Personality tests are the most viral quiz types on the internet. They have been around for decades and never lose their appeal. Quiz marketing experts Boris and Mike share why they’re so effective – and how to create the perfect personality test for your own marketing needs.
Get your free Quizmaster e-Book here.
Take the test “”hat breed of dog are you?
Hi there, and welcome to this episode of “The Quiz Makers”. My name is Mike Hawkins, and I am one of the co-founders here at riddle.com. And today, I’m here again with my co-founder, Boris.
Hey, Mike, good to be back.
Good to have you.
In this episode, Boris and I are going to talk about personality tests, why they work, how they work, and how to make a really good one for whatever marketing objectives or whatever business you’re in.
But before we dive in, I think we probably should take a step back and talk a little bit about personality tests, and how they’re used offline, and how they can be used online.
So Boris, I know we talk a lot about this when people ask us on our customer support chat at Riddle.com. You and I both do most of the customer support, where they’re often coming to us with questions like “Hey, I want to do a Likert scale personality test”.
You can tell they’ve got this idea offline, like those paper tests from back in the day. They’ll give a number of questions with the same scale – say, 10 questions and for each, give your answers from strongly agree to strongly disagree.
For example, do you agree with these statements: “I am brave.” Next: “I am proud.” Question, question, question. That’s one way to do it. But isn’t there a better way to kind of tackle this personality test concept?
Right? If you think about it, when you offer tests on paper, it’s really the only way to do and evaluate these tests. Because you need to count how many ‘strongly agree’ versus ‘strongly disagree’ answers and then get to a personality result. You can’t really run any logic behind these questions on paper.
So a lot of these tests – when they’re brought online work just like that. And a little backstory, we both worked at a company called Tickle.com, which was really big in creating their own online test portal. And we started out being founded by some Harvard graduates doing these highly scientific tests, because everyone at Harvard was used to building tests.
Initially, the company did not work at all. No one was interested in taking the test. Luckily, before we ran out of money, someone came up with a killer idea to do a more fun personality test called “What kind of dog are you?”.
Oops, sorry, but before you go on, Boris – just for some context, this is way, way back in the day… around 2000.
Yes, yes, way, way back!
We were one of the very, very first quiz sites out there.
Absolutely. And and the dog test helped us become I think the 28th biggest website in the US.
Not many people have heard the company name before but you could talk to people about the “What kind of dog are you?” test, and they would be able to tell you their results because they seen and taken that test. Sure, they’re proud to be a poodle or a golden retriever. A pit bull, maybe less so.
But yeah, they were really proud about the results.
So we wanted to take that idea and apply it not just to these fun tests that you see in your Facebook timeline – like “What city should you really live in?” or “What’s the best country for you to retire to?”. There are also more sophisticated and scientific tests that you can build using online personality tests.
And actually, from a marketing perspective, it’s not just only for light and fluffy topics like “Which Harry Potter character are you?” or things like that. Because you’re basically giving each user custom results at the end of the quiz, it’s actually really powerful for product recommendations.
You could use them for “What type of car should I buy?” or “What noise-cancelling headphones?”. Any topic will work – just ask a couple questions, then based on those results, you can then present the user with, “Hey, this is what we suggest.”
Absolutely. And the basis for all these tests is really what we call the OCEAN personality types. These are a set of very common personality traits that you can find in people. So OCEAN defines your openness level, your conscientiousness level, for example, are you introverted or extroverted? Then you tie these to the results.
Let me take the dog tests actually as an example because it works so well. If I tell you the dog name, you have a picture in your mind and we tie that back to people. So let’s say in that dog test, one of the results would be a basset hound. And if you you know, can picture a basset hound, you know, it’s probably the typical British dog. You know, the long ears…
Very low to the ground, as well.
So if you think about a basset in terms of personality, the bassets are laid-back, low maintenance… downtime is their favorite activity, they treasure moments when you don’t have to do anything.
Next – you could now counter that with an opposite a Pekingese (and I hope I’m pronouncing this correctly). Pekingese dogs, you know, they’re classy, fashionable, love to be admired. You know, they walk with their head up high. They’re well groomed dogs, so completely the opposite personality.
These are the personality types. And when you write these tests, you always start out with the personality results. The way we always did this, and I remember Mike and myself sitting in the room on a whiteboard, charting down personality results that we wanted, and then coming up with traits.
Another example would be – let’s take something more serious, we could do a test for a financial institution. So there could be some personality types like “What kind of saver are you?” A ‘savvy saver’ would be one – and another might be an ‘adventurous investor’.
And then you would try to come up with what would a ‘savvy saver’ actually be like in terms of personality? Here we picture that person, right? And we come up with ideas. So now if you do this exercise, you wouldn’t want to ask those boring old types of questions from ‘strongly agree’ to ‘strongly disagree’.
You could ask more situational storytelling-like questions to get the user’s responses. One on our favorites is always “Imagine you’re at a cocktail party, what do you do?” Now, in your mind, there is a cocktail party going on.
Now, tying it back to the dog test – “Would be you tend to comment on people’s attire?” If that would be very strong, you’re a Pekingese type dog. That’s what you focus on – what are people wearing and you comment on that?
But maybe maybe all you care about is enjoying the free food. So you’re laid back – you just want something comfy. Boom! You’re a basset and we would also tie that to the the German Shepherd.
Yes, definitely. And actually the dog example we’re using is kind of light hearted and fluffy. But hopefully you can see how you start at the result level first, no matter what the topic. And we recommend about four to five, maybe six different types of results. Because if you have too few, then when people take it and they share it with their friends, everyone’s going to end up with the same one.
However, you can also go too far on the opposite side. Now technically, with Riddle’s quiz maker, there’s no technical limitation to numbers of results – but I think the most I’ve seen a customer use is 65 results.
Yeah, that’s really crazy. It will be really hard to score that test.
Yes, to map out all those traits that we discussed and points to make sure the outcomes make sense.
But the other part which I thought was really interesting is that these linear scale questions – you know, strongly agree to strongly disagree.
That’s an unnatural question. I never get asked in a like in a real world. “Mike, what do you feel about this? Do you strongly agree or strongly disagree?”
However, my friends might ask me a question along the lines of “Hey, what are you like at a costume or a cocktail party? And that’s a more natural experience.”
The other advice is that the questions need to be subtle, so the user does not instantly see a question and know how to answer to get a certain result. So if you ask, “Hey, are you friendly or aloof?” then it’s obvious to the quiz taker… “Okay, well, that’s gonna be tied to this type of dog.”
You want it to be a step back. So cocktail party is a good example of soliciting info in a more clever, subtle way. You’re still getting the trait, but it’s just a little bit further removed from the personality results we mapped.
Definitely. So you could use the cocktail party and other situations for more serious tests. For example, you can find out if someone’s an extrovert or introvert by asking “When you’re at a cocktail party, are you the one owning the room and chatting with everyone?” In which case, you’re more extroverted. Or are you clinging to one person or hiding behind the bar? Voila, you’re an introvert.
So going back to our financial institution ‘serious’ example. If you wanted to find out an investor type, it might be important to know if someone’s an extrovert and more risk-prone or a risk-averse introvert.
Or you could ask a question like “If you won it as a prize, would you jump out of an airplane with a parachute?” Your answers could range from “I would do it in a heartbeat” to “I would try to find a friend I could give this to because I’m not crazy”. And these would tie to risk adverse and risk friendly?
Questions like these work a lot better because they’re situational. And the other thing why people love these tests, right? They actually work even better than quizzes because they tell you something about yourself.
Yeah, they’re the most viral. Of all our different types of quizzes and polls and surveys, personality tests are hands down the most viral, but also the hardest to create.
Not technically, technically is quite easy with our quiz creator. But mapping out those traits is quite cool.
And the actual quiz creation flow is interesting, you start at the personality results, come up with just the general types, then you map them. So you’re making these questions last.
This flips that 50:50 rule we’ve talked about so often before. Remember, you have to spend as much time writing really good results as the actual questions. So in that “What kind of dog are you?” test, we didn’t just say, “Oh, you’re a Pekingese” and call it a day.
The reason the dog test was so successful is that each of those had balanced and insightful results. They each started with a variant of “Oh, well, this is nice, because you’re this, this and this. But here are some things you might have be aware of.”
For example, for the Pekingese example, a result might read “Hey Pekingese, you love being the center of attention and your friends love it. But be careful – don’t you know overshadow your quieter friends. (And then you end on a positive.) But everyone loves a Pekingese – life’s never dull around you!
Because so we spent a lot of time actually writing that copy, because that’s what completes the loop.
Right? And because you’re never always right, when you create a personality test, try to create answers that give you some leeway. You should never say, you must be result X – instead, say “You’re most likely a Pekingese, because of X, Y, and Z. That’s really important, because then they work so much better.
True. And also it gives you a little bit of an out – just in case for some reason, the mapping are not quite right and the user says “Wait a sec, I’m not like that at all.”
If you are very decisive, and you’re wrong, you lose a bit of credibility.
The other thing which which I like to do – people love taking these over and over and over again to see what other options are out there.
With the dog test, we had people taking it who just wanted to see what combinations got them to a Golden Retriever, a Labrador, a sheep dog, and so on. So we actually have this feature in Riddle’s quiz maker where you can actually show them that.
If you turn on the extended results, your users will see “Okay, we think you’re Pekingese. But you’re also 17%, Labrador 23%, and 26% Golden Retriever.” So people love seeing all the other permutations, which is quite fun.
And maybe to end this… (if you made it all the way to this point in this podcast to this point, we love you for it) – there’s two things we’re going to give you for free actually.
First, Mike and I wrote a book on how to create the perfect personality test where we go into this topic in great detail. And we’ll give you a link to that book in the show notes. You can just download it for free – there’s there’s no registration or strings attached. We just want you to read it.
Second, we’re also going to make that famous “What kind of dog are you?” test available for free to everyone. Just sign up on Riddle.com, create an account, even a free trial account, and there’s going to be a template for the personality test that you can just copy.
And sure, that’s a good light-hearted example. But you can also just see how we mapped all the results, asked these questions, and did the scoring.
Then you can easily adapt it and apply it to your own topic and needs.
Yeah. And that dog test has will probably be one of the most complicated tests in terms of mapping you’ve seen. And we actually created that with a very good friend, a PhD psychologist in animal psychology, Dr. Russ King, who helped us write this test.
So thank you, Russ, for helping us write this awesome test that we can now make available to everyone.
So actually, Boris we’re going to keep this short – the challenge is that we get so excited because we are serious quiz geeks. We’ve been doing this for so long. Normally we keep these very short, this episode has run a little bit longer.
But we’re going to wrap this up with one question. So you’ve taken the dog quiz a number of times. What was your result?
I most often end up as golden retriever, but it’s also my favorite type of dog.
So I kind of try to answer the questions to be a golden retriever… I can’t help it! And because I wrote the test, I’m allowed to cheat a little bit.
Fair enough. Thank you so much for joining us for us. Thanks everyone for listening.
And we always wrap this up – if you have any questions about quizzes or quiz marketing, just pop over to riddle.com. Boris and I actually jump on all the support questions, and we’ll answer you within a minute or two tops. I think our average right now is about 67 seconds. So we’re fast.
Right? Thank you. Mike