Looking back, Fausto Marcigot says he has been to about a half dozen tea parties in Mukuru. Mukuru is a large informal settlement in Nairobi, Kenya, home to roughly 750,000 people, and it is where all of PayGo Energy customers are based. Marcigot, PayGo’s engineering lead, was always on site as the company was getting its first gas customers connected, and the tea parties became a way to celebrate.

“They evolved organically,” he says. “Kenyans don’t like to waste energy, so at the end of a [stove] installation, after we checked everything, when we would teach the customers how to light the gas, they would put water on to boil and invite us and their neighbors over for tea.”

PayGo introduced gas stoves to Mukuru’s residents in 2016. The standard cooking fuels in Nairobi’s informal settlements are otherwise charcoal and other forms of biomass, and kerosene, because there is no piped infrastructure for in-home gas. Canisters filled with liquid petroleum gas (LPG), also known as propane or butane, are technically a third option, but they are both expensive and in short supply. Even if a family was able to pay the high upfront purchase price of a stove and canister, they would also need a large sum to periodically refill a gas cylinder. This leaves most low-income households with few fuel options for cooking.

“Here in Kenya, a lot of the population lives on a day-to-day basis. They often get paid on a daily or weekly basis, and so they can’t really afford a full cylinder of gas,” Marcigot explains. “What they can afford is a small daily payment.” Most households therefore buy charcoal or kerosene as they need it.

There are plenty of downsides to both forms of fuel. Charcoal is slow to heat up and produces a lot of smoke. Indoor air pollution in poor households is a major contributor to respiratory diseases and premature death. Kerosene, on the other hand, heats faster, but it is highly flammable and leaves cooked meals with a distinct aftertaste. Both types of fuel are affordable on a daily basis, but after all of the daily expenditures are tallied over the course of a month or year, the poor’s per-unit energy costs often far exceed their middle- and upper-income counterparts, who can afford to pay for fixed services or buy energy in bulk.

A large number of Kenyans get paid daily or weekly. They buy charcoal or kerosene as needed, but they pay premium rates.

The problem reaches far beyond Mukuru: 84 percent of Kenyans rely on solid fuels for cooking, according to the Global Alliance for Clean Cookstoves. Globally, 2.7 billion people do not have access to clean or modern cooking devices.

The team behind PayGo wanted to change that, in their own backyard at least. Their idea was to introduce LPG as a more affordable option than charcoal and kerosene, but which customers could still purchase on an as-needed basis.

Next level P.A.Y.G.

Pay-as-you-go services are not new in sub-Saharan Africa; in fact, they have become the standard for the provision of services like mobile communication and solar energy-based home power systems. PayGo is charting new territory with pay-as-you-go gas in Kenya, however.

To implement a pay-as-you-go gas system, Marcigot and his team developed a smart metering solution that connects to any standard gas cooking cylinder. The smart metering device consists of a GSM communications unit, a gas measurement system, and a shut-off valve, all of which are bundled into a pre-certified, explosion-proof casing. The metering device is affixed to gas cylinder with a tamper-proof lock and then hooked up to a standard two-burner gas stove, which PayGo buys from Hotpoint Electric Heating Company, a company co-owned by Whirpool and Haier. All of these elements are part of the full package that PayGo offers its customers and delivers directly to their homes.

“They make a small down payment of about US$30, and once everything’s installed, they just purchase as much or as little [gas] as they want,” Marcigot explains. Price-wise, customers who switched from charcoal to PayGo’s gas service can save 25 to 30 percent on their daily fuel costs.

On the software end, PayGo’s system is integrated with mobile money platform, M-Pesa, an almost ubiquitous digital payment and money transfer service throughout Kenya. All of PayGo’s potential customers can use M-Pesa with a basic mobile phone. PayGo collects and processes the M-Pesa payments with another service called Movapay, which also allows the company to monitor gas consumption data from its customers, and in turn, calculate how much gas is in each cylinder. “We monitor that on a daily basis, and based on what’s left in the cylinder, we arrange for a cylinder delivery,” Marcigot says.

“We replace the cylinder before customers run out of gas, so by constantly monitoring how much people are using, we can predict when they’re going to run out,” he adds. “We essentially create a piped-gas infrastructure without actually building the infrastructure.”

The PayGo system also has a number of backend apps for their sales team to enroll customers and chase potential leads. The apps also help the delivery team track where to transport the cylinders. PayGo is now working on building a mapping system to optimize delivery.

“We essentially created a piped-gas infrastructure without actually building the infrastructure.”

While PayGo’s model was influenced by the numerous start-ups that have built revenue-generating pay-as-you-go business models around solar technology, the PayGo team found that working with gas comes with a unique set of complications. “We’re dealing with a highly flammable utility product, and so there’s a lot more regulation around it, a lot more hurdles to overcome,” Marcigot says. “And then you have to match it with a delivery service.”

“It’s no good providing one tank of gas with a pay-as-you-go meter if you can’t then supply gas on a continuous basis,” he continues. “That’s a challenge that the solar companies don’t face, because once you [hook] up your panel, as long as the sun is shining, you can continue to provide the utility.”

Form and function

The PayGo team wanted to give the model a chance for gas all the same. To start, they built several smart meter prototypes to begin testing form, functionality, and experimenting with costs. They then reached out to U.S.-based design firm Catapult Design to help improve the user experience of their model. Catapult helped them zero in on a tick list of questions relevant to PayGo’s engagement with its customers, like: how do customers expect to communicate with PayGo? What information do they want to receive, and how do they want to receive it?

From there they narrowed into more specifics, related to the product’s functionality, like: should there be a power button to turn on the stove? Should there be a screen that displays gas usage on the device, or is it better to text customers with usage information? If they do include a screen, how and where should it be positioned?

“Our job was to help them answer some of those questions,” says Heather Fleming, Catapult’s CEO. “We relied a lot on the prototypes that [PayGo] had already built to communicate some of these concepts with their prospective customers—households that they’d been consulting with who were helping to inform them on what the product and service might be.”

Fleming’s team spent a week in Mukuru with Marcigot and PayGo’s head of product development, Mike Hahn, speaking with community members. They worked with a local translator to interview 20 or so residents, only to discover the difficulty of explaining the concept of a product and service that did not yet exist. “Their target customer really didn’t have a mental model for this service that they were offering,” Fleming says. “[Mukuru’s residents] almost exclusively use charcoal and kerosene. Both the pay-as-you-go piece and the use of an LPG stove were foreign to most of PayGo’s prospective customers.”

“When there isn’t a mental model for something, intuition—or building intuitive systems—is very important,” Fleming adds. “Our job is to try to understand what people want to know from this device and what is the best way to communicate it to them.”

To achieve this, the team relied on props and activities to engage potential customers in hopes of understanding their preferences. This way of communicating proved more effective than simply asking theoretical questions, like: do you want to receive a message via SMS, or do you want it to come from a device that is attached to your home? Catapult was able to synthesize the insights gathered from Mukuru’s residents and put together a set of design recommendations for PayGo’s product team. The team, with support from Catapult’s designers, revamped their prototypes and began preliminary user testing in early 2016.

“We built around five different prototypes to understand what features people wanted on the device and how they wanted to interact with it,” Marcigot says. “We found that simplicity was king.” Too many lights and icons were confusing. Users also did not associate colored lights or buttons to any particular meaning—like green for “on” or red for “off” or to signify danger. There were some clear preferences, however, Marcigot says: “Everyone wanted a child lock and to see their [gas] balance on a numerical display.”

The user feedback served a double purpose: it helped PayGo further hone its product design, and it also gave potential customers a clearer idea of the product and service PayGo intended to offer, drumming up interest in the community.

Inevitable obstacles

PayGo officially launched its service in October 2016 with a 10-household pilot. At the time of writing, its customer base has grown to about 120 households, and the team is hoping to expand to 300 households by the end of 2017. Marcigot says that they have had no shortage of interest from existing customers’ friends and neighbors. There are several key factors impacting PayGo’s pace of growth, however.

First, is the company’s ability to get the technology and hardware it needs into Kenya. “Working with multiple partners all over the world and coordinating those efforts is a bit tricky,” Marcigot says. “Developing new technology always takes much longer than expected; those delays seem to be compounded when you have difficulty importing materials or making changes without access to the right sort of tools and equipment.”

Second, the political tide in Kenya disrupted the overall business environment for several months this summer. The run up to Kenya’s presidential election caused apprehension among the general population, which was worried about a repeat of the post-election violence from 2007 and 2008. Marcigot says that this atmosphere affected sales agents’ work in the communities and created a backlog of new connection requests that PayGo is just now catching up to.

PayGo’s target customers didn’t have a mental model for PayGo's product or service. The startup discovered the difficulty of explaining a concept that did not yet exist.

Third, because pay-as-you-go gas has never been implemented before in Kenya, potential customers and gas companies alike have expressed confusion or skepticism over some elements of PayGo’s service model. For example, the startup is still trying to figure out how to succinctly communicate why it swaps out gas cylinders before the tanks are completely empty. “The concept of someone coming to change your cylinder before you run out is really difficult to explain—not just in Kenya but to anyone in the world,” Marcigot says. “We’ve had to explain that we do this because we’re running a sort of uninterrupted supply of gas and that [customers] don’t need to wait to finish the entire cylinder.”

Finally, the team has had some work to do to alleviate people’s fear of using gas. “A lot of people are actually excited and want to have gas, but it takes some work to get someone comfortable with using it,” Marcigot says. PayGo always begins with potential customers by conducting a safety inspection of the home, to ensure there is enough space to install the stove and canister, to make sure the surface is supportive enough, and to clear any flammable fabrics or materials out of the immediate area. When it completes a system installation, the team provides a fire blanket and teaches customers how to use it.

“Once they are comfortable, potential customers are often ready for the service right away,” Marcigot adds.

Partnerships and productivity

PayGo is limiting the number of systems it installs to 300 households while its commercial meter—the one Catapult helped to design—is manufactured. In April, the startup raised more than $1.4 million for its first round of manufacturing and to support its 300-unit rollout. Three hundred installations are what PayGo feels it can reasonably service with its current funding and staff of 18, but that it represents a large enough sample size to prove the scalability of its model to existing gas utilities in Kenya. “We don’t want to scale beyond 300 [households] without really understanding the inner workings of everything and making sure that the entire service is reliable,” Marcigot says. PayGo hopes to work in partnership with utilities as it expands in the future, because it would enable the company to reach more people more quickly, Marcigot explains.

“Little mishaps in connectivity are part of developing a system. Looking back, it’s nice that we’re at a point now where our customers are unhappy if there’s no gas.”

Creating solutions to bring basic resources to low-income communities is something that many social startups are trying to do, from energy to sanitation to healthcare. “It can be difficult to see how they integrate into a wider industry ecosystem,” Marcigot says. Trying to solve a resource access problem without building pathways into a broader industry is difficult to achieve sustainably, he adds. “You need large amounts of capital to address millions of people.”

As PayGo ramps up its service, the team says that one of its biggest drivers for success is its existing customers. Its 100 or so customers support the adoption of PayGo’s technology in their communities and are helping to prove that gas as a service is viable for low-income, un-piped households. Many come to local cooking shows that the startup hosts to demonstrate its technology; the users participate in the presentations to explain to friends and neighbors how the system works. Like the tea parties, it was something that evolved organically as PayGo has been setting up its business, Marcigot says.

“We never asked them to come, but they wanted to show people what they had in their own homes and tell us about the meals they cooked,” Marcigot says. The cookstoves seem to be a point of pride for many of them, he adds. “It’s really nice and positive to see.”

Enthusiasm for PayGo’s technology is part of the company’s evolving relationship with its community of users, beginning from the time PayGo started testing its concept to now offering its service and recruiting new customers. That has been an important anchor for a startup that is building a service model from scratch.

“We’ve let our customers down on a number of occasions,” Marcigot says. For instance, there have been times when customers have experienced disrupted gas or data connectivity. “Sometimes the systems haven’t worked, or we misread or misinterpreted the data, and then someone wouldn’t have gas for a couple of days,” he explains. Other times, there have been problems with processing payments, where a customer would top up with credit, but the credit would not immediately get logged into PayGo’s system. In those cases, the users would have to wait for the payment to go through before the gas valve on their home device would open again.

“Little things like that are part of developing a system, but when people are accustomed to the service, you want to continue to offer them that service no matter what’s going on internally,” Marcigot says. He adds that their customers’ frustration during those times is a reflection of how PayGo is supporting the transition from costly, inefficient, and rudimentary cooking fuels to a utility service model. “When I look back now, it’s nice that we’re at this point now where our customers are unhappy when there’s no gas, which they should be.”

There are other signs of the potential impact PayGo’s service can have on low-income urban households. Marcigot says he has seen some customers already embracing gas for entrepreneurial and other productive uses. For example, one woman started a small business selling mendazi—traditional Swahili doughnuts. “She’ll get up every day at 6 a.m. and cook her batch of mendazi on the stove,” he says. “[She] says that she’s saving quite a bit of money compared to charcoal that she was using before. In the last four months, she’s essentially paid off her stove in full using the revenue she creates [from selling the doughnuts].”

Marcigot recalls hearing of a saying in Kenya that reflects how deeply a service like PayGo’s could impact the communities it serves—and not simply in terms of lifestyle or entrepreneurial opportunity, but down to a cultural level. “The saying was, ‘A jiko only has one burner’,” he says. (Jiko means stove in Swahili.) The saying seems to be a lesson in patience: that one can only do one thing at time. It is a saying that no longer applies to PayGo’s customers, however.

Marcigot adds, “Our stoves have two burners.”