2

Lean Principles and Process

“Top of the heap … in homelessness!”

That’s what Monica Martinez said to herself after taking the job as CEO of Santa Cruz’s Homelessness Services Center (HSC). Santa Cruz County, her new home turf, held the distinction of having the highest homelessness rate in America. Twenty-five years into the center’s existence, Monica was determined to do something different. She was going to go lean.

For the past twenty-five years, HSC had followed the classic game plan: get an idea, write a plan, raise money (through charitable or governmental sources), and then carry out your plan. If you succeed at getting this far, then you take up what’s next: measure impacts, evaluate, raise more money (see Figure 2.1).

This Plan–Fund–Do approach had served HSC well over the previous two decades. The organization had grown from a scrappy, community-led effort to a well-established provider of up to 145 beds a night for homeless people, with an annual budget of almost $2 million. All that planning and funding and doing had brought immediate relief to some of those suffering on the streets (18 percent of them on any given night), but it had done nothing to solve the problem permanently.

Monica was struck by the similarities with efforts to innovate in the private sector, where over 75 percent of startups eventually failed. In the nonprofit sector, the metrics were more multidimensional than simply whether or not you turned a profit for owners and investors. But the bottom line was the same in many ways. Any startup, any new program eventually had to answer the question: “Does it work?”

By that measure, HSC’s efforts to end homelessness clearly had not. And there were twenty-five years of evidence to prove it.

The New Way: Lean Startups for Social Change

Information technology is the key factor profoundly transforming the way companies, nonprofits, and government agencies can drive change by closing the time (and sometimes the funding) lag between planning and doing. Cell phones, GPS, the Web, social networks, and many other transformative technologies allow social and business entrepreneurs to know almost instantly whether or not their actions, the steps they are taking to change the world, are working. This speed allows social entrepreneurs to bypass Plan–Fund–Do, instead disaggregating the process of innovation into smaller, quicker steps that can eventually grow far larger than ever before.

As steeped as we are in the Internet and cell phones, it’s easy to forget all the ways in which our ability to get information has become faster and easier. For Monica’s work alone, that means that many homeless people now have cell phones, or that her organization is notified by email whenever the county issues a voucher for housing.

Overall, this change in speed underlies the practice of the lean startup. For social change work, lean startup practice is guided by three core principles and by a process called customer development.

Figure 2.1 Plan–Fund–Do … Repeat

Principle 1: Fail Fast! (or, Everything Is a Hypothesis)

Instead of waiting for the plan … to get funded … then implemented, social entrepreneurs practicing lean start with educated guesses (or hypotheses) about how they expect the overall innovation to work. Then they test those guesses relentlessly, failing fast, but also speeding the way to development of solutions that actually do work.

Each element of how you intend to operate successfully is quickly mapped out, often on a tool called the Business Model Canvas. You lay out your core “guesses,” but you test only those critical to early validation of your initiative. The model is broken down into vital components that can be tested quickly when the time is right. You develop hypotheses on the key components of your innovation, including your value proposition, who you’re targeting, how much it will cost, and how much it can raise (covered in much more depth in chapter 4).

The focus is not on planning, but on relentless testing.

In Monica’s case, she and HSC simply started trying new ways to put homeless people into permanent homes. Their early hypotheses centered on the importance of community involvement, developing good will on the part of landlords toward the homeless, and new partnerships (such as one between the local Veterans Administration and the county housing agency).

The HSC team had set a goal of housing 180 people at the highest risk of dying on the streets within thirty months. After twelve months, they evaluated their hypotheses and weren’t happy with what they found. They had, in fact, housed 20 people, but their metrics showed that they would land far short of their goal and end up housing a total of 50 or fewer after thirty months of effort.

The lean startup methodology had emboldened Monica and her team to innovate, to measure, and to test. It also enabled them to learn. A year into the program they focused down to one hard number. In the language of lean startups, it was “the one metric that counts”: the length of time between when a homeless person gets a housing voucher and when he or she signs a lease.

They had failed fast. Rather than taking twenty-five years to understand the persistence of homelessness in Santa Cruz, the Homelessness Services Center had learned in just twelve months that their approach wasn’t fast enough, even though it was still doing some good. They had not let themselves drift into change that was important but, finally, insufficient. Rather, they had experimented, measured, and found themselves wanting. Then they pivoted.

In the process, the lean approach had forced them to constantly revisit the most basic question: What are we trying to do? The single metric they settled on not only drove unprecedented results, but fully embodied their innovation and their intention—getting the homeless into homes.

Fail fast means getting out of planning mode and into testing mode, eventually for every critical component of your model of change. Customer development is the process that embodies this principle and helps you determine which hypotheses to start with and which are the most critical for your new idea.

The Process: Customer Development

Customer development is all about testing the right hypotheses at the right time with the right data.

When Thomas Bukowski thought of a way to help really sick people in low-income countries, one of the first things he did was get out in the field. He flew from San Francisco to Nepal to see firsthand how patients enter the system he was hoping to improve. Watsi, his nonprofit, crowdsources funding to help critically ill individuals get the treatment they need. Thomas wanted to go lean. It was out in the clinics of rural Nepal that he began to understand what that really meant.

One of the shorthand ways lean entrepreneurs refer to customer development is “get out of the building” because the key to its implementation is that the data comes from real people (your targets, your funders, your partners) in the real world, well outside the conference room where you came up with your hypotheses.

Besides relentlessly driving you toward feedback from outside your organization, customer development lays out a process that stretches from the very first hypotheses about what you are creating to the questions of how to scale your innovation and, eventually, how to institutionalize it for long-term sustainability and growth. It’s a four-step process that helps you develop hypotheses appropriate to your stage of development and drives you to use real-world data (from customers!) to test them.

The discipline of customer development is the core framework for lean startups. We’ll be devoting several chapters to each of its four stages. From end to end, they are:

Discovery. The initial stage of customer development involves translating your initial ideas about the innovation into hypotheses, testing assumptions about the needs being filled and the behaviors of key “customers” with respect to those needs. The key output of this phase is the simplest possible implementation of the program, called the minimum viable product (MVP), that allows for in-depth hypothesis testing. The focus is directly on the service or program to discover whether it can achieve its goals in as small an experiment as possible.

For Watsi, the MVP was simple. It test whether money to pay for critical healthcare could be channeled effectively to deliver that treatment. Watsi’s founders didn’t build software to do this. They worked with partners to identify the needy and sent money directly collected from their families and friends. At minimal expense they rapidly proved that individuals from across the world could directly save a life dependent on an expensive medical procedure.

Validation. The second stage broadens the range of hypotheses tested to include the broader environment, what it takes to get the program noticed, placed in use, paid for. This is often the point of first failure, where some key hypotheses start to fail. In almost every case, lean startups learn so much in this phase that they perform what’s called a pivot. They learn lessons, redesign their approach, and turn to a new way of doing things.

The moment when Monica Martinez and the Homelessness Services Center realized that they weren’t going to make their numbers was a pivot moment. They revamped their MVP to focus less on building a network and more on driving down time lags in the system.

Creation. With a proven, simple product in hand and a granular understanding of how targets, partners, and funders react to the new program, you focus in the creation phase on what it takes to grow rapidly. You test a new set of hypotheses around the problems of scaling and replication. How will you reach the average program customer? What medium (online, word-of-mouth, direct mail …) will you use? Who will be the most effective messengers? What will be the range of reactions to different ways of reaching customers? In short, beyond the service or product itself, how can it be effectively delivered to the largest number of people at the least cost?

Watsi hasn’t fully answered this question yet, and the answer will undoubtedly be complex. Beyond delivering money for healthcare, Watsi must understand the best ways to attract donors to its crowdsourcing platform. What will its long-term relationship be with the philanthropic community that supports its infrastructure? Are individual donors willing to see some portion of their contributions go toward organizational overhead?

Institutionalization. The final phase of customer development is institutionalization—shifting from the exploratory mode that launches and scales innovation to an operational mode that embodies all the lessons learned. Many hypotheses remain to be tested, but they tend to focus more on internal processes and how to keep improving and scaling impact in a more institutional way.

Lean startup practices are relatively new in the social sector, and not many of its practitioners’ organizations have reached this phase yet. But as they mature, these organizations can draw on their counterparts in the private sector for inspiration, and we’ll lay out some real-world cases in chapter 12. In a true lean organization, this phase can require the most creativity to keep the organization focused on ongoing innovation and prevent it from slipping back into expensive, inefficient ways of doing business.

Despite all this structure, customer development fosters tremendous creativity. The key is that it allows you to postulate the wildest possible hypotheses and then gives you a rigorous structure in which to test them efficiently and effectively. It helps social change entrepreneurs focus on the right questions at the right time, while taking nothing away from their tremendous creativity. On the contrary, it gives them a structured way to aim high and move quickly.

The second principle of lean startups for social change is all about how to test their hypotheses … how to build your program and succeed in the fastest, cheapest way possible.

Principle 2: Agile Development

Worldreader started with a simple goal—to fill the empty shelves of schoolrooms across Ghana with books. Until very recently, literacy programs to achieve such a goal ran like classic, hierarchical bureaucracies, with long-term plans determined in distant capitals, flowing out into the field to splash up against the messy reality of communities where illiteracy is widespread. Without even realizing it, Worldreader turned centuries of Plan-Fund-Do literacy projects on their heads. David Risher, Worldreader’s founder, said, “We were agile without even knowing it.”

Principle 1 replaces the elaborate assumptions implicit in the old Plan–Fund–Do model with educated guesses to be tested quickly. Agile development (Principle 2) replaces the Plan–Fund–Do method itself with a new way of building things. Agile production methods offer fast, iterative ways of building change. At more advanced levels, they offer a blueprint for how to generate guesses about your project that have the best chance of being right.

Worldreader’s experience started with direct experience with the problems of illiteracy and the absence of books in Ghana’s classrooms. The few books that were there were wildly out of date and out of context (A History of Utah was a title on one school’s bookshelf). You couldn’t make this stuff up, but it was probably the residue of some elaborate previous plan to help educate Ghanaians.

One of the ways agile development works is by creating a story about users, often called use cases. From there, you work backward to scenarios and archetypes—fleshed-out ideas about the place of your innovation in its broader context as well as multidimensional portraits of the people you are trying to reach or serve. In Risher’s case, he returned from Ghana with a clear vision of an entire generation leapfrogging paper books with digital ones, of getting the right tools and the right content to build literacy across the globe.

What he did next was the opposite of Plan–Fund–Do. He did not write up a long, detailed plan, and he did not seek funding. Instead, he asked Amazon for some Kindle tablets and smuggled them into Ghana preloaded with kid-friendly books like Curious George. The kids took to the books immediately and started looking for more culturally specific books, seeking new downloads.

Agile development is an ingredient to lean startups but a diverse field of practice in its own right with many streams and approaches useful to the social sector. What those diverse approaches have in common is that they can all be boiled down to one simple cycle: Build–Measure–Learn. Those three words capture the iterative nature of agile development. They are the way we make the product or service we are offering and how we test our hypotheses about the best ways to make it effective.

Agile development is the methodology that complements the customer-centered data gathering in customer development. As the name implies, agile development is also very much about speed. It is the method that can allow lean startup practitioners in the social sector to find the path to success quickly. Alternately, it’s the method that helps you figure out why you’re failing quickly enough to do something about it, or to revamp your approach as soon as possible.

The main imperative of agile development is to implement your innovation in ways that allow the fastest looping through the Build–Measure–Learn cycle. Contrast Risher’s approach to that used by the famed One-Laptop-per-Child (OLPC) project. OLPC was an idea, born at the MIT Media Lab, to mass produce a great laptop for low-income children around the world. While its main goal was similar (basic literacy), OLPC took more than six years of design and building before producing a viable machine. But by then it had become clear that cell phones were already well on their way to penetrating the same markets with essentially the same capabilities as a simple, basic, relatively indestructible laptop. Almost fifteen years since OLPC was first conceived, they’ve reached over 2 million children. Those are great results, but Risher and Worldreader surpassed that number in two years.

Figure 2.2 The Agile Way: Build–Measure–Learn

How fast is fast? To give you a sense of what “fast” in the agile world can mean, there are companies in the software business that release dozens of versions of their product every day. They take a little extra time up front and build diagnostics into the product itself so they can measure what each change has achieved in terms of product uptake, customer satisfaction, and delivery of essential functions right to the customer.

Most organizations, including nonprofits and government agencies, can’t meet this standard right out of the box. But Build–Measure–Learn is not an abstract concept. Organizations can set goals for the iterative process itself. Worldreader’s first few experiments took months: collecting Kindle tablets, smuggling them into different countries, experimenting with different content and ebooks. By steadily focusing on the rate of iteration itself, Worldreader can now prototype and test entry into a new community or serving a new type of user on a weekly basis.

As important as Build–Measure–Learn are the outputs at each step in that iterative process:

•   The output of build is a minimum viable product (or MVP) based on feedback from customer development. It’s the smallest, leanest possible implementation of the original innovation that allows testing of the most basic guesses about how to make change happen.

•   The output of measure is data, used to evaluate the experiment that you ran with the MVP.

•   The output of learn is what lean practitioners call a pivot, a revision of the original MVP meant to retest the original hypothesis, to test a revised hypothesis, or, if the original hypothesis is spot on, to move the product or service to the next level in creating even more value.

In sum, agile development is the process by which you actually implement your innovation and iterate it until it’s explosively effective, or you give up. Agile is the toolkit for the experimentation that is key to lean startup.

To many in the social sector, the idea of trying multiple different approaches is scary. Innovation often has to happen in environments like jails or hospitals where there is already plenty of risk without adding an explicit focus on risky experimentation. And how are we supposed to fund multiple experiments in environments where people have barely enough to scrape by today? That’s where the third principle comes in.

Principle 3: Efficiency!

Lean = Efficient. Lean’s origins are in companies where you have to make big change with very little money. Which means you have to do everything else we’ve covered in this chapter—test hypotheses, engage customers, develop iteratively—as efficiently as possible, or else you’ll run out of money before you’ve discovered how to grow and sustain your innovation, and then you will go out of business.

But efficiency is even more important to lean startups in the social sector because there’s often more at stake when a social innovation fails than when a business one does. In 1991, I was director of environmental quality for the City of New York. The agency’s budget regularly exceeded $2 billion per year, and that same year the agency was fined $750,000 (less than 0.05 percent of our total budget) by the state for the pollution we’d imposed on the Greenpoint/Williamsburg community through delay in cleaning up a major sewage treatment plant.

Fines against the city were usually left with the city to be spent on vanity projects that staff at agencies picked out. In a major departure from this practice, we convinced the state to force us to spend the money for the benefit of the same community we had polluted. In effect, we would empower the community with its own capacity to understand environmental conditions and independently verify the city’s compliance with environmental laws and standards: we would conduct the first comprehensive, community-based, cross-pollutant environmental assessment in the United States.

We were pretty scared—because failure was not an option. At the end of the project, the community had to be better off, and the only other option on the table was to rebuild a billion-dollar sewage treatment plant from the ground up. Customer development hadn’t crystallized as a widespread practice yet, but that’s pretty much all we did. We met with community members, we made educated guesses about what they needed, we implemented them, and then we started that process all over again based on their feedback.

Despite the fact that we were one of the largest city agencies, we were aware from the beginning that we had to be super-efficient to pull this project off. The competing pressures of the project—diverse community demands, regulatory pressure from the state, budget constraints from City Hall, and the fact that several of the most egregious polluters were other state and city agencies as well as our own agency—meant that we had to show results despite limited time and money.

As nonprofit and government practitioners of lean startup iterate their way toward solutions, they face a number of pitfalls that a steady focus on efficiency can help them avoid. First, rather than being distracted by all the competing demands for innovation that arise in situations like the one I faced, the principle of efficiency forces teams to keep the discipline of building and testing to the smallest possible goal that will advance the innovation. Efficiency means making the leanest possible solution work, and the minimum viable product and customer development are tools for efficiently testing an innovation.

Second, as lean startup projects iterate and pivot based on what they are learning, they face budgetary and political/social constraints on how much they can really experiment. Keeping efficiency front and center kills these two critical birds with one stone.

Efficient spending on the smallest useful tests extends the financial runway for the innovation, giving change-makers more chances to innovate repeatedly on the limited budgets available for new initiatives in government and nonprofits. A focus on efficiency makes it easier to keep stakeholders like foundations, oversight committees, and the public in the loop on why you’re making the choices you are.

Time is the enemy of all great innovations, and efficiency buys you the time needed to pivot until you’ve found the engine of change you’re looking for. In the private sector/lean startup world, small companies use a tool called innovation accounting that tracks how much each experiment costs and seeks to drive down the very cost of experimentation itself. A truism that applies to governments and nonprofits as well is that the only real limit to making your vision come true is the number of pivots you have left.

One early finding in our work in Greenpoint was a potentially elevated level of stomach cancer. A local Hasidic rabbi in fact simply ordered his congregation not to procreate for fear of bringing children into its risk-laden environment. As I mentioned earlier, running out of pivots in Greenpoint/Williamsburg was not an option.

The Greenpoint/Williamsburg Environmental Benefits Project, as it would come to be called, eventually succeeded in addressing many community concerns and, more important, giving residents a sense of real control over their hazardous environment. It developed a comprehensive, community-based monitoring and assessment system that led to a number of further innovations in urban sustainability across that community and the country.

When you’ve got an impossibly big goal with impossibly small resources, efficiency in design, in experimentation, in spending, and in myriad other dimensions is absolutely essential to achieve success.

Summing Up

The lean startup is guided by three principles:

1. Fail fast. Move from Plan–Fund–Do to quick, educated guesses that generate data about your innovation’s effectiveness.

2. Agile development. Quickly Build-Measure-Learn, and do it all over again until you find the optimal product or service, or until you fail, fast.

3. Efficiency. Preserve financial, social, and political capital by spending the least amount for the most result.

The overall process is customer development—a way of testing your guesses with the right people at the right time for each stage of growth in your program. By and large, these principles and process are interchangeable with those used by lean startup practitioners in the private sector, and this chapter has examined them in some depth. Before we plunge into the practices of lean startups, in the next chapter we’ll take a look at key ways the social sector differs from the private sector, and the importance of those differences for lean practitioners in each.