Episode 81: Enabling Businesses with Climate Data with Gopal Erinjippurath
Melissa Perri welcomes Gopal Erinjippurath to this episode of the Product Thinking Podcast. Gopal is the co-founder, CTO and Head of Product at Sust Global, a company whose mission is to “develop data-driven products that enable every business decision to be climate informed so that humanity can thrive in a changing planet.” Gopal joins Melissa to discuss climate sustainability and why climate data is proving to be valuable to all kinds of organizations, how he tested and iterated to build this complex data product, how he’s de-risking bets in a rapidly evolving market, the balance of being mission driven and commercially minded, and the importance of making product thinking part of an organization’s DNA.
Subscribe via your favorite platform:
Here are some key points you’ll hear Melissa and Gopal talk about:
Gopal talks about his professional background, how he got into climate sustainability, and what led him to found his company, Sust Global. [1:29]
Melissa asks Gopal what type of companies purchase climate data products and services and how they use them in a professional capacity.
Your long-term strategy should include holding financial instruments that directly correlate to tangible assets. There are several physical climate risks related to these assets, so ask targeted questions about the climate to protect your assets. [5:26]
Gopal shares how he was inspired to go into the business of climate-related data and insights. [8:29]
Melissa asks how Sust Global tested their climate-based data product. Gopal explains that the first step was “to start with the outcome rather than the outputs and work backward from there.” Creating mockups of the data-based outcome and testing them with the early set of gated customers can provide valuable feedback. [10:42]
Melissa asks Gopal how Sust Global ensures that their climate data product is of the highest quality. Gopal suggests that the best approach is to “sandbox the data capability into an area that one customer cares about and wants to decide on, and then provide them with that data in the simplest form so they can try it and use it for the first time.” [14:22]
Your data should fit three criteria:
temporal - how fresh your database and data product is
geographic - dimensionality of your dataset, how it's partitioned before it is handed to customers, and what interfaces there are
the business problem [16:26]
Gopal highlights the challenges Sust Global faced when creating their product. [19:06]
“You must enable your team to stay on top of things and…to fundamentally have product thinking be part of the DNA of your team,” Gopal says. [20:19]
Gopal looks at capacity building, strategy and execution when he is building a data-based product team. [22:07]
Climate change is a space where it is possible to stay mission-aligned and also be highly commercially minded, due to the rising importance of ESG and climate change initiatives. [24:54]
Resources
Gopal Erinjippurath on LinkedIn
Sust Global | LinkedIn | Twitter | Instagram
Transcript:
Melissa:
Hello, and welcome to another episode of the product thinking podcast. Today. We're gonna talk a little bit about climate sustainability and I'm joined by Gopel Erinjippurath who is a CTO and the head of product at Sust global. He's also the co-founder of Sust global. So welcome Gopal.
Gopal:
Thanks. Delighted to be here. Thanks for having me.
Melissa:
Yeah. So can you tell us a little bit about your career so far? How'd you get into climate sustainability and what made you wanna start this company?
Gopal:
Absolutely. So I'm a electrical engineer, turned geo data scientist. I spent earlier years of my career working on multimedia products very far away from climate, most figuratively, as well as literally. And over the course of the last eight years, I started transitioning more into learning about large scale spatial temporal data sets and applying climate related and environmental data transformation techniques on large scale data sets using machine learning and computer vision. And through a certain bit sequence of events ended up leading insights and analytics products at planet labs that operates the world's largest constellation of earth observation satellites that exist. And through that journey over the coast of three years from 2017 and 2020, we shipped series of products around analytics and insights that were looking at the treasure tool of earth observation, sensing from satellite data into refined, clean, analytical inputs, into workflows that deal with environmental decision making, as well as defense and intelligence decision making and mapping and navigation workflows.
And I learned in that process of shipping the product lines of planet analytics and some other derivatives, I got more interested in climate because that's one of the bigger pieces of the environment that affects all of our lives and saw those, you know, three distinct communities, the community of earth observation. That's looking at observational sensing and data products that are very specific to the historic and catalog event, driven information around climate, and then the community of climate modeling, which is largely in the scientific atmospheric as well as oceanographic modeling environment in the space of earth system sciences. And they look at more forward looking projections around what happens in across the world in terms of different bombing scenarios, across different social economic pathways. And then, uh, the emergent financial ecosystem that is looking at ESG and sustainable financial instruments and saw a void, and the absence of a, of bridges connecting those three communities. And to great extent, that's, what's evolved into the product line via developing at Sust. So at Sust global, our mission is to develop data driven products that enable every business position to be climate informed so that humanity can thrive in a changing planet. So that's kind of the inspiration in which we started in today. We've stood up a product line around that. So it's been long journey into the product landscape of climate data products.
Melissa:
What types of companies would be buying this climate data and how might they be using it?
Gopal:
Yeah, so today we are seeing a growing number of new businesses, as well as, uh, financial institutions interested in, uh, climate data primarily because of recommended disclosure frameworks that quickly becoming close to standard, even though not mandated just yet that's coming. And secondly, there's a increasing desire across the institutional investors, as well as retail investors to invest in a climate consciously because that's on top of mind across the world, be it, uh, heat wave pro area like India, or be it the financial markets in the us or Europe. So you just seeing a lot of emergent awareness specifically, we are selling into ESG advisory businesses that are interested in having the best validated climate related data sets on physical risk exposure. We are also seeing an emergent market in financial institutions, a insurance banking and asset management.
Melissa:
Okay. So in the financial or the insurance companies, let's say you're like a mid-level manager. What types of questions are you asking about climate and how are they using it in their everyday decisions?
Gopal:
So primarily you are looking as a portfolio manager, looking at building either a long shot head strategy, or, uh, you're looking at a long term holding of financial instruments that have direct correlation to physical, tangible assets on the ground. There is a multitude of different physical climate risks that those assets have exposure to. And they can have a direct impact at the bottom line of those financial instruments that the asset managers holding. What they're asking is I hold these assets on my books. I hold these assets over a longer period of time. What is the extent of the exposure? What is the risk? What's the level of severity and how vulnerable am I as a hold of these assets for the long term to climate related risk diversification is on top of their mind. And the thinking about climate diversification as in other dimension of factor risk that they consider.
Melissa:
Okay. So like if I am a portfolio manager for home insurance, let's say, and I've got a ton of homes sitting on the beach in an area where, you know, the water's rising and it might not be on the beach anymore. It might be in the water in a couple years. Those are the types of things I wanna pull and look at to make sure that I'm not over invested or overexposed in certain assets in that area.
Gopal:
That's right. So if you were to look at the gamut of climate induced physical risk, it goes across acute risks, like wildfires, inland floods, tropical cyclones, and coastal floods and chronic hazards like heat bases, sea level rise and water, stress, or drought, and, you know, different businesses have different exposure to each of these. If you're like a refinery or metal mining or bottling plant, you have extreme dependence on water availability and water supply. If you're agricultural commodities oriented player, you are relying on water supply. So the chronic hazards have a lot of impact on your operational costs, as well as, uh, your operational throughput and long term value. If you were looking at real estate, be it, you know, the example you cited or, uh, your investment into a REIT or a REOC or a pool of mortgage back securities, all of them have exposure to the acute risk wildfires, floods, and cyclones, and that directly impact the value in terms of the, either the yield or the direct valuation of the properties over the course of time. So that's where there's a direct correlation between the holding the assets on the ground and physical risk data.
Melissa:
So what made you want to solve this problem specifically? How do you kind of observe it in your, you know, your career and your passing and decide that's what I wanna get into.
Gopal:
So, you know, there's this always this proverbial thinking around, you know, the product manager as the one who's like looking at the mirror and making decisions on his or her of their own, but there's a version of that, not to make that the best case example, but there's a version of that. That's true with me. So I live in San Francisco and over the last 10 years, we've seen an increasing incidence of extreme wildfire events happening within the 300 kilo radius of the city. In 2020, we saw the largest incidents of wildfires recorded in the Pacific Northwest in recorded history. So we're not looking at just one or two years, but if you look at recorded history in 2020, you saw the, the largest incidents. And then my original home city in India is located in the state of. And we saw two, one in 100 year floods at that level of severity happen over three years.
And that's statistically not common. And I felt like there is a unique opportunity to use the emergence of earth, observation capabilities, space, payloads, and space that have been launched over the last 10 years that are beaming data downstream to ground stations, to be bundled with climate related data, towards providing more refined insights around exposure over the near term, which would be one to 10 years. And over the long term would be, which would be all the way up to 30 years. And that information can enable the allocation of capital across the world, be it real estate, or be it in commodities, or be it in new sustainable financial instruments in a climate conscious way. And I saw less of that happening five years ago. I'm seeing more of that happen right now, but five years ago, that was less and, uh, saw the opportunity to be a part of that revolution to be part of creating foundational products that are used across the emergent climate economy, where every decision is made in a climate informed manner.
Melissa:
Great. So when you were thinking about solving this problem, right, you're, you're building a data product. I've heard a lot of people ask questions about how do you MVP a data product, right? You have to go out, get out all of this climate data, all these sources, what did you do to test it and see how your customers reacted to it?
Gopal:
Great question, and I feel like, you know, some of this was experienced and I picked up when I was at planet lab. So I was hired into planet by the CTO of planet labs in 2017, primarily because they had just this problem. They had like a treasure trove of data that they were sitting on, but there were many communities that lacked the geospatial expertise to work on dense semi-structured spatial data that is in the terabyte scale. So while we were building all the capability of transforming that into simplified time series of change or computer vision feature driven features as a service, which you're identifying specific objects in a region of interest over the course of time and tracking their presence or absence, we kind of needed to go through that journey. So the first step there that we followed was really focusing on MVPs and identifying before you actually bake, uh, the cake, what are the ingredients that are more, most desirable to the folks who are gonna consume the cake? And I think the first step there is you need to build out your APIs and your full web stack till you can prove out the value proposition of the data. So start with that, start with the outcome rather than the outputs and work backwards from there. And you can create mockups of the outcomes and test that with a early set of, uh, gated customers who can provide you with trustable feedback
Melissa:
When you're doing the mockups too, for things like APIs, right? They're they don't have as many they're not interface, right? Like they're not like a clickable prototype that you can go through. What types of things are telling you that yes, I'm getting this right. Or I'm reducing my risk that I might fail when I'm actually trying to develop that API.
Gopal:
I would say you can still go through with outcomes. So if you were to think about APIs as like the foundation towards spin clients that have visual components, you can start in two points and I'll start with the visual element first. So you could mock up visual elements, which have the case of spatial data products, which have mapping interfaces, which have <inaudible> interfaces and show what the outcomes can look like for someone who's making decisions with the data you're serving, depending on what choices you make and what you see as dominant signals of interest, you can then work backwards to the API endpoints and your data model. So I would say that customer facing view that drives decision making on the data and the model structure is often very powerful. It's a pattern that I've used a fair bit in terms of scoping and defining API products in the past. The second one you can do is work with a few data scientists or developers who would be direct users of the API because regular consumers are in our retail users or analysts often not the right users for an API. It's normally a developer. So working with developers or proof, the MVP of an API, just in terms of structure and in terms of the query parameters. And in terms of the data model can go a long way in terms of building your own internal confidence that your MVP's the right one to invest engineering effort into.
Melissa:
When we're looking at other factors too, like besides just APIs, when we're building data products, a lot of it is also the quality of the data. If you're looking at one of the things for the users that like, it must be the highest quality data, how do you like test that early on to make sure that they like understand the tide quality? They get that even if it's just like a mockup or an experiment,
Gopal:
You know, one approach, I'm sure there are many approaches. The approach that I've I've used to fair bit is really sandbox the data capability into an area that, that one customer, that cohort of customers really cares about and wants to make a decision with and then provide them that data in the simplest lowest level of effort way so they can try it and use it for the first time. So if you can get to that, that's actually a brilliant, very important milestone because the sooner you can get to that, the sooner you can get on the learning journey of whether the data is high quality and with data products, I think data quality is primary, and it's not an issue till it becomes an issue. If it were to become an issue you wanted to happen sooner rather than data. So that's been the journey we followed, even as we prototype and build out this flagship capability we have at Sust global around climate where good physical risk data and getting that either as in the simplest format, be it, if it tab data in CSV file formats, or if it is, uh, spatial data, having it in adjacent or geo T formats can actually just make it bit easy for anyone working with that data to load up into their own workflows.
So you are taking away the friction and the additional effort of standing up an API and the delivery related engineering effort and making it directly a data to data value proposition. So if you can prove that out then it's about delivery mechanisms via dashboard or API or direct file transfer mechanisms towards getting that data over to them. And now we have enough expertise in the community towards being able to solve that problem. Oftentimes the data's the new thing, and that's the real product. And that's what you wanna prove out first,
Melissa:
When data is your real product, how do you build a moat around it and make sure nobody else can just go get that data and start supplying it? Like, what can you do when you're building a analytics platform or a platform where it's like, I sell data to make sure that you continue to differentiate and win.
Gopal:
That's a great question. And I feel that's still like a unsolved the problem largely in data products, which is we from the engineering and product standpoint. And as we largely the realm of business models. So there's an angle in, there's a dimension I'm trying to flesh out in greater detail in an upcoming blog post that I'm writing up happy to share that with you and your audience in the coming weeks. But the idea, the few dimensions to definitely explore in detail, uh, what's the freshness of the data. So if you are delivering a very static data set, then your data product has some inherent limitations. So the biggest challenge, so think about like a data set of US counties, make it super simple. The county boundaries don't change very often. So if you, you were to package that and sell it to customers, they could in theory, buy it, share it across their teams and buy once, share it across their teams and not have to talk to you again, it's not a very inherently sticky capability.
So I think the freshness and the temporal element is definitely worth thinking through. The second bit is access patterns. How does someone actually access your data? So if it is directly sending them the whole data set through the API or through, uh, visual interface, then that has some limitations, but when you're dealing with global data sets like we do inherently, no one's gonna want all the global data at once. They need specific regions. So you're sticky because of the geographic footprint you cover. And then the third dimension is the business model. Are you licensing the data based on certain set of parameters and how does that work in the real world across your customer base today and in the future? So those are the three dimensions. I would look at temporal being first, which is how fresh is your data set and your data product, the, uh, geographic, or I would say dimensionality of your data set and how that is partitioned before it is, uh, handed to customers and what the interfaces there are. And thirdly, like the business model.
Melissa:
Great. That's a really good advice for that. When you are building Sust global too, what did you find the most challenging part about building a product like this?
Gopal:
The challenges you faced so far are typical of creating a new product in an emergent space. So the first aspect is identifying, okay, where does the product have attraction today? And that's often very clear based on the, in mouth and based on the stickiness of your outlook. And then where are the new venues in which your data, your capability can have traction in the near term and the long term, and in the near term, sometimes it might be very limited, but the mid to long term, it could be emerge. And then reading those signals, that's more of an art, less of a science. It's more of testing early hypotheses rather than a highly quantitative, highly, uh, programmed approach. So I would say having that mix of data driven decision making when there's very little data has been one of the challenging things as we take a new capability like this into the market, which is rapidly evolving.
Melissa:
De-risk some of those bets, you know, you've got tons of stuff going on with climate right now, and policy changes. And a lot of that's like out of your control, what types of things are you doing? And like, especially as a heads of product to, you know, stay ahead of that, make sure that you're not like caught by surprise. What's your, what's your day to day look like in those situations?
Gopal:
So primarily it's having, I think the best way as like a leader on the product side, uh, you could enable your team to stay on top of things is to enable your organization to think product and, and to think fundamentally have product thinking, be part of the DNA of the team you put together. So we, uh, organized in a way where we have three squads that work fluidly across engineering and product access global. We have a data squad, which is central to the new data modeling capabilities that we building. We have a platform squad that's very central to the delivery, uh, content delivery mechanisms that we are putting in place. And then we have a product team that's looking at, how do we bring this product with agility into new spaces that we are building traction in? And I would say the best advice I can give is make product thinking part of the DNA of your organization, make everyone realize that overengineering is like a path to failure in early ventures. It's proving out your capability in, uh, live active usage, driven environments that actually brings your product to life for the near term and the midterm. And if there's no near term and midterm, there is no long term. So having that thinking is pretty critical. And then along the way, inspiring the growth in new markets and making that feedback cycle and learning part of how your team thinks about evolving the product in other dimension that as a leader, you can exercise.
Melissa:
So when you're talking about product thinking, it's funny, cuz that was literally what my talk was on. <laugh> a couple hours ago. What types of qualities are you looking for in the people that you hire to see that they understand product thinking? Or what are you doing with in your organization as a leader to help them understand what that actually means?
Gopal:
Yeah. I feel like hiring on, uh, product management is a nontrivial task. I feel like to some extent, one can say that it's an established domain, but in some ways data, product management is relatively new. At least the way I like to think about it. You know, if you don't look at data, product management, just product management in general, I would say there are three dimensions I care about. One is capacity building, which is, can you inspire your engineering team and adjacent teams to actually work effectively on the most critical items to drive the business forward? So can you help recruiting, can you help bring in the smartest and the brightest into your small team and when the odds are stacked against you in the job environment and the recruiting environment, the second dimension is strategy. Can you think beyond just features into broader strategy on how the product works in the market and how your product goes from one to many in terms of a lineup?
So that's the roadmap level thinking ecosystem level thinking and some of the junior product managers might just be building that out, but seeing some signs of that is always very, uh, exciting to me when I'm in the, in the recruiting mode. And then the third bit is execution, which is, can you drive the day to day delivery of products and features and how well can you do that? How effectively can you do that with a small team at your disposal that you inspire rather than work with authority or so those are the things I look at in product manager. And then of course, mission alignment is important. Their desire and purpose is upon why they're working in products is important. So those are all good things for me to get to know better.
Melissa:
Yeah. And for you, you are actually like a pretty mission driven company, right. And dealing with really hard problems. What's it like to run a company that is so mission aligned and make sure that you are continuing to fight the good fight and not just get completely sidetracked by things that might be easy or super profitable. Like how do you make sure that you are doing stuff that really aligns to your mission and how do you create that mission and that camaraderie around it.
Gopal:
We very fortunate to be operating in a space where it is possible to be mission aligned and at the same time be highly commercially minded. So if you look at the next 20 to 30 years, if your sustainable capital allocation has the do good element, for sure, as a dominant vector, it also has the immense potential for huge return. And we are enabling our customers, our partners on that journey. So that's one of the things that we stand apart on, like with our mission and our, our purpose. So with the mission of developing data driven products that enable every business decision to be climate informed, we're kind of natively at the brand diagram of financial return for our team and our investors and, uh, huge benefits for our customers and then the intersection of social and net positive impact. So I feel when it comes to making choices in product development, as we all know as product managers, you're constantly making trade offs and prioritizing, I always seek the sweet spot in that ven diagram of these two dimensions, which is financial return for our customers and the potential for social impact and growth.
Melissa:
Have you run into any situations where you feel like those things sometimes get at odds with each other and then how do you make sure you choose the right one? What's your thought process around which way do we go? <laugh>
Gopal:
Yeah, I feel we do run into those issues or some choices we have to make, I would say less issues. And sometimes I feel making a suboptimal choice today as like a small business is better than making no decision and allowing it to roll. So sometimes, you know, one of the trades we are to make is do we support a financially lucrative business opportunity that has the opportunity to have near term social impact in a small way? Or do we go after an opportunity to have long term social impact by supporting a nonprofit? And we kind of wanna do both, but in the earlys, we had to make the hard choices there in terms of what we prioritize. And we've come up with some strategies in terms of enabling our data for net positive social good efforts. We are running the first of its kind climate data studio data this month, which is bringing together many nonprofits and potential, uh, sustainability partners into the room and doing a shared on onboarding. And none of them are commercial users of our data, but they see the opportunity to use the data in interesting ways to support the underprivileged communities, support climate justice initiatives and support avenues, where can be better decision making, better knowledge sharing when they have more clean, validated data on the changing climate.
Melissa:
Cool. So you're kind of like supporting these other companies that are also supporting really good causes and through them, you can make sure that you're actually staying in touch with the, the good fight of the climate activists,
Gopal:
Right? And through that, we are actually building, you know, what, in the venture landscape, people often talk about network effects, but there's a subtle element of a data network effect, which is you want people to be able to use and understand your data. And that just takes time being very conscious of that, even in an environment, which is very data savvy with a community that is very data driven, takes time for people to fully understand your capabilities and your capabilities evolving. So having a group of people who are constantly in touch with your capabilities and using your data for social good, just magnifies the impact of what you can actually do.
Melissa:
Yeah. That's really nice. That's great advice for people who are out there. I think maybe trying to get into things that have a lot of impact. And I do hear from people that are like, I wish I could do what I do around these causes that are actually going to make a difference. So I guess what's your advice for people who are out there saying, you know, I'd rather help drive change in climate work for the government work on these different policies. They see the opportunities, how should they think about where to join in or how to start something that might be impactful in those areas and succeed with it?
Gopal:
Yeah, I would say, think about a team that is like learning, like any team you join, think about a team where agility and learning is essential to how they operate. And secondly, think about a purpose that resonates very close with your heart, because then you are willing to go the extra mile or the extra thousand miles towards making a success out of what you're trying to build. And product management in new markets with, uh, new capabilities is just inherently not easy. So being able to retreat and succeed in that space requires all those, the passion, the purpose and trait learning, oriented thinking, and a growth mindset to really come together.
Melissa:
Well, thank you so much for being on the podcast. This has been really enlightening. It's fun for me to hear about how you take such a mission driven company and make sure that you can thrive with it. Where can people learn more about Sust global and yourself?
Gopal:
Yeah. The best place for you all to learn more about trust is sustglobal.com and the best to reach me is LinkedIn. So feel free to send me a note there. I'm always happy to connect with new folks who are inspired by product and data and interested in doing more. We're always hiring for new folks across many different roles in product and engineering. So if any of the, those are interesting to you reach out to me directly on LinkedIn and you know, it's been a delight to be sharing more of the steeper topics with you, Melissa. Thanks for having me and for all of you, thank you so much for tuning in.
Melissa:
Thanks so much for those of you listening. Thank you for joining us and make sure that you subscribe to the podcast so that you can hear a new episode every Wednesday next week. We'll have another dear Melissa. So make sure that you submit your questions@dearmelissa.com and I'll answer them on the next episode with that. We'll see you next time.