by Matthew Syed
These tests answer two vital questions. The first is the fundamental one of, Will people buy our product? If the MVP sufficiently resembles the proposed final product, but none of the early adopters have any interest in it, then you can be pretty sure that the entire business plan is worth ripping up. You have saved a huge amount of time and money by failing early.
But if the MVP looks like a possible winner, you can now find out how it can be improved further. This is the second question answered by the lean start-up approach. You can see what features the consumers like and what they don’t like; you can see flaws in the concept and vary its assumptions as you develop toward the final product. In other words, you have hardwired the evolutionary process into the design of the business.
• • •
And this brings us back to Drew Houston. His problem, you’ll remember, was that he couldn’t raise the funds to get his file sharing idea off the ground. Investors were not confident his idea would get anywhere.
What’s worse, it was almost impossible to create a working prototype. After all, Houston’s basic pitch was that the file sharing product would only prove its value if it could seamlessly integrate multiple platforms and operating systems. To do that in even minimal form required a huge amount of work, based on deep knowledge of the various systems.
But Houston had an insight. He realized that the MVP doesn’t need to be a working prototype at all. All it has to do is mimic the essential features of the final product. Provided it is sufficiently representative it can demonstrate whether consumers really want to buy it and thus kick-start the process of trial and error.
So Houston created a video that showed how the product would work in practice. There was no software, no code, but he didn’t need these for his MVP. After all, how do you decide if you want a piece of software? You often look over the shoulder of someone who has got it, and is raving about it, and watch what it does. That is precisely what Houston did with his video.14
Eric Ries, the technology entrepreneur and author, picks up the story:
The video is banal, a simple three-minute demonstration of the technology as it is meant to work, but it was targeted at a community of early adopters. Drew narrates the video personally, and as he’s narrating, the viewer is watching his screen. As he describes the kinds of files he’d like to synchronize, the viewer can watch his mouse manipulate his computer. Of course, if you’re paying attention, you start to notice that the files he’s moving around are full of in-jokes and humorous references that were appreciated by this community of early adopters.15
The effects were breathtaking. “It drove hundreds of thousands of people to the website,” Houston has said. “Our beta waiting list went from 5,000 people to 75,000 people literally overnight. It totally blew us away.”16
Houston had demonstrated that people wanted the product. It enabled him to raise more capital and continue product development with confidence. But it also enabled him to interact with the early adopters, develop practical knowledge, and refine the product. That is the value of the lean start-up.
Nick Swinmurn, another technology entrepreneur, created a rather different MVP. He reckoned the world needed a website in order to purchase a stylish collection of shoes. He could have gone about this in the usual way: raising millions in capital, creating a vast inventory, and developing relationships with all the various manufacturers: i.e., designing the entire company from scratch from a blueprint. In other words, top-down.
Instead, he toured various shops and asked if he could take photos of their shoes. In return for allowing him to take the pictures and posting them online, he said he would come back and purchase the shoes at full price if customers registered their interest. By this process, Swinmurn was able to test the so-called value hypothesis: do customers actually want to buy shoes online? It turned out that they did.
But he discovered a host of other things, too. By interacting with real customers he learned things he could never have imagined in advance. He had to deal with returns, complaints, and taking online payment. “This is decidedly different from market research,” Ries writes. “If Swinmurn had relied on existing market research or conducted a survey, it could have asked what customers thought they wanted. By building a product instead, albeit a simple one, the company learned much more.”17
In 2009 Swinmurn sold his company, Zappos, to Amazon for $1.2 billion.
• • •
Steve Jobs is a man who is often held up for his vision. He wasn’t interested in feedback and iteration, he wanted to change the world. We will explore how big, creative leaps happen in chapter 10. But in the meantime it is worth noting that when it came to many of his strategic decisions, Jobs harnessed feedback in often powerful ways.
When he took Apple into retail in the early 2000s, for example, he didn’t buy a string of stores and try to make the whole thing fly instantly. Rather, he bought a warehouse and started to test his hunches and convictions, and those of his retail experts. The first approach bombed, as Jim Collins reveals in his book Great by Choice. “We were like, ‘Oh God, we’re screwed!’” Jobs said.
So along with Ron Johnson, his retail leader, he kept redesigning and testing. Eventually they opened two stores in Virginia and Los Angeles, enabling them to test some more. Only when they had learned from direct feedback and early failures did they roll out big, across the nation, with disciplined consistency.18
The lean start-up approach has many parallels in the modus operandi of innovative companies. In its early days, 3M, the technology conglomerate, relied on a team of product developers for new ideas. They would brainstorm, think deeply, and then, when they had developed completed products, they would show them to end users to see how they reacted. It seemed like a rational process—but it was too slow.
In the mid-1990s they transformed their approach by bringing early adopters into the design process itself. They asked them to try early prototypes, observed them as they used the products, noticed what they liked and what they didn’t. This enabled them to test their assumptions again and again.
3M then compared the two approaches. The results weren’t even close. As the author Peter Sims puts it: “A study published in 2002 found that using [the] active user strategy to identify and develop ideas generated an average of $146 million after five years, more than eight times higher than the average project developed using traditional, in-house 3M idea-generation methods.”19
Many other “failure-based” notions are finding their way into business. Agile scrum development and the fail-fast approach are just two of these. Some are doubtless more effective than others. All would benefit from further testing (systems devoted to trial and error themselves benefit from trial and error). None should be used in the wrong context.
But the key significance of this family of ideas, which have helped to develop many of the world’s most innovative products, is that they present a riposte to the historic presumption of top-down over bottom-up.
Drew Houston, the entrepreneur we started with in this section, has learned an important psychological lesson too. To leverage the power of failure, you have to be resilient and open. In other words, you have to have the right mindset as well as the right system. If you run away from mistakes, you won’t get anywhere. “It is a very grueling experience,” he said. “One day you are on top of the world . . . the next day there is a huge bug and the site is down and you are tearing your hair out . . . And guess what: that is still true today.”20
In 2014 Houston’s company was valued at just over $10 billion. It is called Dropbox.
V
There is a metaphor that neatly summarizes these insights. It comes from David Lane, professor at Henley Business School and a leading thinker on complexity.21 The problem today, he says, is that we operate with a ballistic model of success. The idea is that once you’ve identified a target (creating a new website, designing a new product, improvin
g a political outcome) you come up with a really clever strategy designed to hit the bull’s-eye.
You construct the perfect rifle. You create a model of how the bullet will be affected by wind and gravity. You do your math to get the strategy just right. Then you calibrate the elevation of the rifle, pull the trigger, and watch as the bullet sails toward the target.
This approach is flawed for two reasons. First, the real world contains greater complexity than just wind and gravity: there are endless variables and interdependencies. Take a policy as simple as reducing the dangers of smoking by cutting tar and nicotine in cigarettes. It sounds great in theory, particularly when used in conjunction with a clever marketing campaign. It looks like a ballistic strategy perfectly designed to hit an important public health target. But when this idea was implemented in practice, it failed. Smokers compensated for the lack of nicotine by smoking more cigarettes and taking longer and deeper drags. The net result was an increase in carcinogens and carbon monoxide.22 That is what happens in systems populated by human beings: there are unintended consequences. And this is why it is difficult to formulate an effective strategy from on high, via a blueprint.
The second problem is even more elemental. By the time you have designed the rifle, let alone pulled the trigger, the target will have moved. This is the problem of a rapidly changing world. Just look at how IT products are becoming obsolete even before they roll off the production line. This kind of rapid change is only likely to accelerate.
What to do? Professor Lane recommends an entirely different concept of success: the guided-missile approach. Sure, you want to design a great rifle, you want to point it at the target, and you want to come up with a decent model of how it will be affected by the known variables, such as the wind and gravity. But it is also vital to react to what happens after you pull the trigger.
As soon as the bullet leaves the muzzle, as soon as it comes into contact with the real world—this is when you start to discover the flaws in the blueprint. You find out that the wind is stronger than you anticipated, that it is raining, and that there are unknown variables, interacting with each other as well as with the bullet, which you couldn’t possibly have comprehended in advance.
The key is to adjust the flight of the bullet, to integrate this new information into the ongoing trajectory. Success is not just dependent on before-the-event reasoning, it is also about after-the-trigger adaptation. The more you can detect failure (i.e., deviation from the target), the more you can finesse the path of the bullet onto the right track. And this, of course, is the story of aviation, of biological evolution and well-functioning markets.
This reasoning illustrates the balance between top-down and bottom-up. If the original ballistic plan is hopeless, if the bullet just dribbles out of the muzzle, precision guidance is not going to help very much. But likewise, if you just rely on a ballistic plan, however sophisticated, you are going to hit thin air. It is by getting the balance right between top-down strategy and a rigorous adaptation process that you hit the target. It is fusing what we already know, and what we can still learn.
In the coming decades, Professor Lane argues, success will not just be about intelligence and talent. These things are important; but they should never overshadow the significance of identifying where one’s strategy is going wrong, and evolving.
Systems and organizations that foster the growth of knowledge of all kinds will dominate. This is the insight that the high-tech world has been gravitating toward and that much of the rest of the world, with only a few heroic exceptions, is studiously resisting.
Think about the ratio of Unilever again: 449 failures to create a single success. Has your company failed that often, and been honest enough to admit it? Has your school? Has your government department? If they haven’t, you are likely to be off target.
It is pointless getting upset about this. Clinging to cherished ideas because you are personally associated with them is tantamount to ossification. As the great British economist John Maynard Keynes put it: “When my information changes, I alter my conclusions. What do you do, sir?”
VI
To conclude this chapter, let us take one final example that reveals the dangers of trusting narrative above testing and learning. It is from the field of international development and a powerful case study because it reveals that the consequences of relying on top-down intuition can sometimes be measured in lost lives.
Specifically, let us take the scourge of AIDS and HIV in Africa. There are a number of alternative approaches to preventing and treating this disease that, on the face of it, seem highly plausible. All of them look like positive ways to alleviate a pressing (and often lethal) problem. But which is the most effective? What does top-down judgment tell you?
Option 1: surgical treatment for Kaposi’s sarcoma, an AIDS defining illness
Option 2: antiretroviral therapy to combat the virus in infected people
Option 3: prevention of transmission from mother to baby during pregnancy
Option 4: condom distribution to prevent general transmission
Option 5: education for high-risk groups like sex workers
They all sound pretty good, don’t they? You can imagine that each approach has its own charity with its own website, glossy material, testimonies from people who have personally benefited from the program, and promotional video. This is how most charities operate. And, on this basis, you would probably invest your money with the organization with the most convincing narrative. In the absence of data, narrative is the best we have.
But this is why we need to conduct tests, to challenge our hunches, and the narrative fallacies upon which they are often based. And when proper trials have been conducted, it turns out that these different programs, which all look so impressive, have vastly different outcomes. It is not just that some of the approaches are a couple of times better; or five times better; or even ten times. The best of the options listed above is 1,400 times as cost-effective as the worst option.23
On the graph below, the treatment for Kaposi’s sarcoma doesn’t even register.
It is for this reason that many of the most influential development campaigners argue that the most important issue when it comes to charitable giving is not just raising more money, but conducting tests, understanding what is working and what isn’t, and learning. Instead of trusting in narrative, we should be wielding the power of the evolutionary mechanism.
“Ignoring effectiveness does not mean losing 10% or 20% of the potential value that a health budget could have achieved, but can easily mean losing 99% or more,” Toby Ord, a philosopher at Oxford University, has said. “In practical terms, this can mean hundreds or thousands or millions of additional deaths due to a failure to prioritize. In non-life-saving contexts, it means thousands or millions of people with untreated disabling conditions.”24
The problem is not just that the donors don’t know the effectiveness of rival approaches; neither do many of the charities. The power of the narrative fallacy, the stories of the lives being saved, and the testimonies told by people who have benefited are as convincing to people running charities as to those donating to them. Indeed, why would you wish to collect data when you can meet and talk to those whose lives have been saved?
But given that there may be an alternative treatment that can save more lives, benefit more people—sometimes hundreds or even thousands more—our faith in the evidence of our own eyes is often insufficient. It is by testing that we gain access to the feedback that drives progress, and, in the case of charities, saves lives.
One of the ironies of charitable spending is that the one statistic many donors do tend to look at can actually undermine the pursuit of evidence. The so-called overhead ratio measures the amount of money spent on administration compared with the front line. Most donors are keen for charities to keep this ratio low: they want money to go to those who really need it rather than office s
taff.
But given that evidence-gathering counts as an administrative cost rather than treatment, this makes it even more difficult for charities to conduct tests. As Ord puts it: “You might think that organizations would know the most effective treatment. But often they don’t and one of the reasons for that is because they don’t do as much program evaluation as we would like because they’re trying to keep the overhead ratio low. Also, they just generally aren’t aware of these figures.”25
Ord has set up an organization that encourages people to give 10 percent of their lifetime income to charity, but only to those projects with a proven track record of success.26 “Our intuitions about what works are often wrong,” he says. “We have to test and learn if we are serious about saving lives and alleviating suffering.”
Chapter 8
Scared Straight?
I
On a cool morning in the spring of 1978, seventeen teenagers from New Jersey and New York were driven to Rahway State Prison, one of the most notorious detention centers in North America. As they walked up the gravel path to the forbidding set of buildings, the youngsters joked and giggled. They were cocky, had lots of swagger.
The kids—fourteen boys, three girls, of different ethnic groups, aged between fifteen and seventeen—had one thing in common: all had been in trouble with the law. Terence, a seventeen-year-old African American, had stolen cars. Lori, a pretty white sixteen-year-old with a wide smile and large earrings, was a thief and a drug dealer. Angelo, a teenager with unkempt hair and a wispy mustache, had robbed shops in his neighborhood.1
Nearly half of all serious crime in America was, at the time, committed by children between ten and seventeen. Arrests for burglary were reportedly 54 percent juvenile; those for car theft were 53 percent juvenile.2 Rape had been on the rise. These seventeen kids, still joking as they reached the gates of the prison, were not just an isolated group of delinquents, they were symbolic of a wider social problem facing the United States.