Sam Duncan

Powering $100 billion already, Sam Duncan is using AI to make capitalism care about the climate
September 5, 2023
Leo Nasskau

In 2021, the world's largest ESG rating company admitted that it doesn’t even try to measure the impact of a corporation on the world. Instead, their focus is all about whether the world might mess with a company's profits.

Impact investors exist to make the world a better place, but without accurate data on how a company affects society and the climate, that job is impossible. Without cutting-edge AI tech that can process hundreds of pages per company at one third of the cost, it would still be impossible.

But Sam Duncan is changing that reality. Her platform, Net Purpose, exists to make impact investing totally effortless, by using AI to curate investment-grade data across 5,000 companies and millions of data points. Founded in 2019, the company powers over $100 billion of impact funds, and has a mission to completely change how investors invest, and thus changing the very face of capitalism.

Read below for the full transcript:

Leo Nasskau, Culture3: Hi Sam, thank you so much for coming on to the Culture3 podcast. Over the past four years you've built a platform that provides sustainability data on almost 5,000 companies across hundreds of climate data points powering over $100 billion in sustainable funds. It seems like a product, a system that is almost obvious and completely necessary in today's world, but I have to ask, so why did it take so long for a platform like Net Purpose to come about?

Sam Duncan, Net Purpose: Thanks, Leo. Wonderful to be here. Thanks so much for having me. Yeah, I'm really excited to talk about the topic of Net Purpose and sustainable and impact investing. Why did it take so long for this to come about? I guess we're in this moment of an evolution of capitalism, I would say. About 80% of all capital today is actually committed to investing sustainably. That's about $100 trillion or more. And I think that whole trend is something that's taking all of us a bit by storm.

Personally, I want to buy sustainable products, whether it be food or clothing. Companies are also trying to be more sustainable. So, the general topic of sustainability is coming through all different parts of our life. I think that's why now has been the moment to build a platform like Net Purpose.

There are probably two core reasons from an investor point of view of why it hasn't come about until today. One is the focus on largely financial performance for many, many years from an investment point of view. And that's just now starting to change. And then two is just that when we think about sustainability it's actually quite hard to measure what is and is not sustainable. And you need to really figure out how to measure something and build a platform to provide data on a given metric.

Leo: That makes a lot of sense. And on that topic of measurements, I was doing some pre-reading and I saw the world's largest provider of these sorts of statistics and the ratings that go on top of them. MSCI admitted even as late as December 2021, that it didn't even bother to measure these sorts of things, the actual impact that a company had. And it just focused on the financial side of things and the financial risk. What were your feelings about that news? Because I think from the outside looking in, it feels like that this should have been happening maybe a decade ago or something like that. But it feels like this is, with a lot of potentially a long distance to go. What was your reaction to that revelation?

Sam: Yeah, I'm glad you mentioned it. So, and the topic of ESG is one that is also rapidly evolving, given the focus on this particular theme. My reaction to that specific piece of news is I'm glad that the world is opening up to understand that ESG ratings are not currently measuring social and environmental performance. They were measuring and still are measuring financial risk associated with social and environmental factors.

Post that moment today, we are demarcating ESG, which is focused on financial risk, mostly from sustainability, actually measuring whether a company or a product is having a positive or a negative impact on the environment and on people. Prior to that moment, the concept of ESG, it actually was built decades ago. This whole theme has been in play since the 90s, the 80s. And it started as a concept in project finance and migrated into a concept for broader financial markets. This is probably not that relevant, other than to say the reason I think ESG actually started with a focus on making sure it was protecting the investors in emerging markets environment, they were protecting communities that lived by big projects that were being built.

But I think what happened is that back then, most investors weren't thinking about sustainability as something that they needed to invest in proactively. And so the concept of ESG needed to transform itself into a financial risk concept to be relevant to any investor. And I think in the risk assessment on financial performance, but also it's so qualitative because there's been no way to measure social and environmental performance that it's just a bit confused now. And I think what we're doing today is getting clearer definitions around what is and isn't sustainable, clearer quantitative measurements, so we can build the same rigor on social and environmental performance that we have on financial performance today.

Leo: That's really interesting actually to understand that landscape over the past couple of decades. I can imagine in the debates that you and other people in the industry probably had about it. People thinking we just need to tear up how we measure ESG and other people saying we can use what we have and iterate towards something that's useful. And it feels like you guys have taken the perspective that we can turn ESG and how we think about sustainability into a useful measure by focusing on measuring the right sorts of things. I think I'd love to dig more into the landscape of the sorts of things that we actually have to be measuring and how you go about putting numbers on the climate.

Sam: Yeah, absolutely. I love to kick about on this topic. So, I think that's right. I think we're basically taking ESG as a big broad fuzzy concept that no one really understands and deconstructing it into the components of E, which may include the climate, like climate change, I can talk about how to we're measuring that, water usage, waste, like all the pollution and plastics that are in our oceans, etc. that needs to be cleaned up.

Of three major categories of E, which are more tangible, I think, for people to hold on to and understand than an amalgamated ESG score. And then if you deconstruct S, what we're really talking about on the social side is people, frankly, people either employed, like earning an income or being productively employed Many people are grossly excluded from access to financial services, from access to health care, from affordable housing, etc., which reflect global sustainable development goals. The S in the ESG can be deconstructed into those component parts as well, so it's easier to understand.

G is a separate topic of governance, which actually we have focused a lot on in the outcomes. We're more focused today though on the outcomes like results. Are we achieving them or are we not? Whether we've got governance in place or not.

On the climate side there are very common, agreed practices now with the greenhouse gas protocol and numerous other standards to measure CO2 emissions. There are multiple permutations of that. We measure scope one, scope two, scope three emissions from a company perspective. There's also a new concept of scope four emissions as well, which seeks to quantify the positive contribution of a product in wind power. We want to invest in energy efficiency, we want to invest in electric vehicles, we can actually quantify how many metric tons of CO2 we expect those investments to avoid.

Leo: I suppose the challenge across all four scopes is taking it from the country level down to the company level and analysing across all of those different companies. That feels like a really big challenge for you to navigate.

Sam: Yeah, and that's exactly what we are navigating. And actually, it's deeper. You go from country to company to project sometimes, even to product. And as a consumer, if you wanted to know the CO2 impact of, you know, the piece of clothing that you just bought or the apple that you just ate, you got to get that level of detail to make useful consumption decisions. So that's exactly what we're tackling today. It's that quantification on a company level, on a product level and on an asset level of CO2 so investors can appropriately allocate their capital towards better investments from a carbon perspective.

Leo: And how do you go about actually figuring out what numbers to put on it at a company level? I can imagine one approach would be to spend a lot of time working with a company and diving into how they work. But your challenge is also doing this at scale across thousands of companies, which makes that pretty difficult. So, I'd love to dive into the process.

Sam: Yeah, let's talk about the process because it's a really cool combination of AI and humans today. But before we get there, the substance is exactly that. We have built a platform that covers thousands of companies so that you actually have real access to CO2 data on household names, on listed stocks that every investor is investing in.

We get that data from a couple of places today. The main source is actually company disclosure itself. 99% of listed companies are reporting according to the Global Reporting Initiative, GRI, and in our analysis about 70%+ are actually reporting their emissions. That's enormous. We all talk about how there's no data, how we don't know how to get this data. There actually is data. I think we all have a question of how much confidence we have in the data, but there is data to start to use.

I think we are in a place where we need to start using what we have from companies and not let perfection be the enemy of progress. But that's our core challenge today: to collect and collate all of the disclosure that companies make standardised form so that investors can analyse it consistently across companies. And then we also work to validate what companies are telling us by matching what companies report with a lot of the life cycle assessment data on the emissions of a given activity, like electric vehicles or like building insulation or wind power, solar power, et cetera.

Leo: You mentioned two things that I really want to double click on, standardising and validating, because in sharp contrast to most ways companies report financial data, where there's quite a standard way of doing things. I can imagine for the climate, people are sort of making up their own measures. There have been a lot of efforts to standardize that, but it feels like a little bit of a less standardized place. What are some of the measures that companies do use? And how do you go about actually standardising them and making them useful.

Sam: Yeah, so we basically exist alongside a lot of standard setting bodies and our core purpose is to accelerate their work versus reinventing any standards ourselves. So one of those core bodies on climate is the greenhouse gas protocol, which is an established organization that's been around for a number of years now and they have defined standards for how to measure scope one emissions,  what should be included in that, and literally how you calculate scope one emissions.

Leo: And do most companies follow those sorts of standards?

Sam: They do. Yes. I actually think they do. They do because it's a standard, but I actually think they do on a very micro level. If you're someone that's in the environmental team of a listed company and you need to disclose your scope one emissions, what regular person could figure out how to do that without looking somewhere like a standard to figure out how to calculate it.

I think the uncertainty exists because it's not clear about whether those methodologies and standards have been implemented. effectively. There are numerous ways you can misinterpret or interpret the actual guidance itself. And so there is a big auditing industry building up around this topic as well, just like we have on financial statement audits, where auditors are actually starting to assure the process through which companies have calculated their CO2 emissions.

Leo: That makes sense. So there's a lot of the validating work as well as being done there. But is there more work that you do at Net Purpose around that validating thing?

Sam: Yes, exactly. So what we're starting to do, and we've focused first on products and services, is to really validate what companies are disclosing on CO2 emissions avoided. This is a big theme because a lot of investors are trying to invest in companies that provide positive products and services that will help us achieve global goals. Investors want to invest behind clean energy companies instead of fossil fuel based energy companies. They want to invest behind meat substitutes like Beyond Meat, for example, instead of just meat, et cetera. They want to invest behind recycled plastics instead of just plastics, which have a CO2 footprint as well as a waste footprint. They want to invest behind these solutions.

A lot of companies are telling investors, hey, we're doing this, we're doing this, we're providing this solution, but that calculation is very unstandardised. We've put our muscle really behind cleaning up a lot of the solutions data. to make it clear for investors where you can count on what a company says and where you actually need the data that validates that. On the Net Purpose platform from a solutions point of view, you can find company disclosed data as well as Net Purpose validated data using scientific literature. And we expect that we'll move to do that for the same on the operational emissions side as well.

Leo: That makes sense. And this is where you involve that combination of AI and human analysts and researchers to really perform those tasks. I'd love to explore how that works for you and how you're using tech to really scale up and improve the work that you do.

Sam: Yeah, absolutely. We use AI and humans in combination across all of our data collection efforts. It's actually a really exciting time for AI and sustainability right now. I mean, for AI in general, there's obviously rapid advances happening across the world in the strength of the language models, essentially, that we've been using in our process for the last four years, but are accelerating quite quickly in terms of their sophistication.

Today, we run a data extraction process that includes humans and machines. We call it a human in the loop process. And what that looks like, very practically, is our task is to extract quantitative facts, so say CO2 emissions out of a company's public filing. So let's say we're trying to grab the emissions data for Google or Amazon out of their public filing. The way this started was around my dining table with like a couple of analysts doing this manually and then we just basically charted out the tasks it takes to actually do that manually which is an enormous lift.

Leo: Is it just you have to go through an enormous disclosure report, which I can imagine could be tens or, for a big company, hundreds, of pages long, and actually find the numbers that you need? Is that basically it?

Sam: That's basically it. I mean, a sustainability report might be 200 pages long in a PDF. And if you literally, I mean, what a manual analyst is doing there today is like Ctrl+F-ing through a document to find the number. But if you actually Ctrl+F for ‘emissions’, the company probably mentions it like 10,000 times in the document, which is also why you can't use just language, right? Because if you actually assessed a company's CO2 emissions based on how many times they mentioned the word ‘emissions’ in their report, you would have a lot of companies looking really great when actually their emissions profile is not so great.

So our task is not to attract, extract the language, it's to extract the quantitative facts, which also exist in a table form. We need to extract the fact, the time period, the metric name, the unit of the metric, all at once, and do that over a time series of five plus years. So we're actually using more than AI: Tableau fact extraction, OCR recognition, et cetera. What we've just done over the last year is automate every step in that process that we possibly can, from sitting around a dining table to now quite a productionised process with software and tens of analysts around the world.

Leo: So, what is it that you use the human who's in the loop to do? A part of it is training, and a part of it is, and I imagine also part of it is, as you mentioned, validating, and potentially is there also another part of it where there are bits that the model just can't handle and it has a really low confidence about a certain figure that it's come up with, or the company is using a standard that's really vague. Where does the human fit in?

Sam: Yeah, great question. I mean, it actually fits in all the way through. The broad ambition is to automate this entirely. What we've done is take a very pragmatic approach. Our core barometers of the quality of our product relate to the quality of the data. We really take no chances. This needs to be investment grade data and every single fact needs to be accurate and completely transparently sourced back to the filing it came from.

The thing is, AI is actually not really originally trained for that, right? AI runs in a black box type model where it figures stuff out, but you don't always know how it's figured out that that's actually the right thing. The early models were not really that fit for this task. What we did automate is the process of finding where the fact is in the document.

And then we've started to automate the actual extraction of that as well. But the fact and also checking that the extraction is correct in each stage. I think the new innovations will help automate the first round of extraction much more than the models could before. We'll basically just see a progression. We've automated about 55% of the process already and we expect that to dramatically increase over the next months and years.

Leo: That makes a lot of sense and yes, that, I mean, the impact of these tools has been enormous in terms of just the speed in which you can implement them as well. I'm really, so did you start in 2019 sort of automating some of it? I'm curious as to how quick it's taken you to automate that 55% of the process and what that has enabled you to do in terms of the number of companies that you've been able to cover, for example, and maybe where that trend is heading as well.

Sam: Yeah, absolutely. So yes, the mastermind behind this entire thing is our CTO, Alex Woodham, who's been working on AI his entire career, really. So yes, he joined in 2020. That's the duration of which we've been working on this problem.

Leo: What did you do before that time when he joined in 2020 out of interest? Was it all manual?

Sam: Yeah. The way that the process works is you can't just throw a machine at like 50,000 public filings and say 'go get it'. It doesn't work. That's the issue. If it did, we would be in a very different place. What we needed to do before that was actually establish what we need to collect. Is it available? Like is it actually disclosed? What permutations of the metric?

So, if you think of scope one emissions for example: in a document that could be written as scope one emissions, total scope one emissions, million metric thousand scope one emissions. The variation in how companies put that in the file is dramatic and you need to capture every permutation of the machine to know exactly what it's searching for.

Now, the sophistication of the machine improves the more permutations you have and then it learns basically, that's the whole point of AI, learns that this word might, you that word because it's often associated with something in the document, but you need to build that training set for the machine to learn on. So in 2019 we started to build the training set of a very small sample with defining the metrics and getting all of the permutations in for a very small sample of companies and from there we're able to build intelligence on top of that to learn where machines can accelerate from there.

Leo: So really this is a story of, it was basically virtually impossible to get data at the right level of quality, that investment grade quality data, which is the quality of data that you require to really influence the amount of money that you're influencing. And that is what is critical for really having an impact on this problem. None of that was really possible without the AI and the other sorts of data science technology that really you've only been able to do. but people have only been able to use it for the past five years or so.

Sam: Yeah, I mean it was possible but it was manual.

Leo: Mm-hmm.

Sam: Yeah. And manual is not scalable. So it's, but there are a lot of data providers around the world that use manual labor to do this problem in the background. And I will say also my husband is, uh, in tech himself at one of the very large tech companies. And I think the integration of tech and humans is an established thing across the entire universe. I mean, even companies like Meta have thousands of content checkers to pick up every single inflection of every single topic that a human would. This synergy between humans and AI I think is a well-established thing in every tech story around the world. And in this particular case, it's very cool to actually think about the process and how it goes. And yeah, we do similarly expect exponential increases in productivity over the next couple of years.

Leo: Yeah, I think it's just really exciting to reflect on that because the fact that it's a technology that is... the capabilities that technology has given you in combination with analytics is relatively new. It suggests that there's so much further to go, as you were just mentioning. But you were talking a little bit about the impact of that automating 55% of the process in terms of the number of companies that you've covered. What are the trends that you're thinking about in terms of the impact of that automation number of companies and the number of data points that you can cover.

Sam: Yeah, I mean, well, so the direct impact of automation on our side is that the cost of acquiring the data dramatically reduces and therefore the speed at which you can acquire the data dramatically increases. We don't have a straight line yet on what we expect the coverage impact to be, but our ambition is to cover every company in the world with this data and to feed that information effortlessly to investors and to companies and to consumers.

I think the interesting way to think about it is if that's our ambition and we didn't have machines, it's going to take a very long time to do that for every company in the world. If we do and if we can automate 90% of this process, we can instantly accelerate coverage across the entire universe of what we're seeking to cover.

In listed markets, I think the other challenge that we haven't talked about for any investor that wants to allocate capital towards sustainable goals is that a large amount of capital flows into private markets, private companies don't publicly disclose their environmental or their social performance. And so that's actually quite a different data challenge where we'd need to engage with the investors or the companies in order to effortlessly help them contribute their data and measure that so that their investors can measure the impact of their fund.

Leo: Yes, of course. I want to, I'd love to put some numbers on this. How much it would cost when it was entirely manual versus the process you have today to gather all of the data that you might need for a company that people might know. Like if you were to look at Google or someone, how much would it have cost to put together all this data on the Net Purpose platform when everything is manual versus how it is today?

Sam: Yeah, I mean, so the short answer is we measure cost on a datapoint basis. Let's do the math together. We started at near $3 for a data point and it's now closer to a dollar.

Leo: So down by two thirds.

Sam: Yeah.

Leo: And how many data points would you put on a particular company?

Sam: Well, every company, we collect 270 data points over at least three years; we start with three years and then we go back five years.

Leo: So $800 per year of data down to, what, $200 ($270). Which is a huge saving.

Sam: Massive.

Leo: Incredible.

Sam: Rough numbers, but that's exactly what it is. It's incredible. The charts kind of...

Leo: ...fall off a cliff. You mentioned getting that data available to consumers as well. As you were, as we've been talking, I thought something that would just be really interesting, but also just have a lot of virality benefits is just putting a quick site up where anyone can type in like a company that they know, like Google or something, and they could just dive into their emissions. People would love that. It would just be a fun thing to do, but have you thought about it? You must have thought about it.

Sam: Yes, I love this topic because I totally agree. That's what I want. Right? Like every time I buy something, I'm like, okay, how sustainable is this company? What would be even cooler is if the data fed through the online platforms that I buy stuff on, like imagine if Instagram could tag their ads and tell me, okay, this is a sustainable company. And I think companies thinking about that now, we have a data set that would help them do that actually, you know, listed markets, which is 80% of global produce, probably.

So yeah, that's what gets us really excited. I think that the benefits and the use cases for this data are infinitely more so than financial data. And financial data, by comparison, that's a $30 billion industry that exists to feed financial facts to investors, essentially, so that they can justify, like, visions. The thing with financial data is, when I buy a sweater, I don’t think, ‘what's the earnings per share of the company I'm buying this sweater from?’ But I do think, when I'm thinking about buying a sweater, about sustainability: ‘are they environmentally conscious? Are they not?’

I think the nature of the data itself opens up so many use cases that we are definitely committed to supporting. I think when you start a company though, you need to focus, can't do everything at once. just on a stage of our journey where we focus first on investors because they are the ones that need to mobilise trillions to achieve the Sustainable Development Goals.

Leo: Yes, yeah, the focus is, I think, obviously very important; potentially one of the underrated qualities of a successful founder. However, I think it would be very fun.

Sam: I couldn't agree more.

Leo: But nonetheless; I want to dive into your personal story. How did it start for you? I know you have an established career in the worlds of impact investing, but why did now feel, why did 2019 feel like the right time? What was the process leading you to Net Purpose?

Sam: Yeah, so I would categorise my career slash life journey as... two bookends really. I started my career in traditional investment banking at Goldman Sachs, so probably couldn't get more like ‘profit, profit’. I also started right at the peak of the financial crisis, well actually just before the peak. I was one of those lucky people that within 12 months saw the boom and the bust all happen before my eyes.

Goldman's an amazing institution. I mean, full of highly committed, highly intelligent and dedicated people, really rigorous, influencing multi-billion-dollar transactions. And I had the opportunity to work on a couple of those big deals in Australia which were some of the last consolidation plays of the healthcare sector in Australia, which I really enjoyed.

I think the thing for me though is that, healthcare, it was very obvious that all we were doing was optimising our models around earnings per share, basically profit to shareholders, but there were so many other things happening as a result of the deals we were doing that were impacting access to healthcare in Australia, the cost of healthcare in Australia, the employment opportunities for doctors and for nurses, like rural versus urban distribution of medical centres that just never factored into any of our that also always felt a bit half-baked, right? It wasn't the full picture.

So that's the foundation. I think that and the financial crisis got me thinking about the model of capitalism. From there, I spent really a decade investing in impact. I moved to Peru from Goldman and ran a small microcredit program. When I say run, I really got the chance to hang out with amazing women entrepreneurs in Peru who were using $20 loans to start their business and to really make a big change for their families. It was honestly inspiring. And I just thought, wow, I mean, the power of finance is social. We just need to find a way to mobilize more in this direction and represent the social impact that it's having so that investors can intelligently think about what they really want to invest in.

I joined a company called LeapFrog Investments which did that on a larger scale. LeapFrog was launched by President Bill Clinton and is an impact investing leader. LeapFrog invests in companies in emerging markets, serving a growing, emerging consumer segment with financial services and health care. I was lucky enough to join the team and support in measuring the impact of that.

Those two bookends, I guess, Goldman and then impact investing, really identified to me this movement of investors that want to invest for social and financial returns and how that was catalysing a much broader movement towards sustainable investing around the world.

What was missing for me in the setup at LeapFrog was data. At Goldman, I had Bloomberg, FactSet, Refinitiv, CapIQ. I used all those tools to grab facts on the financial performance of companies. And my job was to analyse that information, not ever collect it from scratch, you know?

At LeapFrog, I looked around the sustainable investing industry, everyone was trying to measure their sustainable performance, but most people at the time were literally punching numbers from a PDF file manually into an Excel spreadsheet and then shipping that over to someone else. And it was just like, hold on, this is not going to scale. This is not going to work. It became very obvious, having been in the weeds of it for six years at LeapFrog, that someone needs to build the data infrastructure to power trillions of dollars of sustainable funds. We have data infrastructure to power trillions of dollars of financial flows; someone has to do that fast otherwise we're going to miss the moment. We at Net Purpose sought to work on that problem and that's exactly what we're building today.

Leo: And here you are. Do you think there's a future in which AI is, or other forms of technology like you mentioned OCR and others is doing, how high does that 55% go in terms of automating the process or the entirety of that process?

Sam: Yeah, so I think it actually goes quite far, but I think there's two parts to the process. I think there's, what do we measure? Like that's still landing regulations much more much clearer now than it was four years ago. And in fact, it's coming into play across Europe, the UK, the US, Australia will all have regulation around basically an evolution of accounting standards to incorporate environmental and social factors.

The ‘what to measure’ is still a bit of an intellectual human exercise with standards. Once we land that, what to measure, then it's about companies disclosing data in line with that. So part of the data challenge is the companies themselves producing all of this information. So their internal systems need to change to be able to do that efficiently and effectively.

And then for us today as a company, assembling all that information with AI will become seamless as AI improves, but also as the data itself gets more standardised. I definitely think we're in a world where we can aggregate, we will be able to aggregate this effectively in a few years.

But I think that the standards will likely continue to evolve. I mean, there's been a lot done on the environmental side, in particular climate, but we're way away from standards on water, on waste, on social inclusion, social themes. So there's still quite a lot of work to do. we're helping clean up as well is Which is very human-based and I don't think will go away completely But that's a better use of our human brains than you know getting numbers out of PDF files

Leo: Yes, a long distance travelled, but a long road to go. To wrap up, I'm curious about which challenges you're going to be focusing on over the next couple of years. You mentioned focusing on private markets and, well, I was going to say smaller companies, but consider Stripe, for example; you have startups, private companies, that are tens of billions of dollars.

I'm also thinking about the blockchain (specifically Bitcoin) ecosystem, where you don't know which companies or which organisations are active. And then you also mentioned the other types of challenges about the social challenges as well. What things do you really want to be prioritising over the next couple of years?

Sam: Yeah, great questions, Leo. So, for us, we think about our product evolution in a couple of buckets. I think one is just the depth of what we're able to provide on an existing universe of companies. So, we cover nearly 5,000 companies today. That's 100% of the S&P 500, 99% of the MSCI world and equity by value. In that universe, we are going deeper to help investors align with evolving standards.

So that's new ways of visualising what is and isn't sustainable, making data more comparable as disclosure improves, and adding themes like biodiversity for example which is a new theme that also is evolving and not really well defined. So that's increasing the depth and the quality of the product as it relates to an existing universe but then I think the biggest opportunity is to expand into different asset classes we would say.

Our next adjacency will be impact and sustainability bonds, which is actually, there are investments on an asset basis, not on an entity basis. So, and it's a big market of companies that are issuing these specific bonds to target projects and assets, and a big market for investors that want to invest in positive impact, because you can actually back these projects to accelerate the development of, for example, the energy transition.

Beyond that, private markets are a big topic for us. I also was in private markets at LeapFrog and I still want to build a tool for my former self! That will come. It's also a massive area of funds flows. There's a lot of companies that are choosing to remain private. You mentioned a few just now and so I think this transition in capital markets today towards private is something we want to be a part of. And then I do think there is crypto and other things; it's probably a topic for a separate chat. It's not immediately on our radar, but it is a big topic, big energy use, big climate impacts, big evolving asset class. It's probably not on our 2023 vision, but it'll remain a topic of one to watch.

Leo: Awesome. And to close out and to bring everything together, you mentioned the challenges there over the next couple of years. If everything goes wonderfully and if Net Purpose continues to be wildly successful, how do you want the world to be different? Investors, consumers, and of course the companies themselves, what do you want your dent in the world to be?

Sam: We are strongly of the view that in the future, every investor will be a sustainable investor, just like every investor today is a financial investor. We don't think everyone will be top quartile sustainability performance, but we think that the factors of social and environmental performance will be incorporated into investment decisions, just like financial performance.

If we are successful, we will help that happen really quickly. We are on a mission to mobilise trillions of dollars to achieve the sustainable development goals. We currently power over $100 billion in sustainable funds. And our kind of success criteria is if every investment manager in the world measures their social and environmental performance, we just think that if they do that, we will all be in a very different place and we'll be allocating capital with an entirely new lens and helping to change the code of capitalism, I guess, to a more sustainable form. 

Leo: Wonderful. Sam Duncan, thank you so much for coming on.

Sam: Thank you, Leo. Pleasure.

Links to the author's Culture3 page
Hosted by
Leo Nasskau
Links to Curation, by Culture3, on Spotify
More Podcasts
C3 Spotify

Leo is part of the founding team at Culture3. An award-winning editor, he is also the Chair of UniReach, an EdTech non–profit he founded whilst studying at the University of Oxford. He writes about technology, change, and culture.

Related episodes
See all

“Success in creating AI would be the biggest event in human history, but it might be the last” — Stephen Hawking

Stay up to date with Culture3 and receive our best on tech, creativity, and hope.
Thank you.
Something went wrong – try again?