Gehm’s Corollary: Sell Magic

Alistair Croll
7 min readJun 15, 2018

Science fiction author Arthur C. Clarke proposed three laws, the most famous of which is that “any sufficiently advanced technology is indistinguishable from magic.” In 1991, Barry Gehm proposed a corollary to this law: “Any technology distinguishable from magic is insufficiently advanced.”

This is a critical flinch test for founders. If your product doesn’t seem like magic to your target market, something’s got to change. A magical product means the difference between a long, costly sales cycle and a market that just pulls what you make out of your hands.

Choosing the right market

In a perfect world, a founder would first look for a target market they can reach profitably; then identify an unmet need that market has; and finally satisfy that need with something that feels like magic. That’s how it’s supposed to work.

Unfortunately, startups seldom start by identifying a need through customer development. They spend much more time adjusting their product to a market they’re in love with than they do finding a market that loves what they already have.

There are two clear reasons for this:

  • The startup’s founders have complete control over what they make; and talking to prospective customers is messy, uncomfortable, and fraught with rejection. So they do what’s easier and more fun.
  • Checking market demand forces the founders to confront their confirmation bias, and find out they aren’t the unique and special snowflake their mother says they are, which is never fun.

We can rationalize anything

In The Righteous Mind, Jonathan Haidt talks about his research into Dumbfounding Questions. These are questions we react to morally, and only afterwards think about rationally. In one example, he describes a widow who uses the flag from her husband’s coffin to clean up a mess. Offended, one respondent quickly says it’s morally wrong—but then struggles to explain why. Eventually, she argues that “the flag might clog up a toilet when she disposes of it.”

Moral reasoning happens from passions and convictions, but isn’t based in fact. Startup founders need a healthy dose of such delusion; after all, they’re defying common wisdom in their search for something new and unproven. But when this delusion spills over into product development and market research, it’s deadly.

As an advisor and mentor to accelerators, I’ve lost count of the number of startup pitches that present a technology, and then twist themselves into logical pretzels trying to explain why a market needs what they’ve already resolved to make.

We can rationalize everything.

Startups aren’t about fun

I mean, startups can be fun. But your actual job is to unearth a sustainable, repeatable business plan and eliminate risk from it. It follows, then, that if the biggest risk is “nobody will care,” the most important thing to check is whether you can reach a market that does. Being irrational doesn’t help this cause.

Rather than making something they can sell, many founders try to sell what they’ve already made.

If this is you—and it probably is—you’re not alone. I was one of these founders. You can still succeed. But you have to find a way to make what you sell magical.

Selling to magicians sucks

Here’s a somewhat painful story from my own past to illustrate the point. Fair warning: This is going to get a bit nerdy unless you have a background in network infrastructure and web performance management.

Coradiant, a company I co-founded in 2000, made powerful web performance management tools that measured user experience on websites.

We sold these products to the network operations teams responsible for uptime and operations. The tools were amazing: Where web analytics would show you what people did on your website, we’d show you if they could do it.

That’s some serious detail, circa 2003.

But selling it was hard, because we had ignored Gehm’s Corollary. We picked a market that could distinguish it from magic. What we made didn’t look magical to the operations team, which was filled with reasonably technical people. Many felt they could build this sort of thing themselves, or that they could solve the problem some other way. They understood the problem, but they also understood how a solution might work, and that slowed down the sales process significantly.

Ultimately, we did two things that helped us overcome the delay, and eventually build the business and exit.

  1. We found someone who did think it was magic;
  2. We found a way to do something magical for the current target market.

Finding a market that thinks it’s magic

While operations teams understood how we did what we did, marketers didn’t. We could show them where they were losing hundreds of thousands of dollars. And we could show them who was suffering right now. We added features that looked up the source of a visit and guessed who the visitor was, then exported it to a map. And we made it easy to look up a user by name, or locate their IP address on a map, and to see what happened.

Every page and object in Brenda’s session, down to the very last detail, as it happens on the site.

To marketers, this was unbelievable. They had no clue how they could see what was happening to an important customer, a new campaign, or an irate user, the moment it happened. They could find out what the purchase page was like for actual users, within seconds of a product launch.


Do something magical for the current market

We weren’t magic to network operators. We needed to find a way to give them some kind of magical superpower for them to like us.

In a tech company, network operations staff are responsible for the health of the website. But in many organizations, the things that break their systems aren’t their fault. Marketing teams install new apps, or introduce content errors, or sign partnerships without warning. This makes operations teams furious.

To make matters worse, because marketing controls what’s on the sites, if operations wants to monitor something, it often needs marketing’s permission to do so.

(Think about what you have to go through to install web analytics on a website; you need to change the content on the pages. Remember that this is 2004. Now imagine you have to ask someone who doesn’t really understand the Internet for permission to do so. This makes you Inordinately Angry. Today, the field of Devops is an attempt to resolve this tension, and the best Internet companies embrace continuous deployment and better alignment of marketing, development, and operations. But 14 years ago, that didn’t happen much.)

Operations didn’t think what we did was magic. But when we showed them you could get complete visibility into everything marketing had broken or changed in 15 minutes just by plugging in a couple of cables — and you didn’t need the marketing department’s permission to do any of that— well, that was incredible.

For operations, political advantage was magical.

Market maturity and the ability to distinguish

Different audiences have different ability to distinguish magic from technology. A software company probably doesn’t think wireless VR is magical; a realtor still does. As a product category matures, moving from early adopter to laggard, the ability to distinguish changes as well.

What’s obvious to one segment is mysterious to another.

You may be able to make something relatively mundane seem magical by targeting the right market, or a particular subset of the market according to maturity.

Chase the magic

When you’re picking your target market, remember Gehm’s Corollary. Is the market able to understand how you do what you do, or is it magical? If they can distinguish it from magic, it may be insufficiently advanced. And to fix that, you can alter the product—or find a more easily amazed target audience.

(One more thing…)

There are plenty of other criteria to consider when testing markets, of course. They’re covered in most marketing textbooks, and they include:

  • Shared need: Do those in the target market have the same problems and values?
  • Segmentable: Is the target market easily identified and distinguished from other segments?
  • Profitably reachable: Can you get the market’s attention in a way that won’t break the bank? Billionaires are a great target market that’s hard to reach.
  • Big or fast-growing: Is the demand large enough to fuel your ambitions for growth — or growing fast enough that you can grow with it?
  • Loyalty: If you’re looking for recurring revenue, loyalty matters a lot. If you’re doing one-time sales, it’s less important.
  • Social: Does your market share things? Will it amplify your message? A gregarious, vocal market is worth spending more to reach.

Just remember to check whether your target market can distinguish what you make from magic.



Alistair Croll

Writer, speaker, accelerant. Intersection of tech & society. Strata, Startupfest, Bitnorth, FWD50. Lean Analytics, Tilt the Windmill, HBS, Just Evil Enough.