When we set out to survey people about turning Just Evil Enough into a cohort-based live course series, we knew we had to be clever about it. Not just because we wanted to make sure we got the best results, but because you shouldn’t buy a book about suverting existing systems if the authors don’t do just that every chance they get.

We’re going to share some of the tactics we employed, even if it makes us feel a little dirty. Please don’t hate us.

Audience, constraints, actions

Whenever you’re planning a survey, you need to know who your audience is, what constraints you have, and what actions you’re hoping people will take.

  • Audience: We have two big target audiences: Startups who want a solid go-to-market strategy; and corporate marketers in challenger brands who need to stand out from the competition and put their competitors on their heels. Those audiences have overlapping, but different, needs and challenges.
  • Constraints: We needed responses relatively quickly as we were working against a deadline. We also needed to be consistent with our brand, from color scheme to tone, so it would reflect the personality we’re trying to cultivate: A little bit naughty, but clearly organized and thoughtful. We want our audience to feel like they’re in on a secret, and be a bit playful, and that only works if it’s clear that we’re making valuable, professional content.
  • Actions: We wanted a significant number of responses, both from our existing network and from people we didn’t know. We also wanted those respondents to share the survey with their own networks, and to feel smarter for doing so. And we wanted to understand how much people would pay.

Tactic: A reward that self-selects the audience

To encourage people to respond, we wanted to offer a reward. That’s tricky, because simply offering a cash prize would have given us many responses from people who weren’t our target audience. So we decided to offer a free keynote talk or workshop for their organization—something we usually charge a lot for. This helped respondents self-select, because our target audience would find the offer valuable.

Of course, this creates another problem. We’re choosing the winner from among the hundreds of people who responded, but the more respondents we have, the lower a person’s chance of winning. So why would they share it?

Tactic: In-group tribalism gets shares

Everyone wants to be on the winning team. So we created two nearly-identical surveys (more on that below) and labelled them Team Orange and Team Black. And we said we’d pick a winner from the survey with the most responses, which encouraged respondents to share the survey with their networks urging them to join the same team. We made it easy for them to do this with a click:

A tweet urging someone to vote for Team Orange

Tactic: A little help gets more help

In Chuck Pahlaniuk’s Choke, the protagonist runs a scam to pay for his mother’s care: He goes to fancy restaurants, pretends to choke on food until someone “saves” him. The wealthy Samaritan then feels obligated to help him, and sends him money to help with his life.

This is a common tactic of travelling salespeople: Ask for a glass of water, and your mark is more likely to help you in additional ways. So by asking people to help us with the survey, and then to help us by sharing it, we’re encouraging them to help us further by taking the course or buying the book.

You may wonder why we’re explaining this here, since it ruins the trick. But we’re not actually trying to be evil—and we figure that when you see these tactics explained, you’ll realize we have a lot to teach.

Tactic: Colors influence behaviour

Why black and orange? We’re glad you asked. They’re our color scheme for the site and brand, but there’s a method to that madness. We spoke with an expert on neuroscience and color theory (one of the many case studies we’ve been recording for the book) and it turns out orange is seen as active and provokes responses.

A comparison of responses showing 58.4% of people chose orange

As expected, Team Orange outstripped Team Black by a handy margin.

Tactic: Ranges help anchor pricing

Most people don’t know what to pay for a course. A technique called price anchoring allows you to plant a price range in your market’s mind, urging them to select from the middle of a range. In one study of taxicab payment screens, suggesting a range of tips from 15% to 25% increased tipping amounts by 38%.

We have a pretty good idea of what we want to charge already, based on course size and duration. But we never want to miss a chance to put some of these tactics into action, so we offered four pricing options:

A set of four options including $5,000, $3,000, $2,000 and $1,000

The average price our respondents suggested was $1,792.

Tactic: Always Be Testing

A subversive marketer is always looking for new angles and opportunities. So we made one subtle difference between our two forms. Respondents who chose black were asked, “What would you pay for a 6-week live course?” whereas their orange counterparts were asked, “what should we charge for a 6-week live course?”

Differently worded questions, where orange suggested $1,917 and black suggested $1,610

The responses varied significantly (though an astute analyst would need to test whether color choice influenced the response.) Asking what we should charge resulted in a $300 greater amount than asking what the respondents would pay:

Other details

There are plenty of other details in the survey:

  • We repeated the Call to Action before and after the detailed explanation on the landing page (so we’d get clicks from people who wanted to move quickly, but provide information for more skeptical or curious respondents.)
  • We mentioned other well-known instructors, in what’s known as a “top shelf” strategy—if you think highly of them, and we’re their peers, then you assume the same of us.

Hopefully you can put some of these tactics to work in your own surveys. We make no apologies for being subversive. But we promise to do it in the open. After all, we’re not actually evil—just evil enough.