How To Use A Quiz To Get More Opt-ins

[Jim Stone]

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

[Jason Moffatt]

Good write up.

A good model to pay attention to is the FreeIQ tests. I'm sure many of you have taken them. I heard somewhere that their opt in rate is something like 90%. (Just what I heard, and I have no verifiable stats to back it up)

However it does make sense. After spending all that time filling out the damn IQ test, I sure as heck wanted the results. So of course, like 9 out of 10 people I opted in to find out I was a couple points shy of being a Jenius!

[Jim Stone]

Jason, yeah, IQ tests are a great example.

I wonder if it's 90% of visitors who opt in, or 90% of those who finish the quiz.

Either way that's great. I usually have 40-50% of those who finish a quiz sign up.

Jim

[Richard Tunnah]

Hi Jim,
Thanks this is interesting as I've been considering testing a quiz opt-in to replace my bog standard ones. Also I find intriguing you offer no incentive to get the opt-in apart from the answers.

RIch

[TheNightOwl]

John Taylor's Psychic Optin is worth checking out, too, for a more no-frills (but still effective) solution.

Nice post, Jim. VERY useful stuff.

Thank you!

TheNightOwl

[Kael41]

Since developing my quiz plugin for wordpress and implementing across a wide range of niches, i've seen between a 10-15% increase in CPA and Google Adsense revenue just from psychological engineering.

We're all geared up internally to expect a prize from taking some sort of online quiz

[Kevin-VirtualProfitCenter]

This post is awesome. Any chance I can reprint this on my blog along with appropriate credit and a link back to any site of your choice?

[Jim Stone]

Hey Kevin.

Feel free. I don't know if you also need any permission from the Warrior Forum, though.

I'd love a link to http://www.segmentex.com/coming_soon.php with a link text of "optin quiz".


Nightowl: John Taylor also has a nice little guide for writing quiz questions that he offers with his software.

Kael, I'll have to check out your WP plugin.

Here's the thing. I know I'm only starting to scratch the surface of what can be done with quizzes. From getting leads to market research to segmentation to cross selling to getting testimonials to guiding much of the visitor experience. It's a great interactive way to get the visitor/customer to tell you exactly what they need.

Jim

[FriendlyRob]

This is something that I have never thought of before. I will try to test it on one of my pages.