How To Use A Quiz To Get More Opt-ins

Ever since Eben Pagan released his Psychic Sales Letter software with his Guru Mastermind program, I've been using quizzes to get optins with good success.

I never did use Eben's software, but began developing my own tools.

I've also been split testing various aspects of the quiz-based optin proces, and wanted to share some of my main findings here.

1. Your first question is very important. That's what draws the visitor into the process. Example. I had a quiz for a site that teaches people how to flirt better. Originally the first question was something like: "The most important aspect of flirting is, . . ." followed by some fairly complex responses. Only about 50% of visitors participated. When I changed the question to "Which sex would you rather flirt with . . ." the participation rate went up to over 80%.

2. It seems to work better to have the quiz questions occur one at a time in an iframe, rather than being all laid out on the page. One optin page went from 12% to 15% with just this change alone.

3. If you can get the optin form itself into that same iframe at the end of the quiz that seems to work best, and I saw a bump from 18% to 25% by doing that.

NOTE: all of these numbers come from traffic on the Google Content Network, and are thus lower than they would be coming from endorsed traffic.

4. Make the quiz either an assessment of the person's current knowledge or skills in an area related to your main product, or an opportunity for them to make expert choices about which widget is good, and which is not good. That way you can use the answers to the quiz as your opt-in bonus. I've found that often this is the only bonus you need to get good optin rates.

5. When writing multiple choice questions, make sure your questions allow everyone exactly one best answer per question. Technically speaking, the answers should form a partition on answer space. They should cover all the possibilities, and not leave someone wanting to answer two of the questions equally. Here's a bad question: Which food do you like? [pizza][pie][vegetables] -- a visitor might like more than one of thse, or none of them. Here is a better question: Of the following, which do you like MOST: [pizza][pie][vegetables].

Just read through your questions after you write them, and ask yourself if some people might not have an answer, or might want to answer more than one.

6. I've found that you can sneak a market research question in toward the end of the quiz without much dropoff in completion rate at all. Once a person has answered a few questions, and anticipates receiving a scored assessment, they're more than happy to tell you what their biggest frustration is, or what product they would love to buy from you if you had it. These kinds of questions can tell you exactly which product to develop next, or exactly how to push their buttons on the sales page.

7. If you plan carefully, you can also segment your visitors while you quiz them. If you find out that some of your visitors are guys, and some are girls when they take your flirting quiz, you can get them on appropriate lists, and create the proper visitor experience for them. If you hide these "segmenting" questions in the quiz, you're much more likely to get honest answers to the segmenting questions, and they don't even notice they're being segmented.

8. 5-10 questions seems to work best. There's a tradeoff between investment and attrition. If the quiz is too short, they're less likely to opt in at the end, because they're not as invested in the outcome. If the quiz is too long, there's more chance they'll just tire of the process and drop off at some point.

Hope that gets someone started. I'd also love to hear other people's tips if you've been using quizzes.

Sincerely,

Jim