On the surface, there doesn’t appear to be a close connection between farming and modern technology. But
Milan Dobrota and his startup Agremo are starting to change that. Agremo uses AI technology to analyze
aerial imagery as a method to help in crop production. Agremo’s technology can be used for any crop and
addresses a huge lack of data that is useful for crops and agriculture.
Milan spent time this week speaking with Roland Siebelink on the Silicon Valley Momentum Podcast. They spoke
about how the agriculture industry can benefit from Agremo’s technology in multiple ways and some of the
reasons for Agremo’s early success:
- Why some of Agremo’s best customers are ones that have already tried similar solutions.
- Why Agremo’s focus has nothing to do with competing companies.
- The best way to approach customers when selling a new technology.
- How creating partnerships has been an integral part of Agremo’s growth.
- Why it’s so critical for startup founders to keep learning so that they can be better than they
Roland Siebelink:Hello and welcome to the Silicon Valley Momentum Podcast. My name is Roland
Siebelink. I'm a scaleup ally for tech founders. I'm very excited today because we have a guest all the
way from Serbia. It's Milan Dobrata, the CEO of Agremo. Welcome, Milan.
Milan Dobrota:Hello, Roland. And hello to all of your listeners. It's my pleasure to be here with
Roland Siebelink:Absolutely. The honor is totally on our side. Milan, let's just jump right into
it. What does Agremo do? And what difference do you make for whom in the world?
Milan Dobrota:In Agremo, we are developing the technology, which is actually software technology,
that uses AI to analyze aerial imagery. The problem we are solving is that we are attacking the issue of
having sub-optimal crop production caused by the fact that there are so many unknowns and there is a huge
lack of data about crops and about fields. We are basically replacing the inefficient techniques of
gathering the data, which are usually incomplete, expensive, or inaccurate.
Roland Siebelink:Okay. What kind of crop production do you typically try to optimize with Agremo?
Is it all crops anywhere or are you focusing on specific subsets of crops?
Milan Dobrota:The technology is built in a way that one of the important benefits is that it can be
trained easily to attack new challenges and deal with, theoretically, any crop and any phenomena using any
data sources. Aside from the high accuracy, the adaptability is a very important differentiation of ours.
We can deal with any crop and any phenomena within the crop. But of course, in our efforts to make it more
available and accessible on the market, we are focusing on a smaller number of crops. For instance, such
as corn, wheat, soybeans, barley. Typically, it's cereal, industrial plants such as sugar beet, then
vegetables, tomato, potato, but also in forestry, plantations, and so on.
Roland Siebelink:That's awesome. Okay. Really the core staple crops that are being used all around
the world, it sounds. That makes sense to at least start with those beachheads, right?
Tell us a little bit about the history of Agremo? How did you get into this business? When did you start
the company? On what basis?
Milan Dobrota:As an idea, we launched in 2014. Actually, it was launched as a project within, at
that time a mother company, which was in the software services business. We wanted to build our own
product, of course, in the area where our strengths are. Besides agriculture, it was software engineering
and data science. By the way, there was a lot of hype about drone usage, commercial drones usage in
agriculture and other industries. Luckily for us, there is no more such hype. Now it's just all in the
real business, so we are very happy for that. We started investing in the idea and initially got funded by
an EU grant and it started to grow from there. After, in 2018, we spinned it off completely with a seed
round investment. Now it has, of course, its independent journey.
Roland Siebelink:Absolutely. That's a different trajectory than many of the startups I talk to.
When you compare your history at Agremo to some of the stories you read or hear from other founders, do
you feel it makes a difference that you started off as a corporate venture, finding your independence
later versus the people who just start out in a garage?
Milan Dobrota:Yeah. Actually, we didn't start it in that sense from a corporation because the
previous company was also a privately held company founded basically by the same founding team. But it had
a completely different journey. We benefited in that regard that we knew each other for a long time
already. Speaking of the founding team, we did a business together. We knew the strengths and the
weaknesses. In this sense, it was much easier for us.
Of course, there was still a lot to learn about the new business, about how to make a good product that
will solve important problems in a very easy and not complex way for customers and benefit everyone. It
was a bit easier, but still not too easy for us.
Roland Siebelink:I understand. It sounds like you all came from a software background, is that
Milan Dobrota:Yeah, exactly. My background is actually software engineering and my PhD is in data
science or more concretely computational statistics. I have around 19 years experience already in the
industry. Regardless of all that, getting an entrepreneurial journey, launching and managing a company
like this that is developing a product for a global market and trying to develop the business globally,
sometimes I feel like I don't have 19 months of experience, not to mention 19 years. I'm definitely
learning how to do this business. The most important fact about my background is that I'm learning.
Roland Siebelink:Learning. Yes, absolutely. And I love to hear that when founders and CEOs really
bring that focus on learning in the job. That's such an important asset.
Does any of the co-founders have a particular background in agriculture or was that just a use case that
seemed very obvious to you guys to venture into?
Milan Dobrota:It's actually something in between. By professional vocation, none of the core
founding team was an agronomist or something like that. But we all grew up on farms. We are from the rural
areas, so we've had it in our genes and in our mindset to work in and for agriculture. But our core
experience was in technology, in software engineering, in data science, and product development.
Roland Siebelink:The software helps to optimize yields, optimize crop production. Who's the
customer? Do you sell it directly to farmers or how should I imagine that go-to market?
Milan Dobrota:Yeah, that's an interesting question. The thing is that often technology, you think
of it as magic. You push a button and you have higher yields or more profits. It doesn't work that way. I
just want to emphasize that. Actually, we are giving you much more accurate digital information in real
time of what's going on in your fields. This is the first thing that is very important.
The second question is how and who does that help. It's not just the farmers and the crop producers who
are benefiting. There are benefits throughout the agricultural value chain. For instance, input suppliers
need this kind of technology to help their customers to grow their crops better on one side. They need it
in order to give them the premium service to help them in that. Those who are advising growers, whose job
is to help them to grow the crops better and to bring and integrate technology to them, are more likely to
use directly our technology and then take it to the growers.
To answer your question, typically, it's not the farmers (individual farmers) who are our direct users.
But the large enterprise growers who have already precision agriculture teams who already have initiatives
to implement precision ag techniques, they are.
Roland Siebelink:Okay. Okay. You're really building on an existing capability and existing function
that was already tasked with optimizing yields, getting more and more data. And there, you're making their
job a lot easier by just providing them with that precision data, as I understand it.
Milan Dobrota:Yeah, absolutely. That's the roadmap to technology adoption. But I'm sure that as
time goes - typically, the majority of our customers are either those who are aware that there are
precision ag techniques and something that they need to implement to grow the crops better and those who
are working to implement it. The number of those who are completely unaware of the existence of such
technology or those who already implemented it some long time ago, there are not so many of those. We are
working in this middle with the customers who work actively or are planning to work actively in
implementing these kind of technologies into their workflows, decision-making processes.
Roland Siebelink:Very good. Very good. It's not so much a sales process where you have to start all
the way at the beginning and try to convince them of, "Oh, what if you had better data?" It's really,
they're already looking for solutions, as I understand it. And then it's about convincing them that your
solution is the best.
Milan Dobrota:Yeah, absolutely. It's exactly like that. The majority of our customers or potential
customers already come with concrete questions. They heard about a lot of technology, but now they need to
cross the gap - between what we know about the technology to implementing it as a solution to our real
problems. And this is where the majority of our effort goes. You have to put it in real work to identify
and to resolve concrete use cases.
Roland Siebelink:Okay, very good. If they are already that conscious about their needs, it must
also be that you are typically competing against certain alternatives, whether it's other companies, and
you don't have to name them, of course. But what would you say is your biggest barrier to entry with those
customers? Is it the other competition in the same field? Is it other startups? Is it incumbent solutions?
Or sometimes the biggest competitor is people doing nothing.
Milan Dobrota:Yeah. Of course, there are many good companies acting in the same space as we do. We
have a lot of good feedback from our customers. And we like to say that sometimes our best customers are
those who previously tried other solutions, from our competitors, but typically they are very happy with
what we can achieve in terms of the accuracy that we can provide, the actionability of results, our
service and, all the efforts that we're putting into delivering it and integrating the entire solution
into their business and workloads.
To be honest, we are not focused so much in fighting against the competition. As I mentioned, maybe even
more than once, we are more focused on putting it into real use. It's the change management and the
awareness and how you integrate the technology. And how are you at the end solving the issues that the
agricultural professionals have that we are putting the most effort to. Maybe somebody is doing it better
than us. But I think we are doing fine with having the focus on the customer and not looking so much left
Roland Siebelink:Okay. That's very good. And it sounds like part of your unique value proposition
is to think beyond the software and think about, "Well, how does this actually become part of their
workflow? How can they actually start using it with the things they're already used to?" The services
component of your offering seems quite important. Would you say that's correct?
Milan Dobrota:Yeah, absolutely. We notice that it's much easier to explain and to prove what are
the benefits and the values for our customer than to remove fear of a customer or potential customer about
the complexity that this new piece of technology can bring. They are being bombarded constantly with new
technologies, with machineries. Whether it's drones, satellites, softwares, et cetera. And simply, they
need these complexities removed.
And this is what we are trying, and this is why we are building also the different partnerships in the
industry from integrations to other software, to supporting different hardware and equipment, to find the
partners in each country and region, drone professionals who can collect the data so that we don't shift
all this complexity of establishing the workflow and implementing technology to the customer. We want to
deliver a turnkey solution to them. We want it to be as simple as possible, so that for them, we just
reduce it to what are the benefits and what are the values it brings them. And they can leave the
complexities to us.
Roland Siebelink:Oh, I love that. That's so rare, I would say, in what is still an early stage
startup, as I understand it, to already have that full customer focus in mind. They don't just want the
software, they want the solution that works. The turnkey solution, as you call it, right?
I like also that you're mentioning partnerships there. It sounds like while you're trying to provide a
complete offering, that doesn't necessarily mean you have to do all that activity in-house.
Milan Dobrota:It wouldn't be possible for us to collect this data. This is why definitely we have
to rely on partnering with drone service providers. And luckily for us, this is a growing industry that is
getting professionalized and improving each day. It's also about having a partnership with other
softwares. For instance, integrations with machinery management software, so that the outputs that we
generate - which can be the weed pressure map - can be an input into creating a spraying map that you can
download to your machinery. There are different kinds of integrations that you need to achieve in the
workflow, in the technology stack, so that at the end, what I mentioned, you removed most of the
complexities for the customer.
Roland Siebelink:Okay. Very good. I think it's pretty rare for an early stage startup to already
have such a developed view of partnerships. For the founders that are listening and that are also thinking
of having a partnership component to their business model, what's been your experience in how do you find
those partners? How do you persuade them to work with you, especially when you're still small yourself?
And how do you keep them happy over time?
Milan Dobrota:Yeah. If you come to, what I learned by my experience, no matter what the the size of
a potential partner on the other side - whether it's a one man company or a large global corporation - if
you come to them with a viable win-win proposition that brings a real value to them as well, then it's not
hard at all to start this kind of partnership.
Of course, they may approach it later with higher or lower priorities e.g. in sales execution. How well
will their field force or their sales representative promote the integrated solution, that's another
thing. To start to work with such a company, it's all about having a meaningful value proposition for all
sides. I find it really not easy. It's all about finding the real deal to establish with them.
Roland Siebelink:Okay. You start from a good win-win and then it really becomes a matter of
following up with them and keeping them interested, right? I think many people close the deal and then
think that that's everything. And that's only the start of hopefully a long relationship.
Very good. With that, we're actually moving already into how much traction have you seen with Agremo. You
mentioned customers all around the world, so it seems like it's growing very fast.
Milan Dobrota:Yeah. We do have a bit over 800 active customers from over 70 countries in the world.
Roland Siebelink:Congratulations, Milan. That's awesome.
For those founders that are as impressed with these numbers as I am, can you tell them a little bit how
you got to so many customers in what is still a relatively short time? What's your go-to market? How do
you reach all those customers? How do you get them to sign up for Agremo?
Milan Dobrota:Yeah. These figures, if you don't put them in the context of some other numbers, such
as the average yearly contract value, for instance, they don't mean a lot. It's all about segmenting by
the sales channel in coverage matrix.
Some of the customers which are much larger by number but the yearly contracts are of lower value, we get
them through a partnership with other companies, so that it also adds the integrated solution to the
overall brand. They are coming through the self conversions on a freemium model. Then customers who are
enterprise customers and who generate a larger yearly contract and with whom we have larger cooperation,
we typically get to them through direct or inside sales channel, whether it's inbound or outbound leads.
And finally, we have local partnerships in different regions, by referral and reseller models.
I believe there is no one right answer to any piece of advice that I could give. But it's more knowing
what could be or knowing what to test. Knowing could be too much to say because usually you don't know
when you start it. But testing different sales channels for different customer segments and see what works
best for you
Roland Siebelink:Okay. Very, very good. Definitely more of an indirect strategy, it sounds like,
where you're working also on the sales side more with partner organizations, if that's correct?
Milan Dobrota:Yeah. Again, the way we are reaching a larger number of smaller customers. There is a
bit of a customer journey and one of the important characteristics for us placing our technology on the
market is that by rule, each customer starts with a small proof of concept project, a small pilot project.
The technology is new. The adoption of these technologies is relatively new. They want to test it before
they buy it.
Sometimes, we have a company with whom we have just signed a $1,000 contract for a year, but with a much
larger potential - they're growing hundreds of thousands of hectares. It's all meant to convert them and
push them through this life cycle of a customer before we acquire them as a large customer. I wouldn't say
that it's the most important sales channel, this kind of partnership for us. It's probably more important
from the side of visibility and attracting potential customers to give a look into Agremo. And then we
take them further through direct sales activities.
Roland Siebelink:Okay. Yes, that makes a lot of sense. You get in with just a small commitment and
then when they start seeing how much the data can do for them, then it's really about starting to cover
more of their acreage, as you say, right?
Milan Dobrota:Yeah. That's a typical case, yes.
Roland Siebelink:How about financing strategy, Milan? Are you looking to raise more money at some
point in time or is the strategy more to just be profitable and reinvest in the company?
Milan Dobrota:Yeah. We actually had plans last year to raise the next round to support what I
previously mentioned. And then when the situation last year started to evolve, we postponed it. We want to
do it somewhere in the near future. To raise the next round of funding to support our further development.
This is something that is planned actually for the first half of the this year.
Roland Siebelink:Excellent. Very good. If any investors are listening and interested in the ag tech
space, then, of course, I'm happy to provide an introduction to Milan if you want to talk to him. And for
anyone else, what is the last tip that you would give, Milan, as an experienced founder by now, to people
who are a little bit behind you in the journey? What is the key learning that you have had as a founder
that you would like to convey to other founders?
Milan Dobrota:That's a good thing. And I'm always very careful with giving advice. One piece of
advice that I can give is not to take advice too easily, even from experienced and very successful
entrepreneurs. Not because they lack the ability to give good advice. It's because there is a strong
situational approach to each company. The startup is not the same if you're a startup in Silicon Valley or
in a developing country or in one space or the other. And what's your starting position? What is your
environment? Be careful with how you can project the advice or success of others to your startup. That's
And on the other side, it's just the importance of learning. Being able to learn from your specific
situation. Being able to build to become better than you were yesterday. That's probably the most
important thing that everyone should be having in mind.
Roland Siebelink:And that's an excellent way to finish this recording. Thank you so much, Milan
Dobrota, the CEO of Agremo, joining us from Belgrade, Serbia. It was an honor to have you on the podcast.
This was an amazing interview, Milan. Thank you so much.
Milan Dobrota:Thank you, Roland, and it was my pleasure to be here. Thanks.
Roland Siebelink:Thank you.
Roland Siebelink talks all things tech startup and bring you interviews with tech cofounders
across the world.