Where are all the women founders in agtech, especially at later-stage startups? What does an acquisition really feel like, and when does it make sense? In this Bonus episode, we answer both these questions, as well as build on the key insights from our latest series: “Later-Stage Agtech Startup Lessons” .
This episode features:
Anastasia Volkova - co-founder and CEO of Regrow*, an independent measurement reporting and verification platform. Regrow, (formerly Flurosat), recently acquired US agtech company, Dagan, and raised $17 million in Series A funding.
Matthew Pryor - co-founder of Tenacious Ventures, and partner at the Agthentic Group. He previously co-founded Observant, a precision ag water management company, which was acquired by Jain Irrigation.
Our guests also examine some of the key lessons from the entrepreneurs featured in this series: Michael Gilbert of Semios, Paul Lightfoot of BrightFarms and Charles Baron of Farmers Business Network. .
*Disclaimer: Tenacious Ventures is an investor in Regrow.
Matthew Pryor [00:00]
hardware. I mean, to some extent leaf fills, there's a couple of companies in Australia, one called pear tree that probably sits in that spot the best, which is, you know, some farmers who do get it are still frustrated by the fragmentation. And so they need a pane of glass that kind of sits in front of everything. And, again, that's a good, it's a good use case. But you know, it shouldn't really be needed. But I think it's also interesting when you think of, like open team, and pharma, which is where a lot of that value proposition is, the reason why you want to be in control is not because you're worried about what other people are doing. It's because you're smart enough to be able to build stuff that encodes your value. I mean, I would draw the analogy there, say, for me with Rome research, like, I'm coding my own brain, I don't if you see any of I feel malls, stuff he is he's a partner at Main Sequence. And he he's written pretty extensively about how he uses Rome research to support his VC activity. And and it's effectively a kind of brain programming environment where you can code your proprietary approach to your day. So I think I think that's super interesting. And that's something that we don't talk about, almost at all. Which is like, what is the digitally version? What's the digitally native version of Excel look like? For for farms, like if you if you solve the data problem properly in a way that's largely in the interest of a farmer who's digitally native, then like, that's where you want to be?
Shane Thomas [1:49]
Mine mind map for the farm, so to speak. Yeah,
Matthew Pryor [1:52]
yeah. I mean, yeah, like a mind map, or like, like a table or something like that, like a no code environment that's just fully connected to everything. You know, where we're at, you know, at the moment, that kind of high art, like a, like a switched on producer really wants to drive their business, the way they want to drive it is probably got kind of, you know, macros coming out there, you know, years, and 20, different interconnected Excel spreadsheets or something. And there's not a lot of discussion around that we kind of have this expectation that the vendors are going to solve the problem completely. So that that's one one area where we have historically done quite a lot of work is, you know, what does that aggregation case look like? So there's, you know, those those kind of four use cases, or for usage modes, I suppose, if I'm a farm down at a farmer facing, I think the other insight there is, when people talk about the kind of the fear part, they're largely talking about what vendors will do with the data. Right, not what they will do with the data. And again, I just think that over indexing of those fears has the worst side effect of that has been, we're not even talking about, like those amazing good use cases, you know, what does a no code environment look like, if you really want to drive your business, you know, as optimally as you can, and we accept that that means digitizing and, and, you know, a digitally native version of a farm management approach. I mean, we kind of touched on it a little bit on that kind of lean farming, where we've had contact recently from a group in Australia called farm owners Academy, I think farm owners Academy. And mirrors, I can tell they're basically applying traction to running a farm is fabulous idea. But you so quickly gets to the point of okay, but we need a new a new generation of tools there. Because if you really want to build an operating system, for your digitally native farm, then, you know, it can't just be one vendors idea of what your farm looks like. So I think that's, that's an area that I just don't hear a lot of discussion about the sort of benefits of farm data. And so that, you know, kind of aggregation case, we've kind of got that interoperability case, the, you know, the interchange case, and then the and then the kind of migration case. And so, yeah, I think the response to the first article, largely still was around that. You know, but vendors don't tell farmers what they're going to do and all those things are true, but they really shouldn't be the place we're spending time and yeah, I think the the bit of the article that perhaps wasn't picked up as much was like, we are waiting around for people to solve a non problem. And then meanwhile, like, it's all there for those, like who really want it right, that was the other point of what we're trying to say was, if there's some big economic advantage that's comes from data, if it's big enough, then it's already being used. Because near enough data is already available in that kind of remotely sense setting, right? If you've got on farm storage, in all this kind of stuff, quite likely, I can task a satellite and take photos and estimate, you know, how much grain you're storing yourself. And, you know, if I could see your fields beforehand, have got a very good idea about what your father's didn't get out yet. And so if somehow I can corner a market or, you know, set prices, or predict demand for an input and stockpile it or something, I'm already doing it.
Sarah Nolet [6:03]
It's not true for the individuals, right, like, that's true for the companies, but like, I would definitely get value out of someone like aggregating my spending data and telling me how I could save 10% for a month or like, looking at my health data and giving an insurance but like, I don't do those things. Like there's still a behavioral change aspect for the individual of getting the value out of the data. And I guess you could say the value proposition doesn't outweigh the behavior change. Or just like, some people won't do it. I guess you mean more on the company side?
Matthew Pryor [6:38]
I don't mean more on the company side, I mean, that that's what the fear is driven through. Right, right. Because people say, Oh, I don't want somebody looking at my ex, right? Where, because because in your engineers, right? I don't want I don't want somebody predicting when my ASR is going to require maintenance and ringing me up saying, Hey, we can see that blah, blah, blah, blah, blah, like is that? That's the thing I the narrative is they've got information, high resolution information about me. And they're analyzing something about me, and I don't want them to do it. But the thing, that the question I keep asking myself is, why? Why are you worried about that? Because to date, and in all this conversation, despite all the commentary, no one has articulated here is a way that a farmer suffers significant economic harm as a result of having shared the farm level data. And that's a bit I don't get I agree that there's plenty of benefits that people aren't getting, and their user behavior change. And, you know, there is a kind of creepy factor to it, I suppose, in a way, but we'll get into the benefits, because we've been we just spinning in a circle, worrying about the bad things that are going to happen. And I just, I really struggled to know what actual bad things these are.
Shane Thomas [8:10]
What do you think they're? I mean, I think you just alluded to it earlier. But is there that we need to do we being whether it's the incumbent companies or startups to actually better illustrate what those tangible benefits are? Even if there is behavior change, even if there is some challenges to it, but doing a better job painting the picture of it? Or is it? Or like, what are your thoughts on on that?
Matthew Pryor [8:35]
Well, I think that's kind of what the the reference to the kind of psychology part of the argument is, is, is while ever we keep primarily talking about why you should be afraid, we won't get people to pay attention about why you what benefits you will get. Because it's it's this unspecified harm that we need to protect you against first, before you can get the benefits. That's basically the narrative.
Shane Thomas [9:05]
Yeah, no, I mean, it kind of reminds me of this, this concept from like, I think his name's like Ken arrow, it's like the paradox of information. Or it's like, the only way to price a piece of information is to reveal it. And so the value of anything is that you can access zero and so until you actually begin accessing it, but your point you're holding, you're not if you're actually scared of the implications, you don't take action. And so you never ever get to reveal what the value could be. And of course, there's going to be value that accrues to large companies and incumbents Of course, too. But if I'm afar, I mean, we have this zero sum mindset to where it's like, well, I want to derive the whole benefit as a as a farmer vs. Acknowledging that, hey, you know what, I'll get 85% or whatever the number is, and somebody is, of course going to benefit along the value chain as well, which I don't know if that gets talked about in a call it compelling way to the to the farmer either but
Matthew Pryor [10:04]
yeah, I mean, I think that's I think that is probably true. But I but it is a conundrum right? Because again in talking about it and trying to solve it you the risk is you contribute to Tim Neil from data farming added a really good comment on the LinkedIn post where he basically said, Look, D are far more concerned about making their business more efficient, and you know, delivering better customer service. Now, that's an interesting question, right? Because if, if they are using individual farm livered level data to make that true? Who cares? Like, wouldn't you prefer that they made use of that information to deliver your more efficient service? I think the the question is like, does the reverse happen? Do they somehow work out that they can charge me more? Because I rely? Yeah, I just,
Shane Thomas [11:03]
that's a risk in in literally everything, you
Matthew Pryor [11:06]
literally everything you do? Yeah, that's right. And again, if that information, if, if the advantage they were going to get was so big, because if they figured they could get 10% from every single customer, if you know, 10% more from every single person, might they'd be doing it? Yeah, they would already be doing it, I would already have found a way to increase the price of that product to, you know, take to shorten the delivery time to you know, add a premium to something like they're not sitting around waiting for opportunities to make their business better. That's the other other part of it, I think, which is this kind of like, yeah, it adds to the narrative of, I'm going to get charged for
Shane Thomas [11:51]
Yeah, it's like all any like, not that this is a conspiracy necessarily. But it's like a conspiracy theory, where it's all of a sudden, just like, oh, yeah, this will all happen. And this will happen. And it all kind of makes sense. But you're not looking at the realities of all the other aspects like nobody's sitting there trying to, to your point, trying to extract another 7%, or whatever it is necessarily out to the farmer. But what they would like to do is add so much value to that farmer by you know, sticking with the John Deere example, by actually being able to do some sort of predictive maintenance that helps that farmer stay up and running during harvest, so that they if you don't actually get any tangible dollar value out of that, but they would have had a significant cost. But they wouldn't do that. So that now you want to actually have two John Deere combines instead of you know, another case when that you bought used the other year or something like that, right. So it's, I mean, of course, companies always want to sell more things, but they also know where they can work. They know, they know, excuse me, they know that they need to know that they can't actually just call it what's it called have a super high take rate or anything like that. So yeah. Anyways, it's an interesting, interesting dynamic around the entirety of it.
Matthew Pryor [13:03]
Yeah, I mean, I think I think you so you're the question you kind of asked was, can we make can we make progress here by talking more about the benefits? I think that's what what I was trying to achieve, at least was, I think, in articulating those kind of four use cases is at least establish a clear vocabulary, just sort of say, you know, that if we just simply make it good versus bad, we oversimplify and, and human nature is to index to that, right? Whereas if we sort of talk about it in a more nuanced way, and so you know, sort of talking about, okay, why would you aggregate data? Why do you need interoperability? You know, when do you interchange between systems? Why is migration important? Then maybe we can have a more nuanced conversation and kind of say, well, okay, really, the fees are not very well justified, and the benefits that I could get for my business, in being able to sit in front of the dashboard that I can kind of code my own business rules into, you know, in a way that suits me in a way that takes what I believe to be proprietary and keeps it for myself, so I can mine my own advantages. Surely, that's where we want to be rather than worrying about whether I'm going to pay 2% More for an input, because someone has access to detailed as slides and machine hours and, you know, yield maps.
Shane Thomas [14:43]
Yeah. Like the thing that I come back to all the time around around this is, if you like the percentage of customer facing staff that can actually engage in a conversation with a farmer around some of the things you're alluding to with interchanges and yeah, What that means to let, because that's where I come back to it just started writing about your article in upstream actually earlier today and and it come back to like, if you look at this as a incomplete or, or not exactly comparable example, but you look at weed resistance, which is one of the biggest problems that farmers deal with, right? How many dollars go into that with sorry? How many dollars go into training staff so that if a farmer says, Oh, I'm just gonna apply my glyphosate, you push back on and say, No, actually, you need to include a report in and you do to group, you know, one to one, whatever, whatever, it's a good one. AD, it's good. None of that has actually went into actually educating the staff to support them with having a conversation around, hey, you know what, this is why we should actually be looking at it from this perspective. And I think even you look at the end again, this is not to be too poorly about anybody in the industry. And, and but if you look at like the average senior leader, and in the industry, they're not like, how they got to where they are wasn't because they were super strong and understanding data, they have a bunch of other skills, but the way they tend to approach data is around this. Okay, well, you know, what we need to do, we need to make sure we own it, we need to make sure of XYZ, but it's not actually thinking about it in a strategic way. Again, that's just an average kind of observation. But then everybody in middle management sees that everybody that they talk to, it kind of trickles down the organization. And it just leads to this thing that you talked about to where if I'm, if I'm a staff, and and a farmer says, Well, who owns the data? Or how are you going to use my data, all of a sudden you like freeze up? Or it's like, no, you own the data? Yeah, let's, let's not push anywhere on that. And so you, you reinvent, again, this is just getting at the same thing we've talked about is it just reinforces fear in, in the farmer, and there's fear in the staff to have that conversation. And so you end up with this standstill of just like staring at each other, saying, Well, I guess we'll just stick with the status quo. And that is what it is. But again, there's more things are within that. But
Matthew Pryor [17:00]
that's such a good insight chain. That's such a good insight. Because I think, yeah, that that makes a lot of sense. Right? In the way I would play that back to you was to kind of say that that idea of so a base level of kind of agronomic knowledge is absolutely required, right, in ag retail, you couldn't possibly walk into a room without, you know, a base level of of agronomic knowledge. And so to your point that would include well, like, what is multi, you know, resistance mean, in weed control, and why it's important to have a weed management approach, you know, that is informed by what resistance and then there is no analog for data, right? And that, that, when you're in that territory of my lack of knowledge, my tendency is to simplify and go to a simple argument. And actually, the more emotionally resonant it is, the less risk I have of seeming like I don't know what I'm talking about. Exactly. Yeah, that's, that's very, very, very insightful, would
Shane Thomas [18:17]
see it all the time, even like away from data. But you know, herbicides were well understood by agronomist, whether in retail, or manufacturers or what have you. But then you get into, say, fungicides, it's a harder conversation to have, and all of a sudden, you had to have an understanding of the diseases and cycles and weather and active ingredients in a different type of segment. And all of a sudden, half would shy away from having that conversation. Because a farmer would say, well, actually, no, I don't I don't need it, because I've never had that disease before, why would I have it now and if you don't have the confidence to that conversation with some actual tangible background to pull from, it's a heart you're standing on a house of cards, essentially, instead of some call it a somebody's shoulders that where you feel like you can actually take on on the conversation, engage with it and add value to the farmer. And so you get this weird dynamic of to your point where, you know, you kind of you kind of shy away or you you, you don't push on it like you, you would have you in a topic that you have confidence in and I know I find it in myself, even in other areas that aren't specific to, you know, agronomy or something like that. But when I get into talking about different types of software dynamics, and all those sorts of things, and so probably very similar in, in staff in in the industry that are customer facing.
Sarah Nolet [19:34]
It's interesting, there's two kinds of things that made me think of one, like we talk about the shifts, changing the industry as like tech and climate. But some of the other shifts are like bringing a grant agronomic advice in house like the bait right like as the dealers are bringing in more in house as farmers are getting to scale what they bring in in house. Like is that what changes this conversation? And like what opportunities does that create because we I think we talked about that less as a, it's like a result, not a driver. But to your point here, it actually can change some of the adoption dynamics.
Shane Thomas [20:09]
I think it probably probably does, especially if you look at the equipment dealers with having that sort of conversation, all of a sudden, you've, you've tied in a much more meaningful, holistic conversation around at all Where's data and I mean, agronomic again, I'm an agronomist, but background, so I've zeroed in on that. But agronomy is different than precision, in digital, which is different than how data can be used. And so even though they're all tied in, you need to have understanding, and in each to be effective, everybody has varying levels of understanding of each. And if you don't have a strong grasp of all those, you kind of shy away from different areas of it. And so until you have the confidence to or at least the core competency in a organization or a specific segment of the industry, you probably don't get the forward progress that you want. But that might just be my thought around that. But yeah.
Matthew Pryor [21:05]
Yeah, fascinating. Yeah. So So I guess that that's kind of where, yeah, that that idea of pushing more into the vocabulary right around? Okay, well, it's not just that, right? That there's different use cases for it. And there's different reasons why you should focus on it, and there's maybe then we can create the space to kind of say, yeah, the stuff that you want to have access to is the stuff that allows you to run your business better. And I think the other part, which again, kind of gets lost easily is the difference between what vendors say, right about about what they're doing. And how we make that true. Like, we can only make it true by vendors not acting like jerks like this idea that by defining data better, ie in a legal sense, and defining what ownership is better, in a legal sense, is hopelessly lost by the year and will go on forever. And in my opinion, definitely increases the narrative of fear and delivers absolute zero, as far as functioning solutions. And so I think that, you know, like the, all that the stuff we touched on in the last recording, where, you know, those kinds of data codes, the idea of making it principles based and making them strong, right. So I could say, as a vendor, you know, you won't suffer economic harm, because of something that I do with the information you give to me. Right, that's meaningful, to sort of say, Oh, your data is the original thing you give me, but I own the data that I derived from, like, helping you here. I think that's I just don't want us to go down that path of, you know, taking a kind of definitional approach, we have to stay at principles, and we have to strengthen the principles. And I think also in teasing out, like, it's not just the data, but but the, you know, so So I don't want again, I don't own it, or who owns it, let's not bother, but you know, you can never prevent me from accessing the information that you sourced from me. And, you know, must make it possible for me to get it all back in its entirety in the exact form that I gave it to you.
Shane Thomas [23:59]
Yeah, well, I even think this like just to take that I think that goes back to your comment on on the interoperability because if I'm a farmer, I don't actually want my data in the sense of I if I'm choosing to go to XYZ Farm Management System or whatever it is, you need to be able to get it to there within seven days of me requesting it right like that's what fundamentally what they want and I don't I mean, there's only so many companies that could come close to something like that today but to your point that leads to the lack of lack of confidence in a farmer to say okay, well do I want to commit to going in and doing this when it's going to be so hard for me to pull it out? And and what do I do with it when I get it back? So you just have you have all these second and third order implication questions that come up from it, even from the first one which is big and can't be answered and then you have all these secondary aspects that just started like a domino effect that fall after that, that leads to that lack of utilization or, or engagement with some of these these systems or tools to help utilize data or digital NYSED tool. So what
Matthew Pryor [25:02]
I mean, I think I think, again, I think that, you know, clarity of the use cases actually helps explain also why, why someone should pay, right? Why a farmer should pay. And that's why we tried to tease apart the difference between, say, interoperability versus migration in the interoperability case is like, Okay, I know that there are a couple of vendors involved there. And I, I get a better cyst service. When they're talking to each other, I don't really know how it works, I don't care. But I just, I get to make better decisions, I get to get better insights, because they're talking to each other in real time. And I understand that I need to pay him both. Because, you know, it's like the surgeon and the stutters, right, they're both required to do the surgery. And I get a bill for a buck from both of them. And I understand what it is. Whereas the migration one is, someone else has a better value proposition than you. And it can't the reason for me not to be able to move can't be because I can't get my data back. Right. Like that's, that's, that's the that's the worst possible case. And if that's your reason for not making it possible, that's exactly why people don't like subscription services. Right? You've failed to create enough value I want to leave and you won't let me leave.
Shane Thomas [26:28]
It's like, it's entirely it because like, I look at, like, if you look at like the seven powers framework, if you're familiar with that, you know, everybody, you can really over index and zero in on switching costs, like let's have high switching costs, instead of saying, Well, let's try to figure out how to have higher network effects situation. And that's what you actually need versus having these high switching costs. And, again, in certain areas of strategy, of course, you have the fit, but for what we're talking about it, there's been too much emphasis on the switching cost aspect versus trying to create the the network effects. But yeah,
Matthew Pryor [27:01]
yeah, yeah, I agree. I agree that it's a value creation problem. And putting the migration up as an artificial barrier to prevent churn which which we did touch on in that previous one, I think is it's a terrible reason. And it does harm to the industry in general.
Shane Thomas [27:20]
Sarah Nolet [27:25]
Cool. Can I ask you guys about something else? Yeah. I'm trying to figure out what this is, like, sounds really stupid. All the incentives like indominus, swarm harm about their rays and go to market and all this stuff, the like, idea of the crop intelligence, and stuff that like a jerris, and mineral and like every company is doing is sort of coming up. And I'm struggling to get my head around a framework for, like, the different parts of the tech stack. And to build a bit of a theory of market for what parts make sense to own versus not own. And how that's likely to evolve. And so I guess my question is, when, maybe not, at first, like when someone talks about crop intelligence, like slide, Edgar Iris, what, how would you define that? Like, my sense is, it's like, is it the data collection and the custom or off the shelf hardware that is sensing and collecting the data? And then the ability to build models that are custom for that prop or that use case? And doesn't end with then the closing the loop on the action? Like the effector that does the thing? Like how would you kind of divide up the parts, because obviously, I would divide them and you would love to win together. But I'm just struggling to first one farmer in particular in articulating, say to an investor, like, we believe that the parts to own or these parts and the parts to partner or these parts, but that market doesn't exist yet. And so we need to own more now. So that we can own less later. And I'm struggling to figure out how to say and then define those lines. Does that question make sense?
Matthew Pryor [29:23]
Yeah. Yeah, I mean, it does. I in classic lumber fashion, I would, I would kind of go to the like, why are you even doing it? In the first place? So So I think you're right, there's a pipe, I would, I would call it a pipeline. There's definitely a pipeline. That would be the pieces that make up what you might, what people might call crop intelligence. But I think the reason to do it is to get a better outcome. So the stuff certainly in the kind of, you know, veggies and, and those sorts of things is because you can borrow was, right, I mean, all of this goes to use the first beneficiary. And And most likely, crop intelligence is only of incremental value to the grower. But when the grower considers their relationship with their downstream, and where especially in, you know, fresh produce timing and supply chain timing, and call chain and all that kind of stuff are really significant factors in the net economic advantage, then I think that's where crop intelligence becomes really interesting, because it's not just we're optimizing the growing but we understand enough about the system, we've both sourced and modeled, to the point where our ability to manage it, you know, as a connected as a, as a supply chain connected system is greatly enhanced, and therefore, our value to our downstream is higher, because they highly value, predictability and responsiveness, the right balance between predictability and responsiveness. Because the signals are kind of consumer demand, and that's, you know, some combination of weather and you know, social things, and blah, blah, blah. So I think, yeah, like, there's, there's absolutely the stuff that will make production more efficient. But I think the far more interesting parts are probably the stuff that makes me a better supply chain participant because of my ability to be much more both predictable and adaptable.
Shane Thomas [31:46]
How would you guide? You know, no, no, I was gonna say, I like, I like the way you you put that Not really, because I know where my answer would have stopped is not taking into account all the responsiveness to the downside, which I love that, that comment in terms of that, that would provide and how to think about that in the context of crop intelligence, so to speak, which Yeah, I think those are really good.
Matthew Pryor [32:11]
And I mean, we, you know, when we, when we did that decently deep dive on robotic harvesting, right? Like it, I mean, absolutely, it is true, that you can't get labor. And so having, you know, an autonomously assisted harvesting process is better and will lower production and will be will be interesting to orchardists. But it's far more economically meaningful when you expand your scope, through the packing shed into the supply chain, and you sort of think, well, okay, how much do we need a cool room? Now? How much do we need a great big Installation have kind of wash and saw and grade? Because crop intelligence should be I already know what grade the fruit is, before they even take it off a tree. In fact, I wouldn't take it off the tree unless I was sure it was the grade that I'm currently looking for. And so you, you, you get game is that kind of Thompson esque aggregation theory, like at the moment, the grower gets a certain amount of the dollar and the pick pack kind of delivery, whatever gets a certain amount of the dollar and the wholesaler and retailer, you know, kind of divide up the rest. But you know, real crop intelligence would be well, okay, are we now talking about sort of distributed, Pac shared? And, you know, grading happens? What, at flowering time, like, I know, I should know so much that I can create new economic opportunity, maybe that's the simplified version. The proper definition of crop intelligence means I can create new economic opportunity through the kind of insight and action that's enabled by the crop intelligence. Because otherwise it you know, it's incremental efficiency, and that'll be important, but but but it's limited to production.
Sarah Nolet [34:23]
Do you? Like I agree with all that, that all makes sense to me and is not the part I'm struggling with the part I'm struggling with is more the theory of market for how we get to a world where that is unlocked? And where are the places you'd want to spend money. And it seems like the end effector like the last part of implementing the recommendation to do the pick or whatever is going to be pretty, like we've talked about this kind of that going to a component tree marketplace where none of those individual solutions are likely to be venture scale. So then if you go back a step you There's like the model and the analytics and the knowledge of the crop and then the sensors, whether off the shelf or custom to collect that data. And then there's the like the data sets, I don't know. So my sense is that those three things get coupled. And because you wouldn't necessarily want to own, like owning the dataset, if you can't, if you don't have the model, or the collection for the dataset doesn't make sense owning the models, without the other two, you can't improve it owning the collection, you would do the other part, right, you would do the analytics and own dataset. So maybe the part you could separate out is the data set as like, if you're a swarm farm, you can provide like a startup sandbox for people to come in and improve the other parts. Because I'm trying to the part I'm trying to solve is like, if you're someone from right now, you don't have the distribution and market share, to incentivize app developers to come on your platform. What you could develop, like Cornerstone partnerships to build that out. But what's the theory of market for over time? What parts of the of your operating system intersect with crop intelligence? Like how much do you do you provide the data sets in the sandbox and others build it? Don't you provide?
Matthew Pryor [36:12]
Me I think I think AWS is the analog there, right? Like you provide a T IDs and services and you just keep enriching, you just keep pushing innovation back. And so your early app developers are kind of on bare metal, right? Your next crop of app developers don't need to worry about bringing up individual machine instances, they've now got just, you know, serverless stuff, and it automatically scales, the next set of app developers don't need to worry about containers don't need to worry about data don't need to worry about pipelining. I think I think that's the way I think that's a pretty clearly established pattern that you just, yeah, definitely the vision is that the sandboxing? Is there the kind of YOLO processing of the imagery that's coming in, ideally, that even the camera, right, like, Yeah, I think that that the AWS analogy, there works pretty well. And you just at the moment, you're at the kind of AMI level, right, and so app developers are, you know, what operating system has the machine instance, Scott, and I'm responsible for keeping that operating system updated. And I'm responsible for deciding whether it comes up on demand or whether it runs full time, and I pay more that kind of thing. Whereas today, like literally, I just write a serverless function, and put an endpoint up. And when someone hits the endpoint, Amazon spins up, whatever I don't know, like, law, and the imperatives that sit behind that serverless function are much higher level as well, you know, the data that I can either get access to the pipelines, I can drop, whatever comes to me into that gets processed downstream. I think that's, that's what you want. And so what are those? I think the reasoning to do there would be what are the building blocks, right? If, if were AWS today, with just machine instances, you know, what do we think an app developer wants in five years time?
Sarah Nolet [38:19]
Yeah, that's and I think the overlay there that's interesting is like the app developer ecosystem is maturing, right now. You've got like the steel implement, guys, and so that they've already solved for that. And you can, you can spin one of those up on swarm. But now Oh, no problem. Then you've got the like, weed it weed, the weed seeker guys who are it's metal and software. But they don't, it's still API's. And there's still like, a quite a bit of meat in the middle to build that out. And they don't yet have the ecosystem like they don't yet want to appstore in which to build like, they're kind of not there yet. But then you've got the future of the whatever, bilberries and cropsies, and whoever else. But they then you get into the chicken egg of like any distribution to get them. So it's kind of like what do you Yeah, what set of how do you actually get to the Envision, which makes a lot of sense is is where I'm stuck. I don't have time. But that's, maybe we can talk about that again, another time.
Shane Thomas [39:21]
Yeah, it's interesting to think about that the market dynamics with within that I have to think a little bit more on I'll send you any thoughts I come up with? That's an interesting question within that content.
Sarah Nolet [39:32]
Yeah. I mean, it would be it'd be interesting, like the way we've framed a certain pitch so far, and I'd love to your take on at some point is like there's kind of three categories of autonomy. There's like the retrofit solutions, the point solutions, like do one thing and then the swarm farms in the world, which is like a platform, but a new approach. And so to scale the platform you need four things basically, you need farm hardened, like tech that actually works that does, you know, the robot that can do stuff, but you also need like digitally native distribution. Like Android says, the idea is I wake up, I'm a farmer, I wake up in the mill and I say I want a robot, I roll over, I get my iPad, I order it seven days shows up, like that's what we eat, which I love. And then there's like digitally native support, which goes to the conversation we had before, like, who are the people that are going to fix it? And is the robot designed to be fixed by an ecosystem of third party, and pharma people or not. So that's what they're building. And then the fourth one is like an app store of other people who are adding value to the robot, and like building new apps and evolving it, whether it's like spray drift, or new use cases, or carbon monitoring, or whatever. And so it's that last piece that you have a few you have to get distribution before the App Store people want to build. But what set of features do you have with the App Store people who they are evolving, because there's venture money going in there? Like, it's that kind of theory of market for that last piece, but I'm struggling with in terms of how to explain it.
Shane Thomas [40:56]
That is interesting, just in the dynamics of that to actually have the incentive for them to want to build those sorts of capabilities and plug into your platform or your ecosystem, so to speak around that.
Sarah Nolet [41:08]
And is there anything other than like, the easiest explanation is like we have distribution. We're the only autonomous platform with 70,000 customers, and it's like, that's probably too high a hurdle to clear, because no one like, you'll just have to wait too long. And so those are the apps that are uniquely well suited to autonomy, just like can't come to market then they all go to retrofit solutions. Which then I think there's value left on the table. So how do you like advance the two things together? Yeah, that's
Shane Thomas [41:41]
interesting. Huh? Maybe this one group I should look into them too but swarm
Sarah Nolet [41:52]
right. Yeah. It's one of our customers Yes. Where they have some really cool stuff. They showed me a robot GDC that video jumps they'll robot delivery like it's just awesome like Andrew the the CEO can like personally delivered a robot and left the farm in six hours. It was like from show up on the farm to happy customer go and see you later. Like full installation up and running ready to go. Because now they've pushed to your point like they've pushed the mopping and the paddock setup and all that on to the farmer now the farmers love doing it. But the then Andrew Indra at home can log in pick the robot and like see in the middle of the night spraying a paddock and like see the lights going on? Just like from his iPad 1000s of miles away. And it's yeah, it's pretty sick.
Shane Thomas [42:39]
That is interesting. I'm gonna look more into them.
Matthew Pryor [42:44]
Did you get in touch with Bill Binney and
Shane Thomas [42:47]
yeah, I had a call with with him last week actually, that was really impressed when we think about everything and exciting I mean, I really liked that they how they were thinking about just the crops they target the manufacturers they target I think it's so smart in terms of mean it's easy to get all its do corn and and and all that but nobody's targeting the cereal market in the drier areas and the secondary call it the tier two or tier whatever going home three manufacturers of of products. So and I mean, that to me was interesting because I was always like okay, well you're gonna have these Nkosi and H and deer and they're gonna be entirely integrate. And so if you don't get actually bought then you're out you're on the outside looking in but with their logic I actually was like, Oh, I was wrong in thinking that I think it was so interesting. Interesting group I didn't realize they had that sort of commercial traction actually are gonna be too
Matthew Pryor [43:45]
Yeah, they've done they've done very well in Australia with with those commercial partnerships and pretty well on your as well.
Shane Thomas [43:53]
Yeah, I'm going to try and connect with whoever they have and hopefully go and see it in action is is the goal at some point in the next few months. So hopefully that will work but yeah, thanks for the connection on that they were going was great. Awesome,
Sarah Nolet [44:10]
awesome. By Jane. Thank you can always appreciate the conversation.
Tenacious Ventures Management Pty Ltd (CAR 001275760), Tenacious Ventures Management Partnership, LP (CAR 001298484), Tenacious Ventures Fund II Management Partnership, LP (CAR 001298483), and Tenacious Ventures Fund II Staple Co Pty Ltd (CAR 001298487) are Corporate Authorised Representatives of Sandford Capital Pty Ltd (ABN 82 600 590 887), Australian Financial Services Licence No 461981, and are authorised to provide advisory and dealing in connection with investments to wholesale clients only.