 PM Keri Morgret: Spend more time thinking about context and the big ideas before you start testing.
2:52:02 PM Keri Morgret: Khalid Saleh is now up.
2:53:20 PM Keri Morgret: He is starting us off with a "Which Test Won" quiz for landing pages.
2:55:47 PM Keri Morgret: Conversion optimization is not about random testing, is not about prayer and just hoping something works.
Plan where to start to optimization process.
Implement optimization using a framework.
Improve in iterations
2:58:01 PM Keri Morgret: One of the biggest challenges is figuring out what you should change in the first place. Main image? Product image? Headline?
2:58:47 PM Keri Morgret: Visitor, Trust, Buying Stage, FUDS, Incentives, Engagement. This is part of their framework and how they start looking at things.
3:00:42 PM Keri Morgret: Now talking about developing personas. Logical, Impulsive, caring, aggressive. He has a blog post about this at http://www.invesp.com/blog/conversion-optimization/personas-understanding-the-4-main-temperaments.html.
3:01:45 PM Keri Morgret: He's going through things very quickly. Touching on buying stages, fears, incentives.
3:02:57 PM Keri Morgret: This screenshot is showing where to look for the various aspects he's talked about previously.
3:03:35 PM Keri Morgret: This is the fixed page, and he is explaining what they have changed.
3:04:03 PM Keri Morgret: 109% increase in conversion rates with the redesign.
3:05:08 PM Keri Morgret: Where to start? This slide shows some of the places to start looking.
3:05:28 PM Keri Morgret: Email the speaker if you'd like a copy of the slides.
3:05:39 PM Keri Morgret: email@example.com is the email address.
3:05:52 PM Keri Morgret: Last up is Sandra Niehaus.
Herding Cats: Approaches to Testing that Won't Scratch.
How to stop talking about button color, and get things moving.
done all the basics
gotten to know your audience
identified your best opportunities
Useful approaches to testing:
Think Small. We forget about the small things when we get in the mindset of looking at the huge things.
3:10:46 PM Keri Morgret: She's showing us a "before" example, and explaining how it was difficult to get at the code to test things. They changed just the copy and the "purchase now" button. That made 9% positive improvement in cart conversions. Hundreds of thousands in annual incremental revenue.
3:11:18 PM Keri Morgret: Another small example is being shown.
3:12:20 PM Keri Morgret: Increased size of headings, buttons, added header as some of the small changes. The 15 small changes resulted in 20% increase in site conversion.
3:13:29 PM Keri Morgret: Break pages into manageable chunks for testing (going modular)
3:14:40 PM Keri Morgret: They chunked up the page into top, middle, bottom.
3:15:19 PM Keri Morgret: They had different goals and things they would test for each chunk.
Showing an enhanced version of the modular approach.
What, where, and HOW IMPORTANT is each aspect.
3:16:47 PM Keri Morgret: Niehaus Wireframe Approach. Talked about at http://www.closed-loop-marketing.com/blog/2010/12/02/the-niehaus-wireframe-technique-methods-to-conversion-optimize-new-projects/.
3:17:24 PM Keri Morgret: You can chunk by function, importance, or placement.
3:17:55 PM Keri Morgret: Being persistent is a key, and has been emphasized by all three presenters.
3:19:06 PM Keri Morgret: In this case, they got good results, but didn't stop after the good results, broke things out even more granular for more testing. Another 30% increase in conversion rate after initial good reseults.
3:20:59 PM Keri Morgret: Have we herded all the cats? she's showing a matrix that she uses to figure this out.
3:22:16 PM Keri Morgret: This has a great presentation, really packed with information, and I wish I could have conveyed more of it in this coverage.
3:22:36 PM Keri Morgret: Q&A time.
3:23:02 PM Keri Morgret: How many personas is too many? Khalid: Seven, and they usually try to do just the four.
3:24:03 PM Keri Morgret: Social proof is good for credibility. Variation internationally in this.
Multiple changes: how do you know what worked?
Khalid: needs a hypothesis behind each change. Which hypothesis helped and did not help?
Scott: Recommends two different types of testing. A/B testing with radically different tests. Multivariate testing is after that.
3:29:15 PM Keri Morgret: The thing I notice here is that people aren't talking about impact of outside things like seasonality, changes in PPC campaign, etc. Even if you do make just one change at a time, you need to take other factors into consideration. This was talked about much more in the PPC sessions.
Last question: has asking more questions ever led to increased conversions (on online form)?
Break it onto multiple pages helps. Conversion rates can be higher because you weed out bad leads with all of your questions.
Extra pages in microsite can be used instead of questions in a form (select if you're student, faculty, etc.).
3:32:11 PM Keri Morgret: And we're done with the session and the conference. Thanks everyone!