Ep32: Greg Jackson 'Building the Billion-Customer Energy Company'

Transcript

Cleaning Up is brought to you by the Liebreich Foundation and the Gilardini Foundation. Hello, my name is Michael Liebreich and this is Cleaning Up. My guest today is Greg Jackson, CEO and founder of Octopus Energy. Octopus is a challenger utility in the UK, which has grown over the five years since its founding to serve just under 2 million households, growing at about 30,000 per week. Octopus, under Greg's leadership has also started to expand overseas and to offer its unique technology platform to other utilities. It's a fascinating story. Please welcome Greg Jackson to Cleaning Up.

So Greg, welcome to Cleaning Up.

Well, thank you, Michael, and welcome to Octopus' lockdown office where I am very, very lonely.

Now, I hope despite the fact that it's a place of work, I hope that you've got yourself a brew there because I certainly have, it's evening here and I'm stuck in Switzerland again. So cheers.

Yeah, well, I've got a delicious Beavertown Neck Oil from the Octopus' fridge. Ordinarily, we get the whole team together on a Friday night, we have fridges full of alcohol and non-alcoholic beverages, to bring our team together and through lockdown we're not getting it through very quickly.

Well, so I think, you know what, we'll come back, I suspect, to talk about your leadership style and the culture that you're building at Octopus, because it's very particular, I already know. But let's start with... just give us an update of where are you with Octopus, just so you know, our audience is... they're all super engaged. They're all very smart, but they don't know everything about the energy system. They certainly don't know the UK utility landscape. And they may not know Octopus, strange that that may sound.

So Octopus was founded about five years ago, by a group of people led by me who came from a technology background. We built technology platforms in a whole bunch of industries, working for third parties, you know, from SAP to I mean, retailers, loads of companies. And I think when we sold that business, we looked around to say which section the world has not yet been disrupted by technology. Energy was this kind of sector, which served pretty much every customer in every advanced economy, but was largely undisrupted, I guess, because of regulation. Not only that, but you know, energy is at the forefront of the fight against climate change. So there's this opportunity to use technology, we thought, to drive down costs, improve service and tackle climate change. Where we are today is we built a platform, Kraken, which is designed to be a global platform for running energy companies. We've got 2 million household customers here in the UK, let's be precise, 1.953 million, there'll be 2 million by the end of next week. And in addition, we've got businesses in <inaudible> in Germany and in the USA. And we've also licenced our platform to Origin Energy in Australia to Hanwha who run a <inaudible> in Australia. We've licenced it to E.ON and Good Energy here in the UK. And also we've formed a joint venture with Tokyo Gas to launch Octopus Energy Japan. We've got an office in New Zealand, about to launch there. So it's pretty exciting, Michael.

That's extraordinary. Now let's just back up a little bit. And let's put the UK piece, first of all, into context. So you've got 1.93, you said, million users, and at the end of next week, you'll have 2 million so you are growing at 70,000 a week?

Not quite 1.953.

Sorry. So, 50,000, 50,000 a week

It's 29,000 last week,

Okay. With that you're forgiven for rounding up. These are big numbers. And that puts you were in... Because we've got the Big Six, and then you've got the various challenges. Where does that put you?

I think it makes us number five in the UK market now. Yeah. So when we started, people said, how are you going to kind of tackle the Big Six and well, I think we now are one. You know, and I think that's the customers were serving directly here in the UK. Through licencing...

You licenced to two other groups. So in total, how many households would be in some way being touched by your platform, Kraken?

Yes, total number of households at the moment is probably around, sort of three point. I don't know between three and three and a half million households in the UK. And then the total number of licences globally is now 17 million. So, you know, sooner or later, about 80% of our customers are going to be served by people licencing the platform rather than us. We've got an ambition to get to 100 million customers by 2027 on the platform. And you know, I'm beginning to worry that might not be ambitious enough.

It doesn't sound very ambitious. When you look at all those countries you're going to be in.

Yeah, no, you're right, though. It's fascinating, isn't it? energy is such a colossal market. So you know, our revenue... Now we are four and a half year old business in market, our revenue is 2 billion pounds or a two and a half billion dollars. And yet our global market share is what probably about 0.02% to 0.04%. You know, we've got an incredible amount of work to do to make any kind of difference at all.

Right? But if you look at the UK, you're sort of home, your home and pioneering market. Three and a half million homes, what are they? There are 23 million homes. So, you know, that's, that's not bad. That's 15 odd percent doing the maths, you know, probably wrongly, but on the fly. That's, that's pretty good. So talk to me about that platform. What does Kraken do? Why is it driving so much growth? Or is it Kraken driving it?

Yeah, it's really important that it is. So we didn't start out as an energy retailer that built a technology platform, we started out to build a technology platform. And in order to be able to test our hypotheses about how energy could work, we needed to build an energy retailer to test the platform in market. So Octopus Energy's retail business, although it is incredibly important and, you know, is the driving mission for the vast majority of our team. Part of its job is really to prove the way what we think technology can do to drive the energy transformation, to drive costs down and help consumers benefit from green energy, rather than seeing it as a problem.

And so there's a few, there's a couple of things. There's pushing the cost down. And you also talked about green energy. So what you're saying... I mean, is the green incidental, or is the green core? Let's start there.

It's a really good point, right? So the green's core. If you look at Kraken today, it's a bit like an iceberg. The bit that's visible, is a better engine for running an energy company, you know, it's great for customer experience, customer service, efficiency, you know, driving down the cost of operation, it does everything from sort of, you know, managing the telephony and digital contact with customers, right through to helping, you know, power, loads of machine learning for forecasting, all that sort of thing. But actually, deep under the hood, what it's designed to do is make the transition to renewables a positive economically, financially in energy. What I mean by that is, renewables behave different than fossil fuels, we all know that. You know, with fossil fuels, National Grid, or whatever your equivalent in any country is, in the vernacular, can wake up every morning, there's a sort of fairly predictable demand curve, and they'll just turn on and off, you know, fossil fuel stations to meet that need, in renewable world that you have to invert the way you think, you know, National Grid wake up in the morning, they look at the weather forecast, and then you have to turn demand on and off to meet those needs. Now, the traditional energy system finds that a really difficult idea, but actually, you know, it's really at the core of how you build relationships with customers, including the deep technology for it, then you can really start to match from consumption to time of generation. It is a great fortune, that exactly the same time as we're building more and more renewable generation, we're changing how we consume energy. So for example, an electric vehicle, typically here in the UK, an electric vehicle might hold the best part of a week's worth of electricity for a typical house. So you've got this big discretionary load, even before we get into things like vehicle to grid, what we need to do is be matching the time we're charging that car with the time we've got renewable generation. Once we're doing that, by the way, we can monetize renewable generation better, which means that it can get cheaper, we can attract more investment to it, that then paves the way to sort of decarbonized heating, using the same technologies for matching consumption demand. We can really move to a renewable system cheaper with that kind of tech, those kind of consumer relationships. And I think, for a long time, energy companies, utilities, networks, grids, governments...were all debating, but the only way of really finding out what you can do is just getting out there and doing it, finding out what works with customers. It's how we found that, you know, a smartphone with a touchscreen was better than the old kind of mobile phones. It was only when someone built one that consumers adopted, that we discovered that was the way of the future and it changed society. So I think we can now do the same with energy.

And one of the things that you have done, presumably, you couldn't have done it without Kraken, without this platform, using the traditional software that everybody else has got. One of the things you did was you launched the agile tariff, which is the UK is first time-of-use tariffs, right?

Yeah. So I mean, agile was, the whole point really, was to say, look, we, as an energy company, see dramatic changes in the cost of electrons throughout the day, day to day, you know, month to month. And yet, typically, consumers are passed a flat cost. Now, if we look at other kind of consumer markets, where prices vary, either tomatoes or strawberries, when you get an abundant crop, prices are low, when you know, the crops are less abundant prices are high. And that enables efficient markets to clear and I think, with less waste, you know, and with better value overall for consumers. And we need to do the same in energy. And I think, yeah, the agile tariff, the idea was, you know, we essentially have a formula, which we publish, that connects the half hour, the price of electricity to the price consumers are paying in that half hour. We then really critically, actually, we published alongside that a bunch of APIs that consumers don't need to know about the APIs. APIs are there to enable companies, people that make electric vehicle charges, companies that make heating systems, whatever, to automatically read these price signals, and then optimise on behalf of the consumer. So many of the consumers will literally just get home, plug their electric car in and they don't need to think about the price of electricity, it will just optimise based on this tech. I think the other bit is, of course, consumers can choose to optimise manually on top of that. So the UK has got simple heuristic. Electricity is expensive between 4pm and 7pm. It's also expensive when it's not windy. Now, that just means that basically, if you get home from work and you cook dinner for your kids, switch the dishwasher on at half past seven in the evening, it'll be a lot cheaper than at half past six, you don't need to check the price. It's just gonna happen. So I think it's really interesting for me, that by publishing, creating that product, we were able to see how consumers responded. And it's been, I think it's really heralded a way to making renewables beneficial for consumers, rather than the world, we were stumbling into of people seeing renewables as being a cause of high taxes and so on.

How hard was it? Or how easy was it to pass on that price signal to consumers? Because this is not something that everybody's doing. And it's not something... there's been quite a lot of nervousness about that, because it does mean that at certain times the price will be high and everybody starts, you know, speaking about energy poverty, and goodness knows what. So the regulators have not been pushing you to do this, have they?

No, it's quite interesting. Our first dynamic tariff was actually a daily tracker, the idea of that was that customers never need to think about energy bills again, like, you know, essentially, if you know that you're always following the market. It's a bit like with the stock market, instead of actively trading you can just have a tracker product. So we kind of tried that. But a journalist said to us what about when prices spiked? And we showed, you know, the data, it said like it's, it's just not a big issue in the grand scheme of things. But she said look that people worry about that. So what we did with agile is based upon that feedback, that question, we put a cap in, so the price will never exceed about 35 pence a kilowatt hour. We've modelled it, so we can take the risk on those rare spiky occasions. And the rest of the time... Underneath that the consumers got full dynamism, including by the way the prices going negative. So (...) and again it's the kind of thing you can only do when you're iterating in market. But I think we've built a mechanism that really helps consumers kind of use energy in ways which is sympathetic to the system. <inaudible> do use it in ways which are sympathetic to renewables, then they'll get very cheap electricity. But it still protects them from the worst. And we could have spent months debating that, but we actually spent four days building it and now the world can see how it works.

And it's great you know, if you go on Twitter, you must have seen this a lot a lot more than me, you will see people saying yes, I made £1.20 charging my car last weekend or whatever it was. So there's definitely that sort of the gamer types that really really enjoy that. But how much difference does that agile tariff make in terms of actually moving electricity demand around? Because for that you need not just the gamers, you need, I was gonna say normal people like me, but that might be pushing it slightly.

Yeah, so I think I think let's separate dynamic pricing within the energy system from the way in which the consumer experiences the pricing. So agile tariff is one of I think, 27 different tariffs we have that use price signals to change the way in which people use electricity. By the way, it's not that you go online have to choose one of the 27, right? We make all that very, very easy for people. But so with agile, we saw a 70% reduction in peak usage, usage at peak times, massive change. And most of those people are not the gamers, most of those people are just plugging electric vehicle or indeed heating system. And it's doing it automatic with no thinking. On top of that, some of them just have that simple model in mind, which is, you know, just try not to use your washing machine dishwasher between 4pm and 7pm. Or you tumble dryer, right? Using that side of that you don't have to do esoteric three in the morning, right? Just avoid the peak. I think...Outside of that, we also have things like, you know, electric vehicle tariffs, that have got fixed slots, so 0:30 to 4:30am, for example when it's very cheap. Or ones where you can choose your cheap slot, you know, you can choose a slot 8pm to 11pm. The shorter slot you choose, the cheaper it is. So using all those, what we've found is we can really use this price signals to move consumption around, to flatten the curve, at a very granular level. And yeah, what we'd really hope is over a period of time, you'd be able to take a substation network of streets. And as people are electrifying transport and heating, we can flatten the curve and get dramatically more out of the network as it stands. There are two big benefits of this I think, Michael, the first one is that obviously it reduces the need for additional infrastructure to get more out of your existing stuff. We make it last longer. We can electrify and decarbonize cheaper that way. I think the second one is that, unlike many other times in energy, one person optimising in their own self-interest, using the cheap slot, using the cheap electricity is actually pushing the price down for everybody else as well, because they're reducing the competition for electrons when they're scarce. So we're keeping the infrastructure cost down which is good for everyone. And we're making better use of electrons when they're available. And I think that kind of behaviour, that kind of economic model means that even when you've got relatively few people in the gaming category, and then quite a few people in the optimised category, and some other people are time of use category, that all contributes to a cheaper system for everyone.

Now, we are electrifying transportation, I think that's already clear. That's kind of you know, that the thundering herd is going off towards EV. We are about to thunder, the herd is about to start thundering towards heat pumps, or at least Boris has said, the Prime Minister has said, we're gonna have 600,000 installations a year. That's a lot of electricity. Does the benefit of what you're doing. Is that... Are they diminishing returns? Or are they accelerating returns?

I think they're accelerating returns. It's a great question, right. So essentially, today, you know, energy in the UK. And in fact, even in most, in fact all the competitive markets, it looks like God's plan. You know, the sort of the Soviet-era central planning system, where we may allow a tiny little bit of kind of entrepreneurial competition, right at the edge with a kind of clever consumer. But governments are still planning how much of each kind of generation they're going to have to meet needs. There's still kind of planning these kind of monopoly grids and monopoly charging systems. And the problem with that is just like, it was in 1980, that Mikhail Gorbachev visited London, and asked Margaret Thatcher who did the central planning, because the Soviet Union's best trained economists were doing it. And they had queues at the bread shops. He saw no queues in London. And she said its price signals. Now, you know, without price signals, what you end up with is now milk lakes and butter mountains and at the same time shortages or massively overspent inefficient systems. So, you know, what we're talking about is finding ways to harness price signals throughout the system. So that investment is attracted to the right places, the product of that investment is monetized optimally. And consumer, users are exposed to price signals, participants in companies such as ourselves are able to invest where it's going to be most effective. So what that means, in simple terms, is let's take product like agile right? Today in the UK on a windy day, we literally pay wind farms to turn off their generation, right? Because the vast majority of the UK electricity system isn't exposed to the pricing that says right now electricity is very, very cheap, right? So we just turn off the generation. This is bonkers, right? Because other times where we still turn on coal power stations, sometimes, right? Tonnes of gas, other times, because there's not enough wind. Now, if we can use price signals to mean that, you know, every green electron that we generate can be used by someone, can be put to use, then first of all, we will more efficiently price those electrons, that will attract investment to the right places. Secondly, will enable technologies that shift consumption, that make the most of those electrons to work based on the basis of, for example, cheap electrons during times of high generation. So, you know, what we don't need is governments choosing how much battery storage we're going to have here. I mean, I say governments can be great, but it's all monopolistic, right? As an energy company, we should be saying, first of all, can we get our consumers to use electricity when it's abundant, and when the system is empty. And if we can't, then, you know, for example, what we might try and do is...So right now we'll build some batteries at the end of the street that will soak up electrons when they're abundant, and they'll be available for our customers when they're not. And then you can scale that all the way back to generation. If we follow these price signals, then I think we can attract more investment into generation because more electrons will be monetized, we can drive down the cost of electricity, because we're using them rather than throwing them away. As we drive it down, electric vehicles get cheaper to run. And electric heating becomes ever more realistic. So I guess for me, you know, there is a huge compounding benefit in working like this. It's like data, do you remember when you're um, I mean, you and I are probably similar ages, Michael, I think you're a bit younger and certainly more handsome, but your mobile phone, you used to have to delete text messages because there wasn't enough memory to store them. And now you record every moment of video of your kids' life, you'll probably never even watch them. And they're replicated 25 times on your computer and in the cloud, and everywhere else. You know, that abundance of data, it's not like, you know, kind of the world of data storage providers and cloud platform providers and so on. It's kind of all gone bust because the cost is dropping every year. We found utility for it. And I think energy is like that, you know, we've got this opportunity to relentlessly every year drive down the cost of electricity now, and to do it with zero carbon. And it'll create whole new industries, just the way that ubiquity of data and computing power have done. I've got examples by if you want.

Look, it's music to my ears. You know, I let you go all there. Because this is exactly you know, I wasn't campaigning, but I was trying to persuade people that price signals matter, that we needed to move to reverse auctions to get cheap renewables. Here's a little known fact, I actually worked on perestroika, I was in the Soviet Union in 1990. I like to say, working for the Ministry of Aviation and Special Constructions, it wasn't quite as simple as that. And you're right, there's no I was meeting Soviets all the time, who were convinced that I was lying to them. Because I said that at a building site, you would just go and you know, you if you needed bricks, you would just kind of send out a fax to a bunch of builders, merchants and get the best offer and the bricks would arrive. And they're like no, no, no <inaudible> somewhere in the system. So I hear you completely. I hear you on that. Give us one or two examples of that, because you wanted to. And then I want to ask a question, which is more difficult.

Yeah. Great example for me is I discovered that we kind of got this series of customers doing vertical farming. Now, you know, vertical farming turns out to be farms that are inside warehouse buildings or whatever, in towns and cities. And in theory, it sounds very nice. You know, they grow the crops locally, so you've got food miles, it's a sealed environment. So 99% of the water that goes in the building leaves in the fruit and veg. And there's no need for pesticides, it's a sealed environment. The biggest single cost is energy. And, in fact, I was talking to one and 55% of their costs were their energy bill, electricity. Using formula agile, what they've been able to do is the crops now sleep at the times that electricity is expensive, right? Because it's a sealed environment. So the diurnal cycle, they can just optimise it, based around, you know, some combo of what's cheapest, what's best for the crops. As a result of this, they're now able to be cost competitive with crops that have been growing, certainly crops that being grown in other countries and then airfreighted in. Increasingly cost competitive with things that are being shipped and trucked. And so I think the interesting thing for me is a bit like no one spotted when the iPhone was launched, that that would be the end of the you know, the kind of million dollar cab medallion in New York, as it would pave the way for Uber. Similarly, with renewables, no one's spotting the incredible number of industries, it's going to transform or spawn when we have these endless green electrons.

Okay, but let me ask you the difficult question. The UK is fairly far north, we don't have equatorial sun every day. In fact, the ratio of solar output from the solar on my roof in W2 is 14 in summer to one in winter. So there's no solar in the winter. So we're talking wind, and the wind can disappear for a couple of weeks, two-three weeks. And you're not going to solve that with batteries. And you're not going to solve that by waking up all your, you know, salad greens in those vertical farms, and saying, right, you know, like... You're not waking them up in fact. There's no wind, you'd sort of put them to sleep for two or three weeks and hope to wake them up at the end, but that's not going to happen. So how do we deal with these long wind lulls, particularly when we've electrified heating, because those lulls could happen? You know, just in a moment when the temperatures plummet?

Yeah. So look, I mean, first of all, we don't, the whole idea that we have to solve this problem centrally, is kind of wrong, right? It's muddle-headed, right? Price signals solve problems. Now, you might say, in principle, we have to be able to solve it, because otherwise we can't go <inaudible>. That's cool. Right? I don't have to give you the answer. Just like, you know, loads of things, to which the iPhone or the smartphone was the answer.

I'm gonna push you on that. Because you would have to give the minister the answer. Because you know, yes, we don't want to have central planning, we believe in price signals. But if the lights go out, or if there's no heating in the <inaudible> homes, and they you know, die of, of hypothermia, then <inaudible>

Sorry, Michael, let's be really clear. We'll get onto that in a second. Okay, the problem is easy to solve. The point really being though, that is hilarious that, you know, we don't sit here discussing how we're going to meet our potato needs when there's been a harsh frost in spring, right? I grew up in the 1970s and 80s. And I remember the centrally planned conversations about you know, what was going on with the Ministry of Production and British Leyland, right. So to the non-British audience, the point being that markets can solve problems brilliantly. So when I give you answers right now, which I will do, they will almost certainly be far from optimal, because the market will cream anything any of us think of right? And it will do it faster and cheaper anyone... Just like, people didn't solve the electric vehicle problem through central planning about two, three years ago, the experts were still saying, you know, batteries weren't going to be the answer. And lithium ion had all these production problems, you couldn't get enough precious metals. It just turns out, it's all nonsense, right? Once you've let market forces do their work, we're discovering alternative to precious metals, we're discovering new deposits. We're finding new forms of battery technology. I say all that...

Greg, Greg there are some people in the Ministry of Hydrogen who would disagree with you...

I mean, one great problem with centralised planning, is once someone is an expert in the hammer, every problem looks like a nail. But the next thing then is that and the great thing about competition is that, you know, companies have got investment in loads of different technologies, are all going to compete, and someone will win. Now to answer your question about the UK and Canada, heating and wind and so on. First of all, this kind of capacity myth, right? I've heard people saying, well, you know, if we're gonna build enough wind capacity to handle winter heating, we'll talk about the lulls in a second, but there's gonna be a lot of excess capacity, it's like, fantastic. That means for a lot of the year, we're gonna have endless green electrons that are near enough free, right? That's pretty amazing. Because what it enables us to do is loads of great things that we can't currently do with cheap green electricity, guilt free, carbon free electrons. Now, once we're enjoying living, you know, so much of the year very cheaply. Actually, we can afford as a nation in the UK's case here to have more expensive solutions to handle things like the lulls over winter. Right? So for example, do we handle the loads by first of all, you've got wind farms all the way around the UK. And my understanding is that, you know, if you have the entirety of the circle of it, in our case, the UK is peppered, then you're down to, you know, typically the worst case is four or five days of lulls, might be longer than that very occasionally. We'll get onto that in a second. But once you've got that kind of capacity, you're then able to say things like, you know, are we using our spare capacity to generate hydrogen, right, and we store it, right. And although it's expensive to generate hydrogen, it would be bonkers to use it for domestic heating, for example, to use it for intraday storage and multi day storage. Even for potentially, you know, longer than that. It's a great idea. And that's, that's one way in which we can <inaudible>. But you know, once you've got price signals operating Do we find that, you know, enterprising companies decide that running, you know, a new undersea cable to Iceland is going to help kind of provide cheap electricity during those times? Do we tap geothermal sources in other countries or dig deeper geothermal here? They're all possible. But you need price signals to reveal what is the optimal solution. So the point really being, to your question or opening question, the physics, meaning that we can easily answer the question about what do we do about for example, electrified heating, even on windless days, but the economics will reveal themselves far more efficiently if we have price signals.

Okay, now, so I accept all of that, you know, as I say you're preaching to the converted? I do think that there will be it's legitimate for ministers to worry about that bit when you know, you and I would say, well, that isn't going to be a problem for probably 15 years, really to get that amount of renewables, 10 years, certainly.

But the thing is, if we do it iteratively, Michael, we can stop at any point and pause. Right, so let's iterate our way there. And if we find it a bit difficult, we pause for a while, we say, you know, we'll keep some of those gas plants running, just while we kind of get better solutions. But we can do this iteratively, we don't have to solve it all in a big bang, it's, you know, my gran today, she's 96 years old. And she's living in a sheltered accommodation, essentially not seeing anyone other than her carers is day in day out because of COVID. And she's staying in touch with the world because of a $50, Android tablet. 10 years ago, or just before the iPhone launched in 2007. So 13 years ago, if you said to people how you can solve that problem, they would never come up with that solution. It happened because we got the technology into the market. Competition drove the price down dramatically. And we opened up solutions known have thought of.

In fact, the French government came up with a solution called Minitel. And they were going great guns until of course, the you know, the private enterprise and the price signals all worked and all of that stuff got ripped out and junked. A couple of times you've referred to yourself as an energy company. Are you conflating energy and electricity? Or would you ever see yourself getting into the molecule business, whether it's gas or hydrogen or anything else?

Yeah. So first of all, we sell a billion pounds, $1.23 billion of gas per year, as it stands, and that's going up a lot. I just don't want to be selling gas, right, you know, I want to be moving away from stuff that causes climate change. So our job is to meet energy needs of customers, just with lower prices, and less CO2 every year. Now, we've looked at alternatives we looked at should we be like, you know, fully backing the idea of hydrogen for heating for example. I think when we look at it, if the mission is to drive energy costs down whilst (...) tackling climate change. The answer looks more like electrification, that'll be the cheapest way of doing it.

Right, right. Now, I think we've established that you've built this platform, because it enables the customer and obviously the energy company to kind of play these markets. And to be very flexible. It sounds like it could be very expensive. It sounds like sort of thing that a conventional utility would go crazy trying to build, or... Is it expensive? And how do you control costs?

You mean on the platform itself?

The platform itself. Yes. How do you keep those transaction costs low? Yeah.

So it's interesting, like we didn't come from an energy background, we came from a technology one. And I shudder, I'm terrified at the thought that energy companies are going to go out and try and build technology. You know, the history of companies in non tech sectors, building good technology is is not great. You know, just like energy companies kind of talked about how special energy is, I think tech companies will talk about how difficult tech is. And I think tech is difficult, because at the very core of it is architecture. Now, if you've got a board, made up of finance people, and maybe people who understand operations, there's no one in there that really knows whether the tech that you're building is going to scale, whether it's well architected, whether it's secure, whether it's, you know, whether you should delay a project, because actually you can get the architecture right, or whether you should keep the pressure up, because actually, you know, entropy means that projects will always expand to fill the maximum time you give them. So I think one of the things for me is and it doesn't just apply in energy, it's around the world. boards need to learn about tech.

I will interrupt, I'll let you get on with that in a second. But just on exactly that point, when I joined the board of Transport for London, I asked them how much they were spending on writing code. Or on their IT as a whole. They've never thought about the question. All the board members, and the senior management, members of Executive Committee, not one of them had written more than, you know, 100 lines of code back at school. And I've written probably a few thousand. But that was it. I was like the most tech savvy person. And it turns out that we're spending somewhere on the order of billion pounds a year through TfL and its various contractors, which they never thought about, there's always people beavering away, writing code on behalf of TfL, with no oversight whatsoever from the senior levels. So we tried to change that. But it was hard. It was hard. That was not their culture.

No, it's not easy, right. And I think, let's just explore that a bit. Because, you know, that we live in a world in which technology is the driving force for change, whether it be you know, kind of the obvious like social media and stuff around the corner, like machine learning. And you've got people who are buying this that don't understand it. And so what you'll typically have is a board, who will appoint a senior executive to manage tech, but if you look at what most of them are doing, they're not defining, they're not architecting, they're procuring, right? So then you've got load of people whose job it is to actually kind of integrate different sets of procured tech, you've got all these best in class solutions being squidged together by contracting firms who literally got misaligned interest. Because, you know, they'll make more money, the longer the projects go on, the more they go wrong and need fixing. Now, I'm not saying any of them are acting, kind of with malintent there. It's just all the incentives are misaligned. Now, imagine running your finance function like that. We expect people who are running companies to have, you know, really good knowledge of finance, not just be able to read a P&L, you know, to be able to think about discount rates, and you know, the impact of different cost of capital, all these kind of questions. And they're expected to do that themselves. And yet, when it comes to 'is our code any good', they will typically ask a senior person in their tech function, who will get a less senior person who will get a third party, who by the way, will then get another third party and eventually comes at line eight, then it says, oh, they say it's good. Or they say it's not good. And you can't check yourself. You'd never do with your finance.

I heard you give a statistic on the cost of administration. You know, you talked about how many sort of SKU, stock keeping units, the average energy company has and you compared with Amazon. Can you run through that?

Yeah. One of the observations we made when we were starting the business, so the numbers might be slightly out of date, but Amazon, at the time, had 330 million different products, right. Which, by the way, it has to get to you by, you know, some combination of planes and ships and trains, and automobiles, and, you know, people knocking on your door, when you might not be in, it's got to take payment in whatever way you want to pay right there. And then. And, you know, by contrast, certainly, in the UK, most energy company got two products, gas and electricity, they've got a metre in your house and the pipe and wire to deliver it. They've got a direct debit, you know, typically. And a government sanctioned estimate for what you're going to use every day. And yet, somehow, Amazon's cost of administration was around 2% of revenue. And in the UK, for energy companies was nearly 14%. And you got to say, like, you know, that kind of spoke to the opportunity for tech, right. And that was just tech to smooth up your operations. Never mind the dramatic improvement you can create for customer experience customer service, and then, you know, paving the way to that much more dynamic, renewable future we talked about earlier.

It sounds so interesting, that's a fabulous statistic, that sort of 14% for two products and 2% for 350 million, whatever the number was, you've done some really interesting things about how you organise the company to deliver service, there's some wetware as well as software, so humans that you've organised as well, correct?

Yeah. So I think it's quite interesting. When we employed about 20 people in customer operations, our operations team, it's not a call centre, the person you speak to does every bit of solving a problem themselves. So there's not a front office, back office or any... and there's not lots of vertical specialties. So it's a more fulfilling job, I think. It was the sociologist, Webber who talked about alienation, you know, like. Adam Smith's pin factory maybe inefficient, but it kind of alienated people from the product they were making. Whereas the mediaeval gargoyle kind of craftsman conceived of made the whole gargoyle and so you know, that kind of ownership, we try to give to our customer operations team. And seeing the job through end to end. But when we were growing, we found that somewhere between 20 people and 50 people in the customer operations team, people started to lose that sense of ownership, they wouldn't answer the phone as fast, they thought somebody else might do it. They kind of had a sense that maybe they've done more than their fair share so it's someone else's turn. And so what we did was we were like, how do we reinvent what we were when we were a tiny business, and we broke our customer operations into teams, each team is about 10 people, each team of 10 people look after about 70 or 80,000 customers, and they do everything for this 70-80 thousands customers, each team is completely self-managing. So we don't have a sort of central planning function. And, and they'll tackle different problems differently. They'll organise their days differently. And they'll organise their priorities differently. But the statistics for the performance of every single team, every single individual are all published, you know, they're all available on our data system. So everyone can see how everyone else is performing. And if someone's trying something, that kind of doing better than they're doing well, they can adopt it. Now, this isn't some sort of hippie free for all, because, you know, first of all, every single member of the team is... they're a shareholder. So they all want it to work. Secondly, because everything's published, you know, kind of, you're continually discovering what's better, and there's no real reason, there's no excuse for not not improving. And I think finally, there's the real reward for success, which is, teams that are doing well grow. And then they split and formed two teams create new leadership opportunities for everyone. And the interesting for me was that, as a manager, is every time something is not working, you have a desire to micromanage, to sort it out. And we have to learn not to do that because otherwise the teams won't sort it out themselves. Secondly, as a manager, you have a desire to standardise, you know, people say to me, hey, but what do you do, if one seems not performing? And it's like, well, you know, the whole point here is, for some teams to do better than average. So we've got to do worse. And we have to accept that because otherwise, we wouldn't get the constant improvement. So as a manager, you can look at that going, literally, I know that that team could do a better job. But if we intervene, too much, too soon, we lose the magic of the system. And it's a really interesting way of building this highly scalable business that, by the way, is growing about 50%, during the pandemic.

Yeah, it's great, because you've created a learning system. And I think that, you know, as I look around, I love to find these examples of how learning you know, small, nimble, competitive, also cooperative and learning beats, massive, centralised, standardised. And there are political lessons that I don't think will go into at least maybe not until we meet in person next time we go for a beer in London. But I hear you on letting the teams, not micromanaging. And one thing I used to do was, if we had a problem, I'd invite everybody who ought to be solving the problem, to a meeting, and then not show up.

Brilliant. I love that.

You know, it's such a simple thing. It's like they all know what they are at the meeting for. They all know what the problem is, they know the problem better than I do. Why do they need me? And it works. Also, I want you to talk about the role of machine learning. You talked about it in a very blasé way just earlier, you said oh, you know, we use it as a forecast and call volumes and blah, blah. But embedding machine learning is also one of the secrets of how you've kept the costs down, is it not?

It is yeah, I mean, look, I'm first of all, to really get the most out of machine learning, you need to have a lot of data that is essentially all in one place. Now, of course, you can kind of simulate all in one place by scooping up or putting a meta layer over it. But the more it's really all in one place, the easier it is to deploy a wide range of things so that the machine learning kind of algorithms, the bots can look at as many inputs as possible very efficiently. And then we have everything from, you know, one of my favourites is the workforce planning. So company of our size, we now employ maybe 500 people in customer operations in the UK. And maybe a bit less than that, but that kind of magnitude. That's typically we might have had 10 or 15 people working workforce planning to kind of schedule to work out how many phone calls we're gonna get and how many emails and the kind of what work can be done by each team. And we have zero people doing that .There's just a machine learning algorithm that is continually looking at how many calls are coming in, how many emails are coming in. And it's also looking at how long they're lasting. And then it's able to look at things like, you know how many bills we've sent out, how many meter reading requests, we've sent out, how many people are kind of receiving a price change notice or whatever it will be. And it takes all of that data, and then it generates a forecast. And it generates a forecast for each team. So very granular, 10 people in the team. Algorithm creates a forecast for every 15 minutes through the day of how many people need to be available, for example, to answer the phones. And the team leader can tell it look, today I want an average time of two minutes, and it would create a forecast for kind of the staff requirement. Or you flip it around, say today, I've got like seven people available, because there's a couple of ill or morning training. And it will tell you what the average wait time will be by 15 minutes segment. And you can move people around as required. Or say, hey, look, I know you are due on training. But it'd be great to delay that or whatever it will be. So that machine learning is basically... it's quite funny, I sometimes think like, you know, the big fear of AI is we're all going to be working for robots. And in a way, we're beginning to experiment with that here. And it's working okay, so far. Similarly, every single phone call is transcribed live. Now today we would have about 20,000 interactions with customers by phone and email. So every single phone call is transcribed live, every email, every text-based interaction is already in text format. The machine learning algorithms then read the conversation, and they work out what it was about, they work out, the extent to which we've resolved it. And we have live data, literally live on screens, when we're all in the office of what is going on in the business in terms of what the topics are, what the problems are, what do we need to do to resolve them? So now, the interesting thing there is, of course, as a management team, we're not gonna go micromanage with it. But that might tell us that it's time that we improve the software that reads smart metres, or, you know, the teams need more training on some topics. Or, you know, when the teams come and talk to us about the problem with payment adequacy, we kind of understand what's going on. And teams themselves can look at and say, why is our team getting loads of payment adequacy? The other ones aren't? All of that is not based on the kind of subjective... and one of the team have chosen something in a drop down, which, let's face it, when you get a dropdown, saying what's this about, you always end up giving up in the end, right? It's too tedious.

Usually go down to the bottom and you look for other.

Yeah. You know what? That is... I do that myself when I'm helping customers. And, you know, the reality now is you don't have to worry about it, the machine's doing it. We forecast every smart meter we've got at the half hourly level for months ahead, right? Every night, these machine learning algorithms, calculate your forecast for the consumption at the half hourly level, as far as we can go. And they put confidence intervals on that, based on the weather and all these clever things. Now, all of that then gives us, like with no human intervention, tremendous amounts of efficiency in the business, improve service, but drive costs down. To answer your question about cost you know we now sit on 150 billion rows of data, it's going up by 2 billion rows a week. It's all being worked on by machines as well as humans. And yeah, you know, we were benchmarked against a sort of incumbent rival, our data centre costs are 20 times lower than theirs.

Okay, so I think we've reached the point where we kind of get a picture of what you're doing. And it is it is revolutionary. There's no question. global domination, you know, you talked about you've got 0.02% - 0.04% market share globally. What's the right market share? What's the target? You said 100 million by 2027? That even that, 100 million households, that's going to be out of what, a couple of billions? I mean, it's still only 2%? Is that the right number?

I mean, this really just tells you what's wrong with energy, doesn't it? If you look at things like, you know, ecommerce, or even transport through something like Uber, you know, what we've seen is that global platforms have been able to transform the way in which those sectors operate. Because what you're learning in one market, at scale, you can transfer to other markets at scale instantly through global platforms. And I think, you know, in energy, we have to go that way. Today, even the biggest global energy companies typically have a completely different stack. And I don't mean software stack. I mean, a business stack, everything is different from one country to another. And that's holding back the fight against climate change, you know, we will create a cheaper, greener transformation, if the world is able to kind of solve this problem across all the geographies, roughly at the same time. So I guess, you know, it is pathetically under ambitious for us to think about 100 million customers on a single platform. And yet, you know, as a startup, you know, it was embarrassing, saying we're gonna get a million you know, now we're looking at 17 million licenced, by the way 18 months ago, that figure would have been less than a million. So don't get me wrong, I think the exponentials point in the right direction. It is just scary to think about where it leads.

Well, is it scary? Or do you just need to, you know, you, you're the CEO, so you'd have to be the first person to do it. You've come up with this word, elec-tech. Was it elec-tech?

Is it green tech, clean tech, elec-tech, ener-tech, I mean, who knows, right? We need a word.

I've got kind of burnt, you know, using those terms I have, we all did, and I'm not sure what they mean by...but what I was going to say is that you as CEO, surely have to be the first person to start imagining the billion household, you know, energy platform, because the same thing would have been said about the billion user media company or the billion... and then you got Facebook, and then you got Amazon, I don't know how many users they've got, but I'm sure that I'm sure it's in the billions. And certainly... you know, billions and not scary numbers in other sectors. And if that's what you're doing lean in and say, yes, we will have a billion accounts.

Yeah, I think one of the biggest challenges we've got in this is the government and regulatory systems that mean that, you know, instead of having a global internet for energy, every country has a different national grid. I mean, one of the I don't mean the company, I mean, it's sort of a grid and network system, definitely regulated in every way. So it's not just the way you present to consumers, it's different. It's the way you charge for access. They're all trying to solve the problems by putting batteries wherever they think they need them, or whatever it be. And I think, you know, when we started our company, people said, how are you going to tackle the Big Six, you know, the Big Six were the six big incumbents in the UK, some countries by they've got a big one right. Others are nationalised.

Italy, looking at you.

So there's certainly some in Europe, and you know, at least Europe's got some degree of competition. But then, you know, behind the Big Six, were really big companies, like, you know, the oil companies, big oil, but actually behind the big oil companies, you know, even bigger than them is big tech, you know, like, what is it? I guess, just by way, of example, Shell or BP, what's their market cap? Probably around $100 million.

No, it's still more than that, although it's been going down quite considerably. So it's 200 or 300 billion.

So when we started the company, people said, How are you going to tackle the Big Six, they look like big companies, but you know, they were just big by UK standards. You know, if you were lining them all up, you know, bigger than the Big Six were big oil. And I think if you look at Big Oil, you're typically, you know, Shell, BP places like that. Market caps of about 100 billion, or some unit, dollars or pounds. And then behind them big tech, where you've got, you know, companies that are heading towards trillions. And so one of the problems is this kind of view that countries have of what scale is. Where, you know, if we're going to tackle something on a scale of climate change, we've got to go beyond national level solutions. And that really does mean that countries have got to work out how do they open up not to, you know, we've got this kind of terrible thing at the moment where a country, a government will often define a solution, and then put it out to tender. So you're having a different solution tendered each nation. Now, what the private sector does best is actually create solutions through innovation. And we need to open up markets to that.

Is this not just telecoms prior to the telecoms restructuring? This is British Telecom, or, or whatever? 1980. Is it not?

Yeah, exactly. You know, 1980 was a long time ago,

Not innovative. Driven by sort of some misplaced sense of national security or you know, industrial planning or worry about what happens if the system fails and trying to micromanage it? Is this not exactly the same?

Let's look, you used telecoms example, well here in the UK. You're talking 1980, that was 40 years ago. And today, you know, two of my team members have both complained today about four hour waiting times to speak to a telecoms company. One of them is coming to the office, because it's going to be two weeks before a telecoms company is going to come and fix her internet. Right. So telecoms is still broken. Now, you know, that's bad enough. I'm in central London, and my mobile phone doesn't work in half the building I'm in, right. I mean, this is just unacceptable. Now, you know, if we can live with telecoms like that, how are we going to live with an approach to climate change that's so badly implemented Now, don't get me wrong, if we're not careful, that is exactly like (...) the fight against climate change can be analogous. But if we let, if we find ways to let companies use grids, like the internet, where we will find ways to use the infrastructure that governments have put in place, to get more out of it on behalf of their citizens than anyone could imagine, then we can have a much cheaper, much cheaper, faster green revolution. But if instead, each country's defining its own solution, no, it's gonna cost a fortune and take forever.

So the final question, a final topic, you have broken through to, you know, you're sort of nearly 2 million UK households three and a half million that are using the platform or 2 million under the own brand. Three half million on the platform, you've got 15 million around the globe, you've got this vast growth rate. Prime Minister and the Chancellor know you, not only do they know you exist, they've been to visit. So you're right at the centre of things. Are you optimistic? You know, we've got the UK, you know, hosting COP26 in Glasgow at the end of this year, everybody is focused on green finance, are the planets aligning, and are you optimistic? Or is it a sort of a blip? And then, like the telecoms example, in 40 years, we could still have the same nonsense going on, because there wasn't enough faith in the sorts of solutions that you're proposing.

Yeah. I am optimistic, right? Fortunately, economics is aligning with the environmental need, you know, renewable generation is getting cheaper every year. Electric vehicles are getting cheaper every year. I mean, I was looking on Alibaba, and you could buy a five-seat sedan with 250 mile range for $15,000. So the fundamentals are going the direction we need. Now, whether we will, as societies, optimise and get this as cheap and as fast as possible. You know, it's still debatable, but at least, the economics means that we will reduce our climate impact dramatically. I think governments, I think the pandemic in particular has crystallised governments' views that change can happen faster than they ever realised. You know, the way in which supply chains adapted to lockdown, the way in which companies adapted to remote working. Yeah, I think, you know, someone at Microsoft said, you know, six years of change happened in six weeks in many countries. Well, now let's capitalise on the fact we've learned we can do that in one section, let's do it in the fight against climate change. I think COP 26 is kind of, I think it's the first since you know, the sort of Paris commitments are meant to have been implemented, we've begun our plans. So huge from that perspective. I think for the UK, it's the opportunity to double down really on being a country that's been doing well, in the move to renewables. But the real race now is who can do them faster, bigger, cheaper, and to use the telecoms example earlier, you know, it must have been 10-15 years ago that we started hearing, yeah, here in the UK, about, you know, internet speeds in Korea, the fact that in Poland, you could get, you know, same day broadband connection, in the UK, we had to wait weeks and weeks if we were lucky. So, you know, there's still an enormous amount to be done, to get nations competing on how fast and how cheap they can do this. But I think we are seeing a doubling down of political commitment, you know, the change of administration in the United States as well is really going to contribute to that. And then in countries, increasingly, you know, we're seeing politicians competing on who can do a better job of tackling climate change. It's been thrilling to see that here in the UK. So, you know, economics, politics, and proof of disruptive capability might just be aligning.

Greg, was there also part of the disruption that you just looked at the competitive landscape and said that customers are being poorly served, poorly served by the customer, poorly served by the regulator?

I think one of the biggest issues is that we've got a sector that doesn't understand customers, consumers. People repeatedly tell me customers won't do that, consumers won't do this. And it absolutely reminds me of so many other sectors. We talked about mobile phones, in which people thought customers wouldn't want a touchscreen phone with the internet. And I think, far too often, we've got policy that is being led by that kind of thinking. And it suits incumbents to suggest that, you know, you can't engage consumers or that, you know, we need to solve everything, treating them as a sort of dumb off-taker, rather than a participant. So I think with Octopus, part of the reason for building a very consumer centric brand is really to prove that, you know, when consumers join us, they are more likely to kind of embrace new ways of using energy and to think more about the kind of the cost and the climate cost of it.

Very good. Well, Greg, you seem to be saying to the politicians, stay out of our business, but give us the infrastructure, particularly the data infrastructure, and we will finish the job. So on that note, I think we'll leave it there. But thank you very, very much for joining us this evening on Cleaning Up.

Thank you, Michael. It's been tremendous fun. And, you know, I'm on beer number two. So if we had three or four, who knows what we might say.

There'll be time for that. Lovely to see you.

You too.

So that was Greg Jackson, CEO and founder of Octopus Energy, building the energy utility for the climate change generation, offering green and affordable power to 100 million and maybe beyond that, a billion households. My guest next week is Amber Rudd. She was a Conservative MP, minister under Theresa May, and also under Boris Johnson, Secretary of State for Energy and Climate from 2015 to 2016, during which time she championed the UK's move entirely off coal by 2025 and also led the country's negotiations through to the Paris Agreement. Please join me at this time next week for a conversation with Amber Rudd.

Ep32: Greg Jackson 'Building the Billion-Customer Energy Company'

Listen On

Featured Episodes

Recent Episodes