Upstream Ag Insights - February 5th 2023
Essential news and analysis for agribusiness leaders
Welcome to the 154th Edition of Upstream Ag Insights!
Index for the week:
It’s Time to Call It
Follow-up on Biostimulant Margins
Valent BioSciences Announces the Acquisition of FBSciences
2023 Digital Advertising Benchmarks For Agriculture
Enko Completes its $80M Series C to Progress Sustainable Crop Solutions
5 Ways ChatGPT Will Change Agriculture
Bayer to Distribute Pheromone-based Biological Crop Protection Solutions from M2i Group
Beware Flavored Software
Thanks for reading, sharing and subscribing!
1. It's Time to Call It - Prime Future
My friend Janette Barnard started out her wonderful newsletter this week with the following:
Riddle me this: What do you call a category of companies that raised a ton of venture capital and, a decade later, had not one sustainable business to show for it?
There's (a) category that is 10+ years old, and a post-mortem is timely because, well, it's basically commercially corpse-like.
The category is farm management software, the row crop genre.
Janette does a great job breaking down some of the dynamics and what we can learn from the first decade+ of farm management software. And one area she mentioned is what I have been kicking myself about for a while now:
My hypothesis is that founders of Agtech 1.0 companies, and investors, had the hypothesis that farm management was a winner-take-all market. If you believe that only 1 or 2 players will dominate a market, then it is logical to invest aggressively in growth in order to be one of those winners.
But few markets are really winner-take-all.
In an industry such as farming where the potential user base is so diverse, their needs are so diverse, their business structure and profit margins are so diverse...the pie is so varied that it would be difficult for any one company to take the entire market, simply from a capability standpoint.
I am a culprit of having thought of the space as winner-take-all initially. I was wrong and have been kicking myself about my incorrect assessment for a couple years. Having sat in many meetings where the assumption was that stand alone farm management software would have just a few winners, I’d suggest that Janette’s assumption is accurate.
A winner-take-all market is one where the top players capture a disproportionately large share of the rewards, while the rest of the players are left with extremely little. In a winner-takes-all market. Therefore the incentive becomes to spend to acquire customers rapidly.
I’ve thought about where I went wrong in assumptions and to build on Janette’s comments I think there are a few learnings from the challenges in that area of agtech that can help ensure those same assumptions don’t get repeated.
The first thing to consider is what underpins a “winner-take-all” market?
The answer: Network effects.
Network effects are the incremental benefit gained by an existing user for each new user that joins the network. Every additional user to the telephone creates a stronger network effect. Same with Facebook or any social media.
The assumption in the farm management software space was that as more acres were gained on a platform, that would mean more data accrued leading to a deeper understanding of that customer and therefore deeper value delivery to the farmer which would mean more data for the organization to add value to other farmers which would increase switching costs of the farmer and make it difficult for any other farm management software player to break into the market. Farmers in theory would derive so much decision making value they would gladly play $5, $7 or $10/ac to access this. A winner-take-all scenario.
Because of these assumptions the emphasis became customer acquisition. Hundreds of million (billions in total) poured into pulling customers onto the software.
What’s important to note is that in order for a network effect to take off there needs to be utility for the user and ideally a tight feedback loop. Think of Google. Everytime you search on their platform you derive utility, almost immediately. And Google has a tight feedback loop surrounding time spent on the page, what you clicked on etc. ultimately enhancing their offering and making them better which means you keep going back to Google. The externalities influencing the system are close to none and there are only so many parameters that can influence the outcome.
Now if we think about this from a farming perspective, we get the opposite. The first issue is that the farmer has an arduous onboarding process and not only doesn’t derive immediate benefit, they might not derive any benefit for months, or years. This stems from data challenges and disparate connectivity issues.
Data is very clean, contextualized and connected in systems like Google. The value derived from using Google is immediate (I get the right website). In farming, we get uncalibrated, incomplete data and often manually input data that is inconsistent. This doesn’t even touch on the lack of depth of data such as missing soil information, or nuanced complexity surrounding the interactions of soil, weather, fertility and agronomy information just to name a few. Then factor in that not only is it complex, but it varies by region because the agronomic outcomes are localized and variable. This just scratches the surface of some of the challenges. The end result is a difficult time to deriving value.
This reality challenges significant value creation for the farmer, or agronomist leading to challenged business models and few network effects. That equals no winner-take-all.
To be clear, just because there is a weak winner-take-all dynamic doesn’t mean farm management software is irrelevant. People often say that farm management software is a solution looking for a problem. I disagree. Digitizing aspects of the farm and having a digital system of record for the farm is valuable. But the question becomes one of business models, total value created and value chain positioning to deliver that value in a positive way for farmers. There are sparing examples of success, limitations on scalability and the need for an alternative business model (eg: use the digital systems to augment the selling of physical goods). One can look at an organization like John Deere who boasted >350 million “engaged acres” in a 2022 report of theirs and we can see JD monetizes that portion of their software in a non-direct way and integrates it into a required physical asset used by farmers (equipment).
Of the agtech 1.0 farm management software groups as Janette calls them, we have still seen a lot of these independent FMS groups remain alive, however, we have witnessed these groups rapidly pivot towards areas like carbon management, MRV, marketplaces, layering on analytics capabilities or prioritizing ag retails as their go-to-market approach which has a fundamentally different total addressable market and solves solves different problems. In my mind this delays inevitable bankruptcy’s, it doesn’t eliminate them.
Of groups that fall into this space one stands out to me: TELUS Agriculture. They have acquired numerous FMS companies in an effort to roll-up and connect the data downstream. The sentiment many have referenced to illustrate their bullishness on this effort was the fact that TELUS has a long time horizon and they had done similar things in health care. Given acquisition price tags, current revenue splits between upstream companies and downstream companies in their portfolio and the fact they do not have a strong anchor point in the value chain/product integration (unless they can somehow do this with their mobility offering or something like farm wide IoT for example) it will be an uphill battle to deliver more value to shareholders and deliver on the vision they talk about in the market.
I don’t think anything stated here is novel, but it illustrates and reinforces some of the challenges we see with many stand-alone agronomy companies in the analytics space of the industry, as analytics as a stand alone become commoditized because the bundling of analytics and assets is where value accrues (eg: John Deere). The highlights surrounding network effects also reinforce the challenges with marketplace business models in agriculture (which rely heavily on network effects to be successful).
FMS aren’t irrelevant, they just aren’t winner take all which means they should have had a different investment and prioritization strategy. I am certain I am not the only one who wishes they recognized this before 2020.
Related: Taranis Introduces AcreForward, Setting a New Standard for Crop Intelligence - PR Newswire
2. Follow-up on Biostimulant Margins
Last week I mentioned news around ADM getting into biostimulants could mean the start of a decline in biostimulant margins. The comment wasn’t fully explained and drew questions in from readers so I want to ensure I share where my logic stems from.
First, it’s important to note I was primarily meaning retail margins. However, as products become more mainstream, more competition and “me too” (we think of these as generics in the crop protection world) products arise which can lead to lower margins as well at the manufacturer level (this means programs aren’t going anywhere).
To better explain my thought process is that on top of the challenge with a company coming in without strong specialized people and intellectual support for the product usage, my experience has often been that margin declines with increased product adoption:
Demonstrating value to early adopters gives organizations disproportionate opportunity to capture new revenue and higher margins.
As products become more mainstream, we are attempting to sell proven products, services and systems. Examples include herbicides or bulk fertilizer leading to increased retail competition across products that are commodities: their value is known and everyone is capable of supporting them. This leads to a pricing mentality among customers and a deeper competitive landscape across the market which means eroding margins.
According to surveys, we can generally say adoption of biostimulants is not quite across the chasm yet, so I am ahead of myself to be sure. But when we start seeing organizations move into a previously value added space without the intellectual infrastructure and investment to differentiate, we are surely at the peak of what margins will be in retail and at the beginning of new bundling initiatives in manufacturing to keep margins where they are for longer. On top, as I highlighted in March 2022, there is a pointed emphasis from more and more ag retail organizations to bring bio products in which shows the adoption and utilization will continue to grow.
3. Valent BioSciences Announces the Acquisition of FBSciences - Valent Biosciences
Valent BioSciences LLC, a global leader in the discovery, development, and commercialization of highly effective, low-risk, environmentally compatible technologies and products for agriculture, public health, and forest health, announces its acquisition of FBSciences Holdings, Inc., a leader in the discovery and commercialization of naturally derived plant, soil, and climate health solutions. With this acquisition, Valent BioSciences and its parent company, Sumitomo Chemical Co., Ltd., now offer an innovative and proven portfolio of integrated biorational solutions, including biostimulants, biopesticides, and crop nutrition solutions.
Another acquisition in the biological space of what surely won’t be the last one of the year. I have been suggesting along with many industry observers and professionals that we will see another consolidation trend in biologicals.
This announcement follows the October 2022 announcement that Valent was launching a biorationale unit.
Overall, I think the acquisitio makes sense given trends and the Valent business itself. I am a bit biased though. If I was guessing, FB Sciences biostimulant products would be the bio based product I recommended most frequently when I was an agronomist. One of the most commonly available biostimulants in Canada when I was an agronomist was FB Sciences “Transit”, or PPA (polymeric polyhydroxy acid) so I am very familiar with the company and products.
What more specifically is driving the consolidation trend in biologicals?
To better explain what is driving this consolidation trend I think we can highlight a few pillars:
Macro trends and Industry head winds (EU) - If we look at EU regulations leading to challenges in bringing new active ingredients to market for agribusinesses along with corporate aims and expectations to deliver lower environmental impact products to market, it makes sense that companies want to bolster their internal capabilities in a non-traditional space.
CAGR and Margins - Depending on the market research you want to cite, there is generally a 9-18% compound annual growth rate expected in the biological space. This fuels the “self fulfilling prophecy effect”. Couple this with the fact that margins are generally higher currently in the bio realm (as we can see from Corteva acquisition of Stoller) and we see demand for biological companies.
Lack of intellectual understanding in traditional agribusinesses - If biological products do take off, most of the big 6 agribusinesses (and others) are ill-equipped relatively speaking with expertise and intellectual property. Their best bet is to acquire to rapidly jump up the learning and IP curve.
Breadth of what is biological - Biologicals are broad. From biopesticides, to biostimulants to biofertilizers and more. This means a desire across various players in the ag value chain to grab a hold of companies as well as opportunity to integrate numerous disparate kinds of biological products and products together. This is where the private equity opportunity lies in the space as well.
Lastly, there are still performance inconsistencies in biologicals. A large scaled org with greater market access and resources allows the ability to better find the right fit for products and if we look at large fertilizer or chemistry and seed organizations, they are very good at putting in the work to understand where the right fit is for products, along with having numerous options for novel bundles.
There are caveats and challenges, but it seems highly likely that biological companies remain acquisition targets in 2023. And there are strong businesses out there that deliver opportunities for large organizations. I highlighted Marrone Bio and Stoller as likely acquisition targets in 2021 and in 2022 and they got acquired; two other organizations that I’ve highlighted remain independently owned are groups like Verdesian Life Sciences and Acadian Plant Health.
Related: Bioinput market grows 67% during 2021-22 crop in Brazil - Agro Pages
4. 2023 Digital Advertising Benchmarks For Agriculture - Think Shift
Think Shift put together a great summary of marketing costs within the industry which is worth checking out.
A nice summary of the numbers can be seen here:
But as I read this I couldn’t help but note the Objectives emphasized were Traffic and Awareness. No doubt these are objectives in the world of marketing today, but no business in agriculture (outside ag media) makes money from website traffic or awareness.
Unequivocally, agribusinesses make money from selling goods. Not impressions. Marketing dollars that directly generate revenue are the best expenditures of marketing dollars. There needs to be a better answer to marketing for an industry that generates hundreds of billions in transaction volume annually supported by hundreds of millions in marketing budgets.
The real objective is and will always be revenue generation. Where is that metric measured? It’s generally unattainable. Until now.
As AgVend* pointed out, that means the real value and real opportunity is in a marketing channel that is connected to the transaction!
When a marketing effort (and expenditure) is connected to the transaction there is the ability to measure the true performance of the spend. This enables future KPI targets like “revenue generated per marketing dollar spent” in a campaign or “new revenue generated per dollar spent”. Marketers today can subjectively come to these numbers in aggregate, but they are unable to associate specific dollars with specific outcomes.
AgVend CEO Alexander Reichert commented on the Linkedin post astutely pointing this out:
"Half the money I spend on advertising is wasted; the trouble is I don't know which half.” - John Wanamaker
Attribution is key. And the best way to drive those numbers higher is hyper-relevant targeting and content. This is becoming harder to do anonymously due to changing privacy guidelines (think: the end of cookies) so having first-party data (ie. contact info and consent) will be crucial for successful digital marketing in the future.
Digitally enabled ag retailers are in the best position to take advantage of this trend. Exciting times!
That is spot on.
Not only is attribution being connected to the transaction key, but as he alludes to having the information and consent (1st party data) is incredibly powerful.
We can take this further: When you have consent and contact information through a trusted channel (the retailer) and you are connected to the transaction you can also identify previous transactions meaning now you can be even targeted to the point of transaction and behavioural history which layers on top of contact information and geographic information. This delivers unique opportunities for in-season marketing capabilities.
This all leads to better results for the agribusiness doing the marketing, new opportunities for ag retailers and more relevant materials directly in the farmers hands.
AgVend Network Marketing delivers the first avenue for input agribusinesses to align with their ag retailer partners to target specific customer segments and be able to identify not only what the views, click thru and open rates were, but also what farmers purchased and how much they purchased from the campaign, delivering novel performance metrics like new revenue generated per dollar spent.
Undoubtedly the best avenue for generating improved marketing ROI in the coming years in my mind will be enabled by AgVend via the retail channel focused on true performance and outcome driven metrics.
Disclosure: I am employed by AgVend.
5. Enko Completes its $80M Series C to Progress Sustainable Crop Solutions - PR Newswire
Enko, the crop health company, today announced that it has raised additional funding from Eight Roads Ventures (a global proprietary investment firm backed by Fidelity), Nufarm, Endeavor8, and Akroyd LLC.
This is an extended $10 million onto the companies July 2022 Series C of $70 million, bringing the total capital raised by Enko to about $150 million.
Since its start in 2017, Enko has generated hundreds of leading molecules across all categories of crop protection through its ENKOMPASS™ technology platform, which combines DNA-encoded library screening with machine learning and structure-based design to quickly find new, better performing and more targeted chemistries. Enko's product pipeline is currently led by a range of herbicides that are breaking cycles of weed and pest resistance and are being developed to reduce environmental load compared to current solutions.
The cost to bring a crop protection molecule to market is on average $286 million. While this cost pales in comparison to the estimated $2.6 billion to bring pharmaceuticals to market, it is still extensive. Enko's approach has generated discoveries in roughly half the time and with fewer resources than conventional R&D methods.
Bringing the time down will bring down the costs in part too. Enko is a fascinating company that enables organizations to manage their R&D and produce more environmentally friendly chemistries as well.
Enko is endeavouring to move the crop protection industry forward amidst lower R&D expenditure from large agribusinesses (as a percentage of revenue). Enko has been collaborating with the likes of Syngenta and Bayer on promising new chemistries.
There is constantly talk about biologicals and we will see that continue, but chemistry is not going anywhere anytime soon. There is big opportunity in rapid identification of novel molecules that include higher standards for environmental impact and lower dosages will improve the type of active ingredients in the market place and support unique combinations of chemistries/biologicals. It’ll also be important to manage issues like resistance.
Less R&D expenditure from crop protection and seed companies compared to 15 - 20 years ago has been a constant trend. There has been more emphasis from crop protection companies on share buybacks for example, attempting to attract investors and bolster returns rather than drive long term innovation.
If we go back to the biotech hay-day when R&D was consistently around 12-13% of revenue (vs. closer to ~9% today), there were very explicit areas of biotech and traits that were going to show good returns and competitive in-roads in the market place.
While we still have CRISPR and other synthetic bio tech around being looked at by companies, it seems there has been less emphasis internally and more looking externally to nimble start-ups investing in tech hoping there is a break through. We can even look at Bayer recently getting out of biological R&D and outsourcing it while focusing on the commercialization. This is further reinforced in looking at the lead investor for the Series C: Nufarm.
6. 5 Ways ChatGPT Will Change Agriculture - Precision Farming Dealer
According to Reuters, OpenAI's ChatGPT is now the fastest-growing app in human history, reaching an estimated 100 million active monthly users in just the two months since its November release. An average of roughly 13 million unique visitors were using ChatGPT per day in January.
In the heading link, Todd Janzen points out some of the shortcomings of ChatGPT today. With that said, there are significant opportunities in using ChatGPT for the general industry professional and a more interesting opportunity in using the ChatGPT API that could then be trained on your own company and industry specific data.
There are only a few examples of direct applicability to agriculture today. The key agriculture angle is how it can augment agribusiness professionals:
AgriMarketers wanting to create basic blog posts, copy ideas for a product or a script for a radio placement or podcast. It could be leveraged by sales reps wanting to generate email templates, or senior leadership wanting to streamline an email on change management. General augmentation of the individual increasing the efficiency of administrative tasks and bolstering creativity and ideation. I have used ChatGPT almost daily now for two months and just like the individual that uses Google effectively has an upper hand in knowledge industries, the individual leveraging ChatGPT effectively (through more effective prompts or requests for example) will have an upper hand in efficiency and creativity in the long run.
ChatGPT is a large language model which is a deep learning algorithm that can recognize, summarize, translate, predict and generate text and other content based on knowledge gained from massive datasets. This is also known as Generative AI. If we consider the definition we can infer that if a large agribusiness was to have an API connected into their internal systems the ChatGPT model could be trained on their internal training documents, transcripts, e-mail interactions and more. This is easier said than done, specifically at smaller organizations, however, it presents opportunities to build better training materials for staff regarding agronomics or sales techniques, or even integrating basic self serve intelligence tools for farmers regarding things like tank-mix order that can then be integrated into a digital tool. When I look at the winners in the industry of those using AI at large, it will be those that are not only applying novel ML algorithms, but accessing novel data and integrating it to differentiate themselves in their go-to-market and product offering. There are numerous other examples even towards the farm in translating for hired helpers on the farm.
While these do not come across as game changing, they are small examples that act as entry points for large language models in the industry.
Will large language models eliminate jobs or solve labour problems? Will they become incredibly beneficial to the industry?
In 1850 farming made up 3 in 5 U.S. jobs. By 1970 that number was less than 1 in 20. Today it is less than 2 in 100. Ag technology single handedly eliminated many jobs; from tractors, to pesticides, to seed breeding and more eliminated more jobs than any industry on earth!
Capitalism aims to convert ambition to better outcomes. Our desire for wealth, belonging and a host of psychological comforts drives us to work and build. Capitalism helps incentivize and find new avenues to drive that.
Technological innovation is what keeps this moving forward. Each technological advance unlocks new opportunities; new goods, better ideas, novel approaches. And that economic expansion creates a flywheel of future opportunities.
From the printing press, to the mechanization of vehicles to computers, humans have always feared jobs becoming obsolete. And it has never happened. Jobs have been destroyed and others arise.
Technologies tend to cause employment instability for only short periods. The market always adjusts itself around the innovation, and job growth increases from there.
When the car was introduced, people fixated on the horse mobility business being eliminated. What is unnoticed is how the economy coordinates like a clock work to mold itself around that new technology or approach to create new jobs and new businesses.
The first Ford car was introduced in 1903 with the first Model T being introduced 5 years later in 1908. Look at an old picture of a US city in the early 1900’s:
We see no side walks, horses and carriage, no street lights, no fast food joints.
Today, a random image of any metro area illustrates to us the shear volume of new innovations, approaches, businesses and jobs that have been created because of the vehicle:
Street lights, gas stations, oil change spots, rental car businesses, fast food windows and more…all thanks to the vehicle! That’s a lot of jobs and a lot of opportunity. That doesn’t even consider more obscure examples, like F1 Racing and much more!
My point to this long winded ramble is to illustrate that even though ChatGPT and Generative AI at large might not be obviously directly beneficial to agriculture today, it is likely to grab hold and change the industry in small ways, leading to opportunities for innovators to approach problems in novel ways that are open to trying new things and patient in the beginning with a lack of immense value.
While I do not think ChatGPT will have the same impact on agriculture as the vehicle on society, the example provides a base example and framework for seeing how one small change can compound on itself into a host of different technology, busineses, opportunities, and ultimately new jobs and ways of doing things.
7. Bayer to Distribute Pheromone-based Biological Crop Protection Solutions from M2i Group
Bayer announced a partnership with French company M2i Group to supply fruit and vegetable growers around the world with pheromone-based biological crop protection products. Through the agreement, Bayer will become the exclusive distributor of select M2i products targeting lepidoptera pests in crops that include stone and pome fruits, tomatoes, and grapes.
I don’t normally highlight horticulture crop news, however, I find pheromone products to be notable across all types of agriculture. To be clear, the applicability is nuanced in large scale row crop farms (field size matters to the viability and economics) but these initial efforts in permanent crops are important.
There is a key aspect here in that pheromones pair with digital monitoring and analytics capabilities:
Bringing together M2i’s pheromone products with digital monitoring applications and Bayer’s existing biological portfolio will enable growers to apply the right product at the right time within a holistic approach that benefits their operations.
My friend Lee Vetsch of Ospraie Ag Science also commented on this news in his newly launched newsletter, Agriculture’s Honest Take (highly recommend checking it out and subscribing as he is always full of great insight!), where he astutely points out the concept of commoditizing your complement in this context:
In other words, pest analytics has the ability to increase the value of crop protection products. Therefore, it would be in an agrochemical company’s best interest to commoditize the service of pest analytics.
In the example of insects, commoditized analytics disproportionately have the ability to increase pheromone based products because they need to be applied in anticipation of insects. (Note: If you remember back to the first story of this edition on FMS you’ll notice that this commoditize the complement was the opposite of what agribusinesses were trying to do with analytics and farm management software initially, however, have shifted more in this direction).
Note: For more on the concept of commoditizing your complement applied to agriculture, see one of the most popular Upstream articles of 2022: John Deere to Crop Input Companies: “Your Margin is My Opportunity”
Pheromones in agriculture have a ton of potential because they can be used to disrupt mating patterns of insects, eliminating the risk of insect damage before any of the potential problems “hatch”. Pheromones flip the traditional “threshold” for action, or spray once you see a specific damage level or number, on its head, potentially eliminating any feeding damage.
We have seen other companies like FMC acquire pheromone organizations like BioPhero and groups like BASF Ventures invest in Provivi. Even smaller companies like Semios are using these technologies today.
According to various pheromone market reports the global Agricultural Pheromones market is expected to grow from $2.6 Billion in 2020 to $10.61 Billion by 2030.
In fact, when FMC acquired BioPhero, they stated the following:
FMC expects pheromones and pheromone-based products to contribute approximately $1 billion in revenue at above company-average EBITDA margin by 2030
That’s significant revenue for a company doing around $5 billion today. It’s also notable that they expect higher than average margins. This reinforces that it isn’t surprising to see Bayer look to get more meaningfully involved in the space.
I think if this market growth becomes real and FMC delivers $1 billion and groups like Bayer deliver success as well, we have to see some new business models developed where the pheromones are integrated into a more holistic system. FMC for example, has their own insect prediction tool in Arc and they also have relationships with remote insect detecting companies like RapidAIM in Australia. If you begin to combine this data with underwriting and insurance capabilities, there becomes an interesting new service that FMC can build out.
Related: Bayer and Kimitec join forces to bring the next generation of Biologicals to millions of growers worldwide - Bayer
Non Ag Article
8. Beware Flavored Software - Every
If you’re building the tech equivalent of balsamic strawberry ice cream, don’t expect vanilla scale
Other Ag Articles
From Farms to Incubators: Women Innovators Revolutionizing How Our Food is Grown - Farms to Incubators
Future of Fertilizer... So What? with Tim Hammerich from Future of Agriculture - Tenacious Ventures
Lavoro Announces the Acquisition of Cromo Química - Business Wire
Who were the most active food and agriculture technology venture capital firms in 2022? - AgFunder News