The Right It

by Alberto Savoia

  In the next few sections, you will learn a set of proven concepts and tools designed to de-fuzz, clarify, and sharpen your thinking. Let’s begin with the all-important and all-powerful Market Engagement Hypothesis.

  Market Engagement Hypothesis

  Remember the Success Equation from Part I?

  Right A × Right B × Right C × Right D × Right E, etc. = Success

  For an idea to succeed in the market, a number of key factors must line up just right. It’s not enough to have what you think is a great concept for a new restaurant; you must also hire competent kitchen staff and servers, run an effective marketing campaign to build some buzz, get an initial set of positive reviews, and hope that Gordon Ramsay does not open a competing restaurant across the street. But, as we’ve seen, competent execution, experience, and even good luck can’t help you if your idea is not The Right It. If your chef or staff turn out to be unreliable, you can replace them. If your first marketing campaign fizzles, you can run another campaign. But if your restaurant premise—the concept itself—is The Wrong It, what are you going to do? Change people’s minds? Good luck with that!

  When the market believes, rightly or wrongly, that a restaurant called Cheapo Sushi sounds like an open invitation for food poisoning and decides to not even give it a try, you are done. If the market does not want to engage with your idea, you can’t force it to. Let me put it as succinctly as I can: If there’s no market, there’s no way.

  But what is your market? And how do you define and determine engagement? You need to be 100 percent clear about these questions. Enter the Market Engagement Hypothesis, which I will abbreviate as MEH (a suitable acronym since, thanks to The Simpsons TV show, the exclamation “meh” has entered our lexicon as a way to express a lack of interest or enthusiasm—which, more often than not, is exactly how the market responds to most new ideas).

  The Market Engagement Hypothesis identifies your key belief or assumption about how the market will engage with your idea. Will they want to learn more about it, explore it, try it, adopt it, buy it? And if they adopt it, try it, or buy it, how are they going to use it and how often? Will they buy it again or recommend it to friends? In other words, the MEH articulates your vision of how the market will respond to and use your idea.

  If your MEH turns out to be wrong, then it’s possible that your vision is just a hallucination or wishful thinking. In that case, you better revisit it, tweak it, or move on to a different idea. But if your MEH is right, you have a fighting chance against the Law of Market Failure.

  Because the MEH is so important, it should not only be clear, but it should also be testable and expressed using numbers wherever possible. But let’s not get too ahead of ourselves. Let me begin by showing you what typical Market Engagement Hypotheses look like, and then I’ll show you how to make them better. Here are some examples that we can use as a starting point:

  Idea: Cheapo Sushi, a cheap sushi food truck, home of the 99¢ tuna roll

  MEH: If we make sushi as fast and as inexpensive as other fast food, many sushi lovers will choose it over burgers, tacos, or less healthy options

  * * *

  Idea: Webvan, online ordering and home delivery of groceries

  MEH: Given the option, lots of households will regularly choose to buy groceries online instead of going to the supermarket

  * * *

  Idea: A movie based on the Marvel cartoon character Howard the Duck

  MEH: People love fictional ducks (Donald Duck, Daffy Duck), so they will flock (sorry!) to see a live-action movie featuring Howard the Duck

  * * *

  Idea: Netflix (the company’s original DVD-based business model—before streaming)

  MEH: If we combine mail-based delivery of DVDs with a flat monthly rate and no late-return fees, a lot of people will sign up with us instead of renting from video stores

  You get the idea. The Market Engagement Hypothesis is a short sentence that encapsulates the basic premise of your idea along with how you expect the market to engage with it.

  I have to confess that while I was developing the MEH concept, I was terribly tempted to use the word hope or hallucination instead of hypothesis (as in Market Engagement Hope or Market Engagement Hallucination). Most of the time either one of those terms would be a more accurate description of how people ruminating on their idea in Thoughtland imagine the market will engage with their idea—a blend of hopeful thinking and hallucination:

  Everyone at the office loves my low-fat vegan kale cookies. They all tell me I should go into business to sell them. Oh, and a neighbor of mine who works at Whole Foods said that this is exactly the kind of baked goods their customers are asking for. She said I could easily sell them for $3 each! So I am going to quit my job, take a second mortgage on our house, invest in commercial-grade baking equipment, hire a few people to help me, and three months from now I’ll be rolling in kale-cookie dough!

  Add tangerine trees, marmalade skies, a girl with kaleidoscope eyes, and, well, let’s just say that when it comes to ideas for new products or businesses, most of us don’t need LSD to have hallucinations. Over the years, I’ve had several such hallucinations myself.

  The problem is that sometimes those hallucinations become reality. Sometimes your imagined market materializes just as you expected. And once in a while the market turns out to be much bigger and much hungrier for your new product than you or anyone else could have imagined. In other words, sometimes your Market Engagement Hypothesis turns out to be correct—your idea is The Right It. Of course, you will still have to execute competently, and many obstacles still stand between you and market success, but over the years I’ve learned to count on the following fact: If there’s a market, there’s a way.

  If there’s real demand for inexpensive sushi, one way or another Cheapo Sushi will find a supply of low-cost fish, so it can offer a tuna roll for 99¢. People can be very creative and will usually find a solution once they have hard evidence (not just belief or hope) of market demand for their product. Formulating the precise definition and finding proof of existence of that market demand are exactly what the Market Engagement Hypothesis is designed to help us do. It’s an important tool and a nonnegotiable first step in beating the Law of Market Failure.

  Now that you have a general idea of what a Market Engagement Hypothesis looks like, let’s learn how to turn it into one of those sharp tools I promised you. It all begins with numbers.

  Say It with Numbers

  Are you familiar with the saying “Not everything that counts can be counted, and not everything that can be counted counts”? It’s a good quote—words to live by. But this is also true: “Some things that count can be counted—and should be counted.”

  One of the most valuable habits I acquired by working at Google is to avoid vague terms and to use numbers whenever possible. If “data beats opinions,” then the best way to express that data is to say it with numbers. For example, instead of blurting out, “I believe that if we make the ‘Subscribe’ buttons a little wider, we’ll get a few more clicks on them,” a well-trained Google employee would transform “a little wider” and “a few more clicks” into specific quantities, turning that fuzzy opinion into a testable hypothesis:

  Fuzzy opinion: I believe that if we make our “Subscribe” buttons a little wider, we’ll get a few more clicks on them.

  Testable hypothesis: If we make our “Subscribe” buttons 20% wider, we will get at least 10% more subscribers.

  By saying it with numbers, a fuzzy belief becomes a clearly stated hypothesis that can be tested. In this case, an obvious experiment would be to split users into two groups: group A (original button size) and group B (a 20% wider button), and to compare the clicks between the two groups.

  Test results: Using a sample of 1,000 page views, we conducted an A/B test. The results indicate that when we increased the width of the “Subscribe” button by 20% (from 100 to 120 pixels), we got 14% more subscribers.

  If these initial r
esults hold up after a few more tests, the team will have compelling evidence—YODA, not opinions or speculations—that a bigger button will result in more subscriptions.

  Fuzzy thinking, along with opinions, is like catnip for the Beast of Failure—an open and irresistible invitation for trouble. Nothing removes fuzziness from your thinking like numbers; and the best part is that those numbers can be just rough estimates at first. In fact, it’s a mistake to use more precision than the situation allows. That’s why in the example above we used round numbers (e.g., 20%, 10%) in the initial hypothesis. At that point we were just making an educated guess, and it would have been premature and presumptuous (not to mention ridiculous) to use more precise numbers—that’s what experiments are for. For example, after several more rounds of testing we might learn that the optimal width for “Subscribe” buttons is 124 pixels (24% wider than the current size) and that such an increase in size will generate an average of 13.8% more clicks.

  Now that you know what I mean by “say it with numbers,” let’s apply this approach to our Market Engagement Hypothesis.

  XYZ Hypothesis

  The Market Engagement Hypothesis is a critical first step and an essential tool. But, like a pair of scissors or a knife, if our tool is not sharp enough, it won’t be very useful. The way we sharpen a Market Engagement Hypothesis is by rewriting it using another tool: the XYZ Hypothesis. I developed this tool out of frustration during one of my office-hours sessions at Stanford. Here’s what happened.

  I was trying to get a small team of engineering students to apply “say it with numbers” to the Market Engagement Hypothesis for their personal air-pollution monitor idea. The students kept coming up with fuzzy descriptions of what they thought their market was and how potential customers would engage with their product. Here’s an example of what the students were coming up with:

  Some people who live in very polluted cities would be interested in a reasonably priced device to help them monitor and avoid air pollution.

  How many people is “some people”? What cities qualify as “very polluted”? What does “would be interested” imply? What does “reasonably priced” mean?

  We were meeting on campus in a whiteboard-lined room, where a previous group of students had left a bunch of mathematical equations on the board. I looked at those equations and had an idea. I jumped off my chair, grabbed a marker, went to the whiteboard, and wrote out the following:

  At least X% of Y will Z.

  Then I explained: “X% is a specific percentage of your target market. Y is a clear description of your target market. Z is how you expect the market will engage with your idea. As you may recall from your high-school algebra, X, Y, and Z are the letters we use to represent unknown variables. And at this point that’s exactly where your idea stands—you are dealing with many unknown variables. But you can begin by making educated guesses about those unknown variables, running some simple experiments to test your initial hypothesis, and making adjustments as necessary.”

  Finally, the students smiled and nodded—I was speaking their language. After a few iterations the fuzziness was eliminated, and they had a respectable, testable, say-it-with-numbers hypothesis:

  At least 10% of people who live in cities with an AQI level greater than 100 will buy a $120 portable pollution sensor (where AQI stands for Air Quality Index, an objective measure of air pollution).

  Note that the initial values for X, Y, and Z were just starting points—best guesses based on the minimum market size the students believed they needed for their idea to be viable. Is 10% a good estimate of the market? Is greater than 100 the right AQI? Is $120 the right price? Probably not. These initial numbers may prove to be way off, but at least the students defined what “some people,” “very polluted,” “would be interested,” and “reasonably priced” meant to them, and they could test to see if the market agreed.

  In addition to having the virtue of being testable, the XYZ Hypothesis is a great tool for getting teams to make their implicit assumptions explicit. One student’s version of reasonably priced was $200, while another one thought that they could not possibly reach 10% of the market at that price and that the device would have to cost $80 to $100. The two students didn’t know that they had different ideas about pricing, but when forced to put a number to “reasonably priced,” the disagreement was unmasked. Which student’s price is right? We don’t know—perhaps neither. It’s quite possible that no significant market exists at either price point or at any price point. It’s quite possible that people are simply not interested—for whatever set of reasons—in a portable pollution detection device. Ultimately, the market will decide what “reasonably priced” means; but for the time being the students met halfway and compromised on an initial price of $120.

  The XYZ Hypothesis format proved to be a wonderful “de-fuzzer.” It replaced broad and imprecise terms (“some,” “very,” “reasonable”) with precise counterparts and the vague notion of “interested” with the specific action “buy” at the specific price of “$120.”

  Before XYZ Hypothesis After XYZ Hypothesis

  Some people At least 10% of people

  Very polluted cities Cities with AQI index > 100

  Interested Buy

  Reasonably priced $120

  After that first success, I suspected that the XYZ Hypothesis format would prove to be a valuable tool, so I asked one of the students to take a photo of the whiteboard to commemorate that Eureka! moment.

  And I am really glad to have that memento now, because my suspicion proved correct. The XYZ Hypothesis has become a permanent and valuable part of my toolkit and one of the first things I teach. In fact, if I only have a few minutes to help aspiring entrepreneurs or product managers, I use that time to explain XYZ Hypothesis and help them express their idea using it—and it always succeeds in clarifying their thinking and bringing to the surface any misunderstandings or disagreements among team members.

  Venture into the Unknown

  As I’ve already briefly mentioned, X, Y, and Z are the letters conventionally used in science and mathematics to represent unknown variables. The letter X, in particular, is often used (even in popular culture) to represent mysteries, elements that we don’t yet fully understand or whose existence we cannot prove: The X Factor, The X Files, Planet X.

  That which we don’t know or fully understand represents both a danger and an opportunity. This makes these three letters very appropriate for our task, because bringing a new product to market is akin to a journey into the unknown—a journey that could bring us great rewards or result in failure. It’s like entering a colossal dark cave that is both full of treasures and full of traps—home to the Beast of Failure and the trolls of Thoughtland.

  X, Y, and Z are also used to describe, measure, and graph things in three-dimensional space. In our case, the unknown three-dimensional XYZ space we are exploring and want to map and understand consists of:

  X: How big a slice, what percent, of our target market can we capture?

  Y: What is our target market?

  Z: How and exactly to what extent will the target market engage with our product?

  No sane explorer would venture into the unknown without the basic tools (compass, sextant, charting kit, etc.) needed to track his or her position and measure progress. The XYZ Hypothesis is the first tool in our market exploration toolkit. And it’s an essential one, because it gives us an objective way to measure and map our steps into that dark unknown that is the market.

  XYZ Hypothesis Examples

  Market Engagement Hypothesis and the XYZ Hypothesis are so important that I want to make sure you understand them well. So here are a few more examples of how to use the XYZ Hypothesis format to clarify a fuzzy Market Engagement Hypothesis. All examples are based on ideas that have come up during my lectures, coaching sessions, or brainstorming exercises. They are meant for illustrative purposes only. Don’t focus on the actual ideas or how silly they might appear to you, but rather
on how they are expressed in XYZ format. And, yes, I know that some of these ideas have actually been tried.

  Idea: An Uber-like service that picks up and delivers laundry

  Fuzzy MEH: Most people who use coin laundries dread the experience. Many of them would pay a few extra dollars to have their laundry picked up, washed, dried, and returned in a reasonable amount of time.

  XYZ Hypothesis: At least 10% of coin laundry users will pay $5 extra to have their laundry picked up and returned within twenty-four hours.

  * * *

  Idea: An ice cube–based air cooler for cars that don’t have air conditioning

  Fuzzy MEH: Drivers short on cash with broken or no air conditioning in their cars would buy a cheap ice cube–powered gizmo to help them cool their car.

  XYZ Hypothesis: At least 5% of people without air conditioning will buy a $20 air-cooling gizmo when the average temperature gets over 100 degrees.

  * * *

  Idea: Beer for dogs

  Fuzzy MEH: Many dog owners don’t like to drink alone; quite a few of them would buy a dog-safe beer, so their best friend can drink with them.

  XYZ Hypothesis: At least 15% of dog owners will add a six-pack of beer for dogs for $4 when they buy dog food.

  * * *

  Idea: Super Squirrel Collectors’ Edition book

  Fuzzy MEH: Fans of the Super Squirrel comic book series will go nuts (apologies) for a limited-edition, high-quality collection featuring their favorite rodent superhero.

  XYZ Hypothesis: At least 50% of the 220,000 subscribers to the Super Squirrel comic will buy the $100 Collectors’ Edition.

  These examples may be silly (I find that makes them more memorable), but the XYZ Hypothesis is a serious and powerful tool; and it becomes even more powerful when we combine it with our next thinking tool, hypozooming.