Silicon Savannah: Expanding Access to Clean Cooking Fuel


A significant amount of the world’s population — over 2 billion households — cooks with dirty fuels. For most, there simply isn’t access to clean cooking fuel.

In Kenya, Nick Quintong (CEO and Co-Founder) and his team at PayGo Energy have designed an IoT solution that expands access to clean cooking.

In this episode, Nick shares:

- Why startups are abundant in Nairobi

- How a piece of smart hardware brings clean cooking fuel to Kenya

- The journey of developing new IoT technology

- Advice for negotiating the right metrics for success

Never miss an episode of Over The Air by subscribing wherever you listen to podcasts.

Listening on a desktop & can’t see the links? Just search for Over the Air in your favorite podcast player.

Currently significant amount of theworld's populations cooking with dirty fields that they can bine some smallamounts h. They can afpord every day. Our device allows in Te bygas in thesame passion so that they can afford it and Guass as much cleaner than cookingwith charcolar Karosine in the household. You are listening to over the Air Iotconnected devices and the journey brought to you by very in each episode. We have sharpunfiltered conversations with executives about their IOT journeys,the mistakes they made the lessons they learned and what they wish they'd knownwhen they started e everybody welcome back to over theAir Iot connected devices and the journey today we're going to be talkingabout energy, Iot and Africa, with co founder as CEO, Pago Energy, NickQuintong. Nick thanks for going on the show man, thanks for having your en sonick. For those who don't know, tell us a little bit about Pago Energy and theproblem you guys are trying to solve, and for now thirty thusano foot viewis,fine thanks, frying yeah, so pigo energy ourm mission is to unlock cleanennergy, the next billion and so by the next billing we meet have the billionhouseholds that are currently alack actess to clean cooking fuel today, andso our technology is a smart meter that locks in Eddi gas cylinder and it turnsit into. It pays yo good advice so that customers can actually buy gas andsmall amounts, currently significant amount of the world's populationscooking with dirty fields that they can bine some small amounts they can affordevery day our device allows in the bygas in the same fashion so that theycan afford it and gass as much cleaner than cooking with Charcola parasine inthe household. I guess one thing: That's kind of interesting to me islike your background personally you're not from Africa, originally believeyou're raising the United States spent some time at ge kind of a like a don'tKno, more traditional career path. Can you talk a little bit about like howhow a guy came from that background- and you know, is now foundingcofounding a tech company in Kenya like what does that path? Look like yeah. Itwas kind of kind of a random bat, so yeah as thatGE great company, Great, receive greattraining there. Originally I'm an Oregon boy. I grew up in band if I'm Maway into the corporate world and yeah around year, four five, I startedgetting this kind of creeping existential anx about what am I doingin the world. I think it's something that's been plaguing our generation, and so I wanted to try todo something a little bit closer to the impact side so coming into like myfifth year G, as are looking for some opportunities, and I found one to be akiva fellow Kepa you're not familiar with that company, their one thatthey're non profit they're, one of the first companies O do crowd sourcing,they're, focused kind of in the microfinance micre lending space, anemerging markets like Kenya. So I was accepted into this Kiva fellows programand they placed me in Nirobi to work on this kind of Kiva Zet project, so put aplace me anywhere in the world. They placed me kind of randomly, maybe notRaun randomly for them, but randomly for me, an Irobi and once I got there,I mean Kenny has an awesome place. Canan people are awesome, there's tone,I'm an outdoor guy, there's tons of do outside, so I really loved the countryand then, when I got into the startup, because system people are work at somereally big kind of media challenges, therelots of Smart People here and yeah.I really like the vibe of the scene here, but I like the pore working on. Itype of people that are drawn in this market and yeah love the country. Sothat's Ho I ended up here is the Silicon Savannah. I understand the termit is like. Is that related to Kina specific, or is that, like Africa? Youknow that region of Africa or like this? Obviously, this podcast is not aboutIot in Africa specific, but for many people this might be one of the fewtimes they're hearing about IOT...

...development in Africa. Do you give uslike twenty seconds on Silicon Savannah? What that means? Yeah I meansicansSavanna is really focused on on thy Robi and what's been created here. So there's been a lot of investment inecosystem development, and so there's one there's some great universitieshere in Eust Africa, specifically here in Ni, Robi pays like universty inIrobi, strathmores and great schools that R that are putting out talentthere's a lot of focus from development finance institutions in this part ofthe world. The UN is based here a lot of embassies ar are located here, sothere's a fair amount of funding and a lot of educated people, and there wassome initial kind of startups at started. You know popping up here aboutten years ago since then, there's been a lot of investment and accelerators.There's been pop ups at kind of venture capital firms and there's been you know,a few kind of win. Some kind of early success stories come Ati. Kenya,opinions are also very innovative people theyre also quick adoptors, annew technology there's over. Ninety percent of the population uses momilmoney here. A lot of innovations come out of this part of the world, so it'sbecome a real hub for startups in heast Africa, and so when Thi SAS Lokansabout Savannah they're, really talking about Whas, been built here in Tharobycool welllet's get back on track I'. I'm really interested to hear about t ethe technology solution you guys have have developed so you're. Looking atthe energy market in Kenya, you know you're, seeing this like nondemocratized access to clean burningfuels or like for most people to get access to to natural gas like veryexpensive or may you know not like the unit economics, don't work at theindividual level. Can you talk about like the process of you landing on anIot solution as the solution to address that problem, yeah, so just kind of setthe stage, so I mean when you're looking at a place like my Robi most,the population is living. These really bensely populated a informal typessettlements, and so you know cooking with, like natural gas, you know in thestates or in Europe, really isn't possible in these areas that you can'tput piped gas infa structure into a place. That's kind of informally set upYOU KNO! There's there aren't tarmacc roads, the houses are, are not you know,built to the traditional code that you're used to this. Is I a reallyreally challenging environment right D, especially so t'd be very dangerous andexpensive to put Natal Gasin. These areas are also growing, really fast,urbanizations happening in Africa and most part in a lot of emerging marketsreally rapidly, and so these places are actually you know, they're not static,they're growing every day, so it's a really challenging environment. If youlook at the livelihood of the average household here, they're living on lessthan eight ars a day, some a lot of houses less than five dolars a day anda lot of their income streams are unpredictabble. We're doing you knowsix seven things in a month to make ends meet, and so, when you look atthat person's day today, it makes a lot of sense in the by charcoal wood orcarosene that they can buy in small amounts and that's what our foundingteam saw early on as they're kind of kind of pattern of behavior on cookingand for curing fuel. So we saw folks teing up in long lines to buy thesefuels. We started doing the back of the NAPP in math. We're like wow they're,actually spending a lot more money to cook, with charcoal than I actuallyspend cooking with gas every month, and so first is like well. This is a realinjustice, but then it also s. The next questionwas like well. Weu know what are the real barriers to cooking with gas, andwe understand why we can't put pipe gas in the home, but why can someone notafford a gas cylinder? I Don' know if we know they're spending more than thata daily basis. So we realize that really just that gas is only sold inlarge amounts. You have to buy the cylinder itself with traditionallywould be. You know, podcast infastructure. You have to buy thatyourself, which is the cylinder, and it can only be refilled on large amounts.So,...

...although there are so many analoguesolutions to this, we thought hey, there's a real opportunity for us tocome in. You know with an Iot device to not just change the format in which gasis sold, to try to find a way to sell on a fractional basis that matches upwith the ditiday life of the customers were targeting, but the Iotiq omponentcould also help us be really efficient with the supply chain, because theareas again that we're working in are really challenging to do distribution,Tha, there's, no addressing system that there's not street names and so havinginformation flowing from the home having better control o the assets wethought could also really improve the supply chain. So it felt like the rightsolution and it felt like a way, thet really displaced. A lot of these dirtyfuels that we knew were not the right product for customers, just just all.They had access to got it and like okay, so you ere now like narrowing scope. WeI think that we, the audience- and I are starting tounderstand, like the problem you know and how you guys are looking at thesolution, makes a lot of sense to me. Talk about the device that you guyshave built. You know to actually begin to meter this out. Some of the likeunique problems that you needed to solve. Can you talk about thatsubmetering device? I don't even know if that's the right term, but that youguys have built and like some challenges along the way that you know either technical or userinterface, or you know whatever that you guys have ave needed to solve inorder to be successful and like I actually talk about the device itselfsure yeah I mean to like any Hardwar Company, I mean developing hardwaresthat a nightmare s s. We had a place that we wanted. We had a very clearidea. I really elikant solution and then we kind of had the bump, alonguntil we landed on on what we've got today. Look when you, when you look atthe technology that we needed to solve this problem, a lot of the things underthe hood are not novel right. Wou have a gas cylinder gas. Is You knowpressurized so that it's in liquid form right, and so it's coming out at highpressure, so we have to regulate it to get to low pressure once it's regularedregulated to get low pressure, because that's what we can cook at right reallyefficiently once ye regulate it, we have to measure it and that's also notnovel. There's lots of devices out there that can measure gas. You know inflow once we've measured the gas. We need to communicate that measurementback toe kind of a system, so we have some kind of a communication, a modulePCB. It's kind of storing that information right and then you know ina prepay format, you need a to be able to shut off access right to gas. Itfolks haven't haveen paid for it. So you need some kind of a bow fright.That's going Ta, that's going to lock. So all these components are are notnovel that are under the Hood, and so the first question is you N W? Do weneed to develop an entirely new device at all? Right I mean this is a gasmeter essentially well there weren't gas meeters out there that were reallythat could do what we needed them to do. BETTERAR managing gas at high pressure,most gas meeters are dealing with. We have so much lower pressure than ot ofa cylinder there, weren't meters out there that would lock on top of ecylinder and would be tamper proof the way that we need and then also from anaccuracy perspective there weren't meters out there like in a commercialsetting that we're selling gas and he's really like a micro amount. I meanwe're talking about focus that need tod buy. You know fifty grams wort Hof gas,that kind of accuracy we weren't finding in meters that were availablein the market, so pretty quickly were backing into a totally new device again,knowing that hey a lot of the components may not be novel but that wewere going to put them together. Will be so we sertthem that journey.Obviously, there's a lot of prototyping going on. We didn't actually land onpoint in thevice in the cylinder off the bat we actually tested putting adevice instead of a stove, but we started foind that really quick. Lothat hey this we got. We have to get to the household level and understand howpeople interact with this device and the environment that it's going to bein and pretty quickly realize. Okay, there's a lot of safety concerns tohaving this built... bilt into a stove that it needs t af we're going to reallymake this thing. Tampered proof is and o have to lock on the on the top of thecylinder, so people can access more gas and we need to be able to control asmuch as we can. The safety side in our own office so having multiplecomponents in the household would be more dangerous. So how much can you putunder the same kind of housing and Qa within a controlled environment beforewe get it into the household? So we start doing all these learnings once wegon of got out in the market and obviously it's been an intative processlike any other hardward company, but we've landeed a proct that still kindof has those same core componentry, but we've had to make some trade offs andwe've had to spend some time really understanding what the customer needshut. Gas companies need to be able to do this really efficiently and provideultimalion experience for a customer that is going to keep them from. Youknow, excited to cook with gas and keep them from cooking with other fuels. Soour audience is a lot of business leaders that are thinking about or inthe process of developing a connected device, ift solution whatever for theirfor their business. You know we hear a lot at very that heythis thing that we're developing there's a lot of off the shelfcomponents. It's you know, this thing is is Mo, is mostly you know,Aplusbplese different off the shelf components configured in a unique anddifferent way, or the application of the thing is unique and different, butyet, as you've discovered, you know, a NB ND C may all be standard off theshelf components, but the application or their integration is not at allstandard. If you were talk like speak directly tothose people out there, like what learnings do you wish nick had known atthe beginning of this process that he knows now about what look? May lookrelatively straightforward because, as you said, the things are not novel. Thetechnology is relatively off the shelf at the individual component level, butthe application or their integration is not. What would you say to business wheth?you re out there hat. That is in a similar position about the journey thatthey're about to embark on. There's a lot ot that I would do differently ango back and take this thing from the top. But look, I don't have a technicalbackground, so you know I leaned a lot on the tonabl guys on our team when wescope this thing out of the beginning, but for my perspectiv there's a coupleareas where we might have been able to we actually some oft Zom we did well,then some erors whery might have been able to potentially speed things up.The first is once you understand, I tink the the experience o you're,trying to provide your customer and what youbelieve are the risky assumptions like in your business model there ar a lotof ways to detest those without having you knowfully big device and so for us working with gas, not just from a safetyperspective, but also disfurther. You know the standards, regulations aroundselling gas and measuring gas right, there's a lot of hurdles that you haveto overcome, just to get your first devices out into the market, and soyou've got to find clever ways to get out there and test as soon as possible.We had some prototymes wo, you're testing. You know what how a customerinteract with it and they were incredibly low, fidelity cardboard yeahusing low fidelity prototypes. You can learn a lot about your customer withoutgoing through the Rigamorole of developing, even like initialprototypes and trying to put those out in front of you know, frodic customers,and I think that the some of that I think companies could doright is say hey. I see that a lot of these components are off the shelf. Icould put somein together rather...

...quickly and get that out to market, andit's actually at least for gas right, there's a lot more. That you've got todo to probably get that fully approved to get that out there and safe in frontof someone's get that n the hands your customers and you might have skippedsome really risky assumptions that you could have rooted out was somethingthat was. You know much more kind of low fidelity like even cardboardprototypes like like what we used. I Thay it. So I think that's that'ssomething that we did. That was probably a good thing. It helped us land on anevice that I think was much closer to where, where we need to be- and I thinkwe save a lot of time and money by doing that- testing of front on theother side of it. Something that I think is it's somehin that you should beparallel processing is that, although you may have something thatagain has lot of off the shelf components to hit the unit cost thatyou need to hit for this to be a commercially viable product. Usuallyyou can't have all off the shelf components under the hood at some pointin an artist start, needto do custom componentry when you go to the valueengineering process, you're going to need to start making some some tradeoffs and the thing will get a lot more custome over time just to to get yourcost down, and so I think the earlier you can start thatprocess. Wil will allow you to avoid a stop start, and I think for us.If we did things a bit sequentially. I think that was tied to the timing around our funding. To behonest in our band with, but if we could ave parallell process that pathethe scale and that path to know a place where our UNICONOMICS were we're,looking positive a lot sooner and then it could have probably saved a lot oftime off our path. So, although you're you're, taking your time and usinglower fidelity stuff up front, I think should be taking that information andfeeding it into what you believe is the more scaledproduct. You'R T have in the future, and that's your entire thought processaround not just was under the hood, but your Glibal supply chain, Hanin, sethat,manufacturing everything you need to do to hit both the unit cost and scalethat you need to really have this thing. Take off. So thank for us. You know wedid the front end right where we were doing a lot of the really importanttesting and interating really quickly and not doubling down on design tooearly. But on the other side, I think that we probably could have done some of thevalue engineering work. A and thinking around you know what a scaledproduction should look like. What our global supply change hould look like,which had done it a lot earlier and let to kind of a stopstart, so you knowtake aways there around parallel planning. You know like getting yourexisting prototype as tight as possible, while also thinking about Vto, notprototype, all the way through then beginyou know thinking about you to. Wesee that a lot. Can you talk about this idea of, like accidentally, shamingyour customers during that manual weighing process, because they wereembarrassed at the lack of gas being used? It made them. Can U expand onthat yeah? You N, to keep yourself honest right, so you're trying to bescrappy and do a kind of a clever test and for us you're trying to test youknow people's willingness to pay. You know you know for to sell gass andsmall amounts and the way that we did it. We thought was really clever goingaround and weighing people's cylinders every day it didn't require any. YouKnow Technology Development Av get to the kind of the rout of the challengeearly. The problem is when you knock on someone's store, veyday and you weightheir cylinder and then you wrihte down the weight, and then you know you knowther take payment a thin to top up or you know if they didn't cook anythingthe night before they're having to look you in the face, as you waigh thecylinder and say sorry, you know I cooked with something else or nor I ormaybe it hs, really low income. You Know House Wild Ey, they weren't ableto have a Mela nite before, and so we really felt that we were skewing thedata, but by this really invasive approach and so Wen. We didn't thinkabout at the beginning, and I think we...

...had to look back at the Dattaafterwards and say: Hey. I wonder how much you're driving consumption justyou know because of the you know the social pressure that we write, that wewere giving them by knocking on their door every day. It's something that wehear a lot where early Beta customers start to root for the the product company. You know they'repart of it. They want to see the things succeed, and so they end up, as you said, skewing the outcome and-and it has all of these unintended downstream effects where people aresaying hey. Look like usage is higher than expected. You extrapolate thatforward and say you know we could maybe see you know this much impact, and butof course that's you know related to like the personaleffect that you pointed at. Can you talk a little bit about likeanother thing you mentioned hat that I wanted to kind of pull on a little bit.Is We hear a lot where the device developer? You know either the personat the company that's in charge of it or the company itself has these metricsof success and they say: Look we're going to get fifty units out in thewild and that's what we're going to consider success, but maybe thatperson's CEO or that person's venture capitalist says well, hang on, you know,five hundred is actually the benchmark of success. Did you guys have like? Wasthis something that resonates with you? Is this an experience that you guyshave had and if so like? What lessons learned? Can you share about likesetting the goal post in the correct place and getting everyone bought inaround the correct goal, post, location, yeah? This is one o that most deeplyfrustrating things about w launching at a hardware. Harwar company is that fora hardware company to hit it between the UPRIGHTS is more challenging. You know andthere's a huge difference between fifty units and even a hundred units orthousand units, and depending on, where you're at an youre in your developmentprocess and so getting to a place where you think that you reach the top of themountain, and then people telling you you're not actually there youve gotthis whole other thing to go up is is really demotivating for the team andreally frustrating so we've definitely had that happen. I mean. The firstthing is that if you can have investors that have a background and investing inhardward companies, that will help a lot, because they can start to at leastempathize with your position and understand how challenging is even justget to that next. It that next point so for us, one of the challenges thatwe've had is that you know we've got a lot of different folks that are thatare setting some of these parameters and wehave to really get deeper on. What is this sumption that they want to test?And I think that's been the real question. Is this a question of aroundthe technical fesibility of what we're doing you want to? You want to see howthis thing actually works out in the wild, and you need a certain amount ofyou know: Aa Sample size in your mind that shows that and wea really dig intoit like what is it? What is the actual data points that you need to understandthat this technology works and that it's feasible at small scale and atsome significantly larger skills? As that the question: Is it more aboutuser engagement and you have real concerns around Hoduc market fit andyou want to see enough cycles, af data and that can come in different forms.That's actually like the number of customers or it's a smaller group ofcustomers, O Er a longer period of time, for it's more data points right fromindividual customers. So we spent a lot of time as we get becamemore mature. I thing as a leadership team bigging into what is it we'resolving for and when you got different folks in the room that are trying tokind of SCO what their mean, what they believe theriskye assumption. Is you start to put those out of separate assumptions thatwe're testing so that Nical, as some D, is a technical concern that we need totest? Let's make sure those books are covered if it's more of a market baseconcern let's separate those out, and...

...we might be able to address thesedifferently so beally on. I think we just trusted. You know the folks weretalking to D Inot. Now that they're, not Tom, they meant to change the goalpost. It's that some unders hat disconnect also between the folks thatare on your board that are telling you hey. This is what we believe the marketis looking for from our experience, then you go out to market an slopd inthe face and they're looking for something different. So again, I thinkyou have to figure out what is it that we're testing for, and you have totriangulate it from the fost close to the business and, ultimately, what thatthat th audience is going to make the pround decision, which, if it's thenext funding round right it might be or this next round of investors they mighthave a different looking field in the folks yare dealing with today, threethings that I think we've talked about today. That are likeinteresting that that you at home, you know, might want to consider as youryou know, thinking about your project. One is this idea of paralleldevelopment. So I get a product out in the market that utilizes as much offthe shelf as possible to get that initial feedback. But at some point youneed to begin to build your view to and parallel with that, not wait until thatprocess is completed to you know, I think, there's a an idea of like really lookingcarefully at your Beta adopters. You know you want to be mindful of like is:Is there any opportunity for this data to become skewed by the you know, thepeople, a that we've included in Arbeda and and their behavior, maybe not beingrepresentative of o the broader market and and then the third one is which ishe one? You just talked about this idea of goal posts and getting keystakholders really bought in about what success looks like across you know allthese keep parameters so, like units deployed, you know probably likesomething around like utilization. You know market adoption, consumer behaviorthings like that, like really defining that, so that everyone agrees whatsuccess looks like and the team can go drive towards that Nick wwe're justabout out of time, but I did want to ask you one last question you mentionedat the top of the show and user listeners from the United States wouldbe familiar Ben to Oregon one of these towns. That's totally blown up. I didnot know anybody was actually ever from. Then it seems like the kind of placethat no one is actually Y A. I live in Bosmin Montanas, the same thing I thinkthe last board and raised Bosman person left ten years ago. What what'ssomething when is the last thime you've been back to Ben, do Oregon and forlisteners out there that are familiar with the town? What's the thing thatyou think has chained you've seen change the most about Ben to Oregon?Give us twenty seconds. Anyone. Even though Bens change a lot good at bend.Organ, it's a magical place greet place to grow up. You've got the CaskaMountain Range, Tho, the shog river, running through town. I think a lot ofthe things that I love about, bandard are still there. When we move there,there is about fifteen ousand people, so I didn't wasn't born there, but Imove there. I was four or five years old. It was mostly a lumbertown and nowthere's about a hundred thousand people there. I don't know what they're doingfor work the economy there. They have a bit of a Sart up scene, there's a bigkind of scene around growth right, but people therere are't too worried aboutthat they're, more outdoors people, so they're skiing, they're playing golfthey're out in the woods they're having a good time soin. The main differenceis yeah the scale I mean it's. It's huge Jus't gone up by like a factor ateight an in size. The economy is totally changed from being a lumber,town and yeah, but it'd sil ti that the people are the same theretheyreoutdoors people they're having a good time and it's a beautiful place to live.I love it. Ladies and Gentlemen, Cofounder and CEO and a Fatgo Energyand Longtime Ben to Oregon Resident Nick Quintong. Next, thanks for beingon the show today, right thanks so much yorbout me, you shouldn't have to worry about IOTprojects dragging on or unreliable...

...vendors. You've got enough on yourplate. The right team of Engineers and project managers can change a pivotalmoment for your business into your competitive Ange. Various close knitcrew of ambitious problem, solvers, continuous improvers and curiousbuilders know how to turn your ideas into a reality on time and up to yourstandards, with a focus on mitigating risk and maximizing opportunity willhelp you build an Iot solution that you can hang your hat on. Let's bring yourIot idea to life, learn more it very possiblecom. You've been listening to over the AirIot connected devices and the journey. If you enjoy today's episode, make sureto hit subscribe in your favorite podcast player and give us a rating.Have a question or an idea for a future episode. Send it to podcast at verypossiblecom. SEE YOU NEXT TIME.

In-Stream Audio Search


Search across all episodes within this podcast

Episodes (17)