Implementing a bike-sharing service that has a real impact on city transportation usually means, among other things, getting the underlying system of docking stations right.

You’ll need a “dense network of stations across the coverage area,” advises The Bike-share Planning Guide, published by the Institute for Transportation & Development Policy. “The utility of dock-based bike-sharing systems depends on the presence of a fairly continuous network of stations,” agrees the Shared Mobility Toolkit, from the Shared-Use Mobility Center, “and building the network is a relatively capital- and labor-intensive task.” The process also requires careful planning to make sure the stations are arranged in the most effective locations—and that they don’t have negative side effects on their built environs.

But what if you could build a bike-share system with no stations at all, as some new enterprises in China are trying to do in a handful of major cities? One high-profile example is mobike, which launched last year and already has a fleet in the tens of thousands in Beijing. Its chief executive is a veteran of Uber’s operations in Shanghai, and it is backed by more than $100 million in investments from financial firms such as Sequoia Capital and Warburg Pincus.

Mobike’s approach relies heavily on its unique smartphone app and technology built into the bike’s patented design. Most significantly, the bikes don’t need a docking station or even a parking dock. Instead they are equipped with a special locking mechanism on the back wheel, meaning users can theoretically leave them almost anywhere except indoors and a few other locations. To locate an available bike, users consult the service’s app, which presents a map that uses GPS technology to point out the nearest available mobikes; you can reserve one through the app to make sure nobody else snags it first. The app also generates a QR code that’s used to unlock the cycle.

The company is still too new to be fully proven, and it faces competition—including another dock-free enterprise called ofo. But its stationless model may be as intriguing from a planning perspective as from a consumer’s point of view.

Zhi Liu has tracked the development of bike-share programs in China for years. Formerly with the World Bank, where he focused in part on urban transportation issues, Liu is now director of the China program at the Lincoln Institute of Land Policy and the Peking University–Lincoln Institute Center for Urban Development and Land Policy in Beijing. He notes that it’s important to understand the context in which these new businesses evolved.

China has a long history with cycling. But even for enthusiastic bike owners, rough and heavily trafficked roads make for a challenging long-distance commute in modern Chinese cities. So when bike-sharing schemes emerged in a few cities around 2008, as a complement to metro and bus options, the idea was quickly embraced. In 2011, the National Transport 12th Five Year Plan explicitly encouraged urban centers to develop bike-sharing as a useful addition to existing mass-transit systems.

“Planners and municipal governments now consider shared bikes a key component of public transport,” Liu explains, “because it helps solve the problem of the so-called ‘last mile.’” That is: You use public transport, and arrive at a station—and you still have another mile to reach your real destination.

Government programs in China didn’t face the same land-use challenges that might arise in a U.S. city, because urban land is state-owned. But other challenges persisted. By 2011, when a World Bank conference focused on domestic and international experiences with shared bikes, the major discussion was around “management and sustainability,” Liu says. “What business model makes sense?”

A mix of solutions emerged. In Hangzhuo, a government-led model involved setting up a state-owned company; today this is reportedly the largest bike-sharing system in the world. Other cities have experimented with various public/private hybrids, searching for a balance that would make bike-sharing cheap enough to attract users but profitable enough to cover costs. 

The latest wrinkle is businesses such as mobike and ofo, both of which also operate in other Chinese cities. These will clearly need to find that same economic equilibrium. But, perhaps because they’re both lavishly funded, each seems more focused for the moment on building ridership and acceptance.

Ofo overtly targets students, using lighter bikes with combination locks, university-centric distribution, and a very low deposit (13 yuan, or about $2). Mobike’s target is more likely to be an urban professional and/or cycling enthusiast. The deposit is 299 yuan (a little less than $50); rental is 1 yuan per half-hour. Its cycles are heavier but also more durable and distinct. “I do hear a lot of people talking about it,” says Hongye Fan, a Beijing-based consultant for the Asian Development Bank and investment manager for China Metro Corporation who has tracked bike-share programs. “It’s an innovative model in China and spreading very fast.”

Fan, previously an infrastructure finance and asset management consultant at The World Bank, points out some of the more intriguing side effects of the stationless models. Rolling out a major bike-sharing system can be, by necessity, a top-down process that doesn’t leave much room for flexibility once dock locations are built out—or, she notes, for “really thinking about and analyzing: What is the real demand from the citizens?”

Bike-sharing is a useful response to the last-mile problem, she continues, but “there is no universal last mile.” In fact, a station fixed in a spot that’s out of a particular user’s way could turn the last mile into the last mile and a half. An almost Uber- or Zipcar-like system that’s more overtly shaped by demand could avoid that.

And there are at least some experiments along similar lines elsewhere. A striking example is Copenhagen-based AirDonkey, essentially an app-based sharing platform that allows bike owners (including, notably, bike shops) to rent out their cycles to others. The startup hopes its model can work in other cities, even those where traditional share systems are in place.

Of course, such approaches involve other challenges and hurdles. Theft has been an issue for mobike, as it would surely be in almost any city in the world, although the company has said it’s a containable problem. Also, the demand-driven model could mean lots of bikes end up clustered in spots that are more popular as destinations than as starting points—meaning they’d have to be physically redistributed.

And, as Fan points out, planning would still play a crucial role in addressing problems that startups can’t—like designing and ensuring proper infrastructure, such as bike lanes, that makes bike riding safe and practical. But that’s true everywhere. Bike-share programs have proliferated wildly in recent years—Africa just launched its first, in Marrakech—and with an estimated 600 systems in place around the world, funding and implementation strategies vary. “We have not found any particular model that fits all cities,” Liu says.

Truth is, we probably never will find a universal solution. And that’s precisely why mobike and other new models—taking shape in China, the country with the most extensive bike-sharing systems anywhere—matter. Exploiting tech innovations in clever ways offers some compelling new potential routes to follow. Let’s see whether others take these ideas for a spin and where that leads. 

Rob Walker (robwalker.net) is a contributor to Design Observer and The New York Times.

Photograph: ofo

When we organize meetings in Latin America, we sometimes hire simultaneous translators to allow those of us with limited proficiency in Spanish to follow the conversation. These translators are a gifted bunch, capable of processing words, context, meaning, and nuance in nanoseconds. From time to time, they get tripped up in amusing ways. One commonly used word in our meetings is suelo. It comes up frequently when we discuss políticas de suelo, which translates as “land policies.” But suelo also translates as “soil,” and, as some translators would have it, we’ve participated in high-level discussions of “urban soil policies.” This left me reflecting on whether urbanists might learn something from agronomy.

Like many of our partners, the Lincoln Institute of Land Policy has ambitious goals. For example, we hope to use innovative land policy to mitigate or adapt to global climate change. We seek to promote financially resilient cities. We plan to help governments at all levels find the revenues needed to invest trillions of dollars annually in infrastructure. Our goals are embedded in the New Urban Agenda (NUA), an agreement signed by United Nations member states at Habitat III, UN Habitat’s recent Conference on Housing and Sustainable Urban Development. They also are aligned with the Sustainable Development Goals (SDGs) that replaced the Millennial Development Goals in 2015 to guide global efforts to achieve sustainable development that balances environmental, economic, and social objectives by 2030.

There are an estimated 650,000 jurisdictions on our planet. These range from around 30 megacities with populations over 10 million people; to 4,321 cities with populations exceeding 100,000; to more than a half-million places with fewer than 10,000 inhabitants. Implementing the NUA and achieving the SDGs will require reaching most of these places. How is it possible to change the path of development in so many locations?

Organizations trying to improve social, economic, or environmental outcomes at a global level typically work through theories of change—logic models that outline a process through which specific tactics and activities align to produce a desired outcome. A simplified theory of change might be: 1) find a successful social or policy innovation; 2) study it to understand why it succeeds; 3) export the innovation to new places; 4) measure its success; 5) repeat steps 3 and 4 until no longer necessary. 

Most theories of change include ways to scale successful interventions through replication and other means. But there are fundamental problems with this “franchising change” model. First, we are not very good at learning from success or even accounting for it. We can observe whether a project or program is successful, but we usually provide only untested hypothetical accounts for why it works. Often our hypotheses are wrong, and attempts at replication wither and die. In other cases, it is impossible to replicate key elements of a program. Thus, for example, the celebrated successes of the Harlem Children’s Zone have not been repeated elsewhere.  We have yet to see the scale or impact of the Champlain Housing Trust copied in other cities that face insurmountable affordable housing shortages. And although there is increasing interest from cities around the world, we have yet to see any that have successfully imported Sao Paulo’s practice of institutionalizing land value capture in its stock exchange.

Perhaps we fail to transplant these successes because we can’t clone the unique leaders who drove them.  Or maybe we can’t mobilize the kinds of resources that one can find in New York, Burlington, or Sao Paulo. Or perhaps it is simply much harder to replicate success than we think.

I’ve spent the last three decades trying to address global challenges like poverty, inequality, and climate change with interventions that could grow sufficiently to meet the scale of these problems. I believed in the promise of innovation—social, scientific, or policy-related. I, like many of my colleagues and contemporaries, believed that my job was to find a magical idea or practice that could spread virally, by replication, or through spontaneous combustion, whatever it took. I thought of myself as an explorer looking for a sturdy potato to bring back from the far reaches of the Andes to feed the teeming masses of Europe.

I’ve only recently come to understand how badly I misconceived my job. It is fairly easy to scour the globe for innovations and only a tad more difficult to construct a hypothetical account for their success. But it is really hard to transplant a novel policy, tool, or practice, and it can be costly to relocate creative new measures and watch them wither on foreign soil.

Looking back, it is not surprising that we were unable to scale social or policy innovations through replication. Each new approach unfolds in a complex social, political, and legal ecosystem. We reduce this complexity by guessing at the salient elements of each complicated context to account for success. It is difficult, if not impossible, to do controlled tests to confirm our hunches. So instead we use trial and error, uprooting successful projects, programs, or policies and planting them elsewhere, hoping that they will take root. And they rarely do. When replications fail, it is easy to attribute failure to a deficiency in the destination. But if we paid more attention to preparing the ground to receive new tools, practices, or policies, we might have more luck at replicating success.

This is where we can take a page from the agronomist’s playbook. Soil, too, is a complex ecosystem. It is composed of minerals, organic matter, and trace elements that offer plants sustenance. But the process through which different plants extract nutrients from the soil is a very complicated process.

It starts with the roots. In natural settings, the stems, leaves, and flowers of plants and their roots evolve to adjust to the complexity of the soil and the variability of climate. With the invention of agriculture, we interrupted this evolutionary process in order to cultivate non-native species in new environments. Through trial, error, and scientific inquiry, agronomists learned a lot about how to cultivate plants that are native to one place in new terrains. Thus, the potato, imported from the New World, became a staple in the Old World in the 18th century. But failure to account fully for the complexity of soil and environment generated some terrible unintended consequences, such as widespread blights that led to mass starvation in Ireland and Finland. 

Uprooting a vegetable and planting it elsewhere is a crude way to replicate success. Growers of certain crops have more sophisticated ways to overcome the joint challenges of soil and climate complexity. They do this by treating a plant as two systems—the root system that delivers sustenance from the soil and the fruit system, or scion, that produces the desired output. Vintners find successful local varieties of a plant and combine their root stock with the fruit stock of a different desired variety of the plant. Skilled practitioners help them to weave these two systems together. This job was celebrated by John Steinbeck in The Grapes of Wrath:

The men who graft the young trees, the little vines, are the cleverest of all, for theirs is a surgeon’s job, as tender and delicate; and these men must have surgeons’ hands and surgeons’ hearts to slit the bark, to place the grafts, to bind the wounds and cover them from the air. These are great men.

For example, a winery in Sonoma, California, that wants to produce wine using a Sangiovese varietal might import the fruit stock from Tuscany and graft it to the root stock of a Zinfandel vine that thrives in the local soil. The California vintners do not need to be soil scientists to replicate a successful Tuscan grape, but they do need to identify the vines that have successfully adapted to the complexities of the local soil and use their root systems to sustain and promote the growth of their chosen varietal. And they need skilled practitioners to graft the two parts of the plant together.

As we think more expansively about the practice of introducing new policies, tools, and approaches to the thousands of places that want help finding answers in land, we are learning a lot. We are learning about ways to prepare the ground to adopt new practices—understanding the “rules of the game” that define the local policy space, for example, and proposing revised rules to enable new policies. Or studying the local institutional ecosystem to identify all of the important stakeholders and inviting them to the table to help initiate new practices. We are learning that successful local people or organizations are the “root stock” that will sustain imported innovations and allow them to thrive. And we are learning that grafting an imported innovation onto this local root stock is a delicate task.

Many organizations focus on identifying and rewarding urban innovation—the magical interventions that help us overcome problems that result from our insistent efforts to urbanize the planet. At the Lincoln Institute, we are paying more attention to the process of replicating success. We will continue to document and share what we learn from transplanting innovation. Whether cities use land value capture to pay for infrastructure, create permanently affordable housing through community land trusts, or improve public schools with more resilient public finance systems buttressed by the property tax, each intervention will need to take root in local soil to succeed. We hope to be there to monitor and report on this success.