Every Flush You Take
Silicon Valley is watching your water habits. That’s probably a good thing
--
The rain is battering away at a conference room window in downtown San Francisco as Peter Yolles, the founder of WaterSmart, tells me how his company is making a business out of convincing people to consume less water.
After three years of a historically unprecedented drought in California, I had come up with what I thought would be a great idea for a tech story. Silicon Valley innovations are supposed to be all about changing the world and improving people’s lives, right? But what, I wanted to know, was it doing about one of the world’s biggest looming problems: the challenge of supplying enough clean, fresh water for a thirsty, thirsty world?
All roads immediately led me to WaterSmart, a venture-backed startup that combines data analytics with behavioral science to persuade people to cut back on their water use. Scott Bryan, the director of Imagine H2O, a seed accelerator that focuses exclusively on companies operating in the world of water, told me that out of the dozens of new companies that had gone through Imagine H2O's program, WaterSmart was doing the best.
Historically, Bryan told me, venture capitalists have been reluctant to invest in water-related startups because the water business is an extremely capital-intensive industry that is both fragmented and heavily regulated. You don’t get big returns on investment on a Silicon Valley time frame by investing in infrastructure. But WaterSmart’s model, which focuses on changing human behavior, doesn’t touch any infrastructure.
“WaterSmart is a great story,” he says. “They have been able to demonstrate a five percent conservation savings with a piece of software. They’ve had real success with it. It is exciting.”
When you take a close look at the change-the-world rhetoric that floods out of Silicon Valley, it’s sometimes hard to see how that new chat app seamlessly integrating video and targeted advertisements is meaningfully improving our standard of living. What makes WaterSmart fascinating is how it applies the current state-of-the-art logistics of Silicon Valley — readily accessible venture capital, cheap cloud computing, and sophisticated data analytics — to tackle a social and environmental problem of the highest order.
WaterSmart’s solution is to suck gobs of usage data up into the cloud, crunch it and then target wasteful water consumers with specific recommendations on how they could do better. We worry, for plenty of very good reasons, about all the data Silicon Valley’s tech companies are collecting about every aspect of our lives. But if the trade-off is a more sustainable world, maybe this time it’s ok to go with the flow.
In the week before my interview with Yolles, more rain fell in San Francisco than in the entire year of 2013. Mindful of the downpour outside, practically my first question is whether WaterSmart’s success might actually be explained by the drought. The company started deploying its software almost exactly when the rains stopped. Isn’t it an easy sell to get people to conserve when the landscape is so parched? And won’t it be much harder for his company to keep getting traction when the rains return and the immediate pressure slackens?
Yolles is more than ready for my questions, because answering them goes right to the heart of how WaterSmart works. Contrary to what one might expect, he says, it’s very difficult to get people to cut back on their water consumption by appealing to their concern for their environment. Perhaps even more surprisingly, it turns out that appealing to thrift — savings on your utility bills — also doesn’t move the water meter needle.
So what does work? Shame!
“Research shows,” says Yolles, “that only one out of ten people are motivated to save money and only one out of ten are motivated to save the environment. But eight out of ten will do so to keep up with the Joneses.”
The observation, drawn from research in the field of social psychology, doesn’t provide the most uplifting insight into human nature. The most effective way to get you to cut back on your long showers and half-empty dishwasher loads is to lay a guilt trip on you by proving just how profligately you are wasting water in comparison to your neighbors? Yuck. But it seems to be true. WaterSmart’s data, culled from 35 municipal utility customers and over a million individual water consumers, (and confirmed by independent evaluations) indicates that a year after deployment of its “behavioral water efficiency solution,” the utilities recorded consistent decreases in water consumption of five percent or greater. For budget-strapped utilities in the parched West, the savings make for a convincing pitch. Last October, WaterSmart announced ten new utilities had signed on as clients.
WaterSmart is capitalizing on a classic big data play. The water business is incredibly fragmented — California alone has more than 500 municipal water utilities. A single utility, explains Yolles, doesn’t have access to enough data to construct the the kind of detailed user profiles that make accurate — and compelling — comparisons possible.
A five percent reduction may not sound like much, but the implications of a successful rollout of WaterSmart’s software across the tens of thousands of municipal water utilities that serve customers across the United States are huge. Globally, the demand for clean (and cheap) water keeps rising, propelled by population and economic growth. But supply is limited by fundamental resource constraints and the huge capital costs involved in building, and increasingly, maintaining water infrastructure. Every drop counts.
When I ask Yolles how he ended up making a career out of water, he goes back to his earliest memories. “My parents were avid sailors,” he recalls, “and I spent almost every weekend when I was a young child before I was five years old on a sailboat.”
After graduating from college, he launched a six-month sailing expedition across the Pacific. Somewhere between the Galapagos Island and French Polynesia, he had an epiphany.
“When you travel out in the middle of the ocean,” Yolles says, “you see the same three miles, this circle of water that moves with you constantly. Right there I decided that water was something that I’d be interested in making my profession.”
He followed through. With the possible exclusion of a dolphin or a coho salmon, it’s hard to imagine anyone who boasts a resume more soaking wet than Yolles.
In 1992, he began his career working for the Pacific Institute, a water issues think tank headquartered in Oakland, Calif. After stints at the Environmental Defense Fund and GE Capital’s Structured Finance Group analyzing water infrastructure, he spent three years at the Western Water Company buying and selling agricultural water rights, and then another seven years as the point person on water issues at the Nature Conservancy.
The idea that became WaterSmart crystallized when, after moving into a new home in Marin county in 2001, he decided he wanted to get some sense of how effective his newly installed low-flow toilets and fixtures were.
“I wanted to see if the actions I was taking were actually going to make a difference, so I went online to my utilities website to see if there was a way to track my water use. I was lucky, there was a link that said ‘consumption use history,’ so I clicked it, and it said ‘under construction, check back soon.’ Which I did — I checked back in 2009 and that page was still there, unchanged.”
The utility, the Marin Municipal Water District, is now a WaterSmart customer.
By 2009, Yolles was itching to do something more entrepreneurial in the water world than his work at the Nature Conservancy. He and his co-founder Rob Steiner didn’t have to look far for a promising business model. In 2007, a company called Opower had demonstrated that the peer-pressure, keep-up-with-the-Jones insights from social norms research could get people to cut back on their electricity consumption. What worked for electricity, figured Yolles, should also work for water.
Yolles saw a compelling opportunity in the tools available to him in the booming tech economy of Northern California. (The marketing boost delivered by the crushing drought that arrived just as his company got up and running was “all part of the plan,” he jokes.)
In contrast to the energy industry, however, the business of water has historically attracted little interest from the venture capital community. (In 2012, the venture capital industry poured 69 billion into renewable energy and just 1.5 billion into the water sector.)
There are many reasons for the reluctance. The business cycle is often measured in decades; building new water infrastructure is time-consuming and costly for the more than 500 utilities who control water in California. The market is also heavily regulated, and by law, most utilities are constrained from seeking a profit.
WaterSmart sidesteps most of these problems. It has no physical infrastructure, outside of what is required to maintain data-servers in the cloud. The basic pitch is attractive to any budget-constrained public institution; sign up with us, and we’ll lower your cost of doing business.
At first glance I was confused as to how this works. Water utilities generate revenue by charging for consumption; if you cut back on consumption, revenue obviously falls. So what’s in it for the utility?
Yolles explained that most utilities must purchase their own water supplies, and even if they own their water sources, coping with increased demand is an expensive proposition.
“By reducing demand we can help avoid the need to purchase new water supplies or develop new water supply sources,” Yolles says. “Using behavioral water efficiency we are able to provide water utilities with the same quantity of water at a much lower price.”
Yolles told me that demand reduction via WaterSmart costs between “250 and 500 dollars” per acre-foot of water. San Francisco, he noted, is willing to sell its water to other utilities for $1700 per acre-foot. Desalinated water produced by the brand new state-of-the-art desalination plant in Carlsbad, Calif. (just north of San Diego), costs a whopping $2200 per acre-foot.
The math is easy, says Yolles. Pay $250 per acre-foot to WaterSmart to reduce consumption or much much more to cope with higher demand.
“The cheapest water is the water we already have,” he says.
The best proof that WaterSmart works happens to come from data generated by my own water utility. The California Water Foundation conducted a year-long trial of WaterSmart’s service with EBMUD, a water utility that serves 650,000 customers in the East Bay, across from San Francisco.
In the EBMUD pilot, WaterSmart sent 10,000 utility customers a “Home Water Report” that used what are called “injunctive norms” to tell households how they were performing.
The injunctive norms took the form of emoticons. If usage is less than the 20th percentile, the report includes a smiley face and the message “Great!” Between the 20th and 55th percentile, a customer sees a neutral face and the message “Good.” Consumers above the 55th percentile got a worried face and an instruction to “Take Action.”
It sounds almost absurdly simplistic, but it works: “HWRs appear to be effective at delivering information on ways to use water efficiently that households can and, judging by the measured effects on daily water
use and program participation, do act upon,” said the report. What’s more, the report also found that households that received WaterSmart’s HWRs were more likely to engage actively with their utility in other ways, concluding that “efficiency programs can provide an effective conduit for channeling customers into other utility conservation programs.”
Behind those simple utility emoji lurks a level of data sophistication that simply wouldn’t have been possible ten or fifteen years ago.
WaterSmart combines data derived from water meters and surveys that households voluntarily fill out with a wealth of other information. “We have a sophisticated algorithm,” says WaterSmart marketing director Jeff Lipton. “We are looking at millions and millions of data points. We look at consumption data, and we compare that with external sources, like property records — how many bedrooms and bathrooms does your house have. We use satellite pictures to see how much outside irrigatable land you have.”
Lipton said WaterSmart even pulls in data from the National Oceanic and Atmospheric Administration to calculates the rate at which water typically evaporates in a given region. All the data are combined to create user profiles for the purpose of making the most accurate comparisons.
It’s the kind of number crunching “that gets increasingly accurate as our data set grows,” Lipton says.
In other words, the more utilities that sign up as customers, the more accurate the profiles that WaterSmart can create, and the more targeted their comparisons and recommendations for action can become — and the more locked in becomes WaterSmart’s control of this particular market niche.
The intimacy of the data collected by WaterSmart raises some obvious concerns. I noted to Yolles that in other businesses where similarly granular data about user behavior is collected, such as healthcare and the auto industry, there are serious concerns as to the ultimate downstream uses of such information. A credit reporting agency, for example, might be quite interested in yet another source of data on household habits.
“That has not been part of our business plan,” says Yolles. “Our mission really is trying to help water utilities and customers save water. In terms of data, privacy and security we do not share personal identifiable information with any third party, period.”
Of course, what happens to any particular database down the line after a startup goes public or is purchased by some larger entity is always an open question. One lesson to take from the story of WaterSmart is that there is simply no aspect of our existence, down to the number of times we flush the toilet, that will escape the appetite of the information matrix. So be forewarned.
But there’s another, more hopeful lesson. It is impossible to look at Yolles’s career and not be convinced that he is dead serious in pursuing his goal of sustainable water management — a task that is utterly crucial to human survival on this planet. Silicon Valley has developed a suite of tools that are usually deployed to get us to buy something that we don’t really need. It’s encouraging, in this case, to see them used for something that really matters.
