It is day one for sessions at the NYC Search Engine Strategies Conference being held this week at the NY Hilton. There are over 5000 attendees, and believe me, every session I've sat in on has been overflowing with people. There is strong interest in Advanced topics, with a keen eye on what the future holds for the Search Marketing industry.
I'm delighted to report that although I'm not there in any official capacity as a usability speaker, the relationship of user experience, persuasive design and conversions is inching its way into many sessions and conversations.
The following is are my takeaways from a talk called "Pay Per Conversation" from today's sessions.
Moderator: Jeff Rohrs, ExactTarget
Jeffrey Eisenberg, FutureNow
Sandra Cheng, Product Mgr, Google
Jeffrey is standing in for his brother, Bryan Eisenberg, because his wife just went into labor. This room is filled to capacity. Unlike the first session I attended this morning (SEO: Where to Next?"), where it was overflowing and I had to sit on the floor, this time I'm firmly planted in the front row.
Jeffrey leads off the session by having us consider the typical PPC (Pay per click) term and reworking it to "pay per conversation". The reason for this is that we don't want to make money per click. We make money per conversion. He also pondered that rather than "SEO", we view it as "search experience optimization". As he says, "There is a signal by the searcher and if we hear it right, we can get them to take action." So let's look at his ideas on how we can listen and meet user/searcher expectations:
1. Web Analytics consulting can show you were the disconnects are. These are where our user expectations are not met. Did they search on a phrase and your page appear but not meet what they expected? Your goal here is to study your data to discover traffic patterns and unmet needs.
2. There were several references to what we now called "information scent". While there are many different definitions, essentially "scent" are cues to your visitors to keep them interested. They'll stick around when pages are relevant and click paths are goal driven. Links, for example, should contain content that will promise to take your users where they want or need to go. Where there is scent, there is momentum. Study your drop off data and note where they have lost the scent of what put them on the trail to your site. Information Scent creates motivation. The objective is to avoid losing so many users up front. Pay attention to the signals users are looking for.
3. Keywords don't fail to convert. Rather, how pages are relevant to keyword searches makes the difference.
4. Personas. There are simple vs robust" personalities, logical vs emotional, quick vs deliberate, methodical vs spontaneous. User behavior absolutely is taken into consideration when determining design, content, and tasks. Eye tracking studies have proved different patterns of page usage can be traced to different types of personalities. Some people will find something quickly and then move on. Motivation is where users will focus. Geico appeals to an emotional need with the use of its lizard, who promises to save you hundreds of dollars. However, if you have sold the idea and then present your visitor with a technical and complicated form to fill out, you risk losing your visitor.
5. (I loved this!) "Plan, improve, measure and plan again, over and over...." Conversions are a continuous improvement process. Align customers with business objectives and consider their behavior patterns. By studying patterns, this changes what you add and enhance on your web pages. Don't throw anything against the wall and see what sticks.
6. There is a White Paper on creating personas on FutureNow. Their Twitter account is twitter@thegrok Jeffrey had a great set of slides to compliment his talk and also showed several case studies from some big brand companies, where something as simple as changing the content on a call to action button increased conversions. He illustrated the value of testing several pages to see what converted best.
Sandra was next. She's an excellent speaker. Very clear and easy to follow. She strongly emphasized the value of using Google Analytics in her presentation but any testing is fine. She showed how testing can help you figure out what works best for your users. Some highlights from her talk include:
1. Not everybody who comes to your site will do what you want them to do (but lets try to nab what we can).
2. Try to avoid bounces and abandonment by making sure the directions are clear.
3. The best way to understand what's confusing is to watch a friend use your web site. You know where everything is already. Where did they get stuck?
4. Get analytics to get data. You need an idea where people are going on your site and what happens when they get there. Reports will prompt you to ask the right questions about your traffic. Data shows where they come from. You will ask questions, get answers and make fixes, all with the help of your reports. Look at landing pages. Where do they enter your site? Google Analytics shows bounce rates. Do they land and leave before clicking anywhere? Bounce rates represents opportunity. What changes can you make to help them to stay? Look at funnel reports. These are goal paths and show Where they come in and out/ Page leaks (where users leave) are also opportunities to patch leaks.
5. Internal site search are great sources that show customer intent. People type into your site what they are looking for. This is how they tell you what they want. Study where they go when the search and also investigate where they left. Look at search terms and un-met needs. Perhaps the search results page is too confusing. or you don't carry the product they're looking for. You can get a great picture of customer intent with on-site search.
6. Test page or ad copy. Look at non-paid keywords and bounce rates and compare with paid keyword strategies.
7. Let visitors design your pages for you. Compare page content by making and testing 3 variations of the page. Google will track responses and show you the winning combination. Run mulit-variate testing. Test an image vs without an image near a call to action prompt. It is not rare for 20-50% conversions improvement by making small changes that appear in this kind of testing
8. Remember that "best practices" aren't always the best for YOUR site.