John Deere Leaps Unlocked Event Highlights and Analysis

Diving into autonomy, sense and act, digitization, talent, sustainability and the defensibility of Deere's business.

May 29, 2022

John Deere Leaps Unlocked Analyst Event - John Deere (Video)

On Thursday, John Deere had their Leaps Unlocked Analyst Day where they went through their technology and their vision for how it will benefit farmers and the agriculture industry in the future. Primarily, it was bringing to life products and concepts that they have talked about and highlighted in various other external facing documents (like their 2021 Annual Report) and interviews.

I watched the event and attempted to pull out some of the interesting aspects (there is a lot in it, it’s well worth the watch). Many of the concepts I have highlighted before, most recently in my overview of their See and Spray Ultimate Launch and 2021 Annual Report. I will try to emphasize some of the newer information, but highly suggest checking out the aforementioned link if interested in a more comprehensive overview of Deere’s technology endeavours.

Overview

John Deere puts out extremely well thought out and professional messaging. They are very pointed with what they want to have emphasized. One obvious one is that they say that there is $150 billion in total market benefit to be unlocked from their technology offerings, through decreasing applications rates of herbicides or fungicides or minimizing harvest losses. Said another way, that is $150 billion of savings to their customers that they can access through selling their technology, primarily Software-as-a-service. Extracting just 10% of that value created, or saved, would give them access to $15 billion in new revenue annually. They want 10% of their revenue coming from a recurring source by 2030. With 2021 sales numbers of $44 billion (this includes construction) and a conservative 5% CGR of total revenue, that means that by 2030 they would have $6 billion in recurring revenue, starting from a very small amount (that they don’t state) today. They see 40% of their revenue to potentially be recurring in the future as well, and if you look at the potential of unique part services in the future or financial products, this doesn’t seem far fetched on a long enough time horizon.

To put the service revenue into context - Apple as of 2021 had about 22% of their revenue gained through services (Apple TV+, Apple Music, Apple Fitness etc) at between 65 and 70% gross margins (which is fairly typical for software service driven revenue).

Thinking back to John Deere gross margins of 15% in 2020 and 20% in 2022, it’s interesting to think that if they execute on this SaaS strategy of 10% of revenue from services with a level of gross margin like Apple it could mean potentially 30+% of company gross margin dollars! Also notably “smoothing out” the cyclicality and inconsistency in their revenues:

Source

Finally, what’s notable, again sticking with the loose assumption that Apple as a consumer business bears some resemblance to John Deere as an agribusiness is that as Apple has rolled out their services, they have seen increases in loyalty and customer retention.

The other emphasis was their tech stack:

They have actively focused on owning and integrating the entirety of the tech stack - either developing or going out and acquiring, something I have talked about being a smart move and what I think will be core to driving Deere’s success. When customer success is so dependent on their experience, tightly integrating the technology and not relying on partnerships is the best way to execute in my opinion.

They broke out 3-hour event by numerous sections, essentially in alignment with their tech stack and capabilities. To keep it consistent I have loosely followed their event breakout below.

Digitalization

They started with digitization, making the statement that “connectivity is as important as sunlight” to farming. While there is some hyperbole there, the fact that Deere is talking about this with such emphasis illustrates it’s importance, even stating they are looking at putting satellite reception capabilities onto their tractors (eg: to access satellite internet connectivity such as Starlink). They currently have 300,000 connected machines, targeting 1.5 million by 2026 and currently have over 250 integration partners into Operations Centre, the most of any agriculture company really making them an important hub.

They have increased the number of “engaged acres” within Operations Centre, moving up from 315 million in their 2021 annual report, to a reported 340 million as of this week. This is well on their way to 500 million engaged acres by 2026 they are targeting. Engaged acres with smart machines creates a systematic approach to driving farm success. In my mind, this should go up and be hit simply based around equipment upgrade cycles of farmers.

What pleasantly surprised me for 2022 was the fact that they defined what an engaged acre is.

I have advocated for this within Upstream in 2021:

Deere now claims over 230 million “engaged” acres. There is no definition of what engaged means. A set definition of engaged acres would be beneficial in my opinion at least as a starting point.

This is an important step to understanding the relative digital position of organizations, though far from perfect (every organization won’t define it the same).

Here are the definitions of “engaged”, “highly engaged” and “sustainably engaged” from John Deere:

I like the fact that they have levels. It’s important for Deere to monitor their customers utilization of tools and systems. When looking at how well their precision strategy is working, seeing that they are increasing the engagement of their customers will be an important KPI to reinforce traction along with being useful to support upgrades.

There were two points emphasized: Operations Centre creating a digital twin of the farm and Operations Centre as the Operational and Agronomic Command Centre for the farm.

I’ll start with the latter positioning. Being the operational command centre makes total sense. The emphasis around agronomic command centre is new, though not surprising given the lynchpin that their equipment and Ops Centre is to execute on precision agronomics. In their last 3 annual reports “agronomy” (or other forms of the word) shows up just one time. It came up at least 11 times during the 3 hour presentation this past week. I’ve touched on Deere’s position agronomically through their sense and act technology and ability to influence agronomic excellence. Leaning into agronomy illustrates in my mind a proliferation into precision and agronomic services more consistently at their dealership level (more on that later) in the future.

The digital twin is a great aspiration. They are in a position to enable much of it. One of the challenges is manual entry of data. For example, today anyone who has looked at any farm management software information will know that the products used, like seed or herbicides, are often incorrect. This challenges not only the digital twin, but the aggregated roll up of data to support decisions that Deere talks about that many other companies have been working on for years as well.

There are some workflow aspects that can be developed through a nice plan to execution UX, but there are also opportunities for collaboration with crop input companies. Take this abandoned BASF patent application as a starting point of establishing a seamless way for accurate data through scanning technology tied into every input product package to ensure accurate data within Operations Centre. It seems obvious to me there is a need for crop input manufacturers and companies like John Deere to work closer together to enable the best possible outcomes for farmers. Whether we are talking about the digital aspects like above, or considering through how see and spray service offerings are executed on effectively (more on that later) to not only deliver seamless experiences for the farmer, but superior outcomes in areas where Deere doesn’t have the expertise and input companies need the access to the equipment enabling next generation practices.

Overall, the John Deere summary of operations centre being the “agronomic and operational command centre” is a great frame. Within that there will be Google Maps-esque approaches to the entire farm operation; suggesting specific practices to do at a specific time or understanding the time to complete a job or how many acres could be done before a weather event as well as dashboards for delivering decision making insights.

Share Upstream Ag Insights

Sense and Act

The Sense and Act capability is one of the most compelling areas of Deere’s future. Once you put camera’s and sensors and connect machines, the list of information you can collect and use becomes compelling. Before getting into the outcomes of the acting portion, the sensing portion can fundamentally deliver better insights into seedling mortality informing variable rate seeding for the following year or provide a map of highest areas of weed pressure informing a specific place to direct an agronomist towards.

The sense and act capability is “plant level management, at scale, with ease, in real time”. The sensing is one thing. Being able to act with precision in millisecond level fashion delivers an unmatched capability to overcome what borders on a wicked problem that is trying to optimize farm outcomes while managing massive constraints and interdependencies.

One of the insightful take aways around sense and act came from the Q&A period where Deere stated that they are focused on see and spray (weeds) and executing there, but in ~2 years will be looking launch a see and spray fungicide, then insecticide and then a sense and act fertilizer (nitrogen).

Weeds are an entirely different problem than the next two problem. Weeds have two dynamics that makes them more manageable:

They are stable. Meaning, a weed stays in one spot in the field and grows at a linear rate, plus once a crop gets to a certain size it typically can out compete the weed if it comes up after the application was made.
Reactive. Unless using soil applied products (which see and spray offers no value to), herbicides are applied in response to seeing a weed present in the field.

If moving to see and spray fungicide next, this brings an entirely new dynamic. Consider the following:

Fungicides (most generally) need to be applied proactively in order to be effective. Fungicides are better thought of like a vaccine (protection provided when used before exposure). Using a fungicide after you see the disease lesion, as implied by “see and spray”, means a much less efficacious outcome.
Type. There are two types of fungal disease. Monocyclic and polycyclic. Monocyclic infect once per year and tend to cause higher rates of losses, both in yield and quality. For example, white mould in soybeans or canola is monocyclic. Once you see a lesion occurring, the fungicide value has dropped immensely, in most instances much greater than 95% (read: rarely you would ever apply a fungicide after seeing the disease). Polycyclic diseases are reproducing every 4-10 days and moving progressively throughout the field and up the plants. Examples include septoria in wheat or anthracnose or rust in corn. If you don’t see it on a plant one day, it can be infected the next.
Movement. Fungus can come from within the field, but often spores move in with wind or animals for example. No disease today can change with one brief wind gust and a shot of rain.

We can assume if Deere wants see and spray for fungicides to be successful, they will not be reacting to the disease lesions themselves.

That leaves the reaction to be around susceptibility of specific plants or areas of a field. Looking at plant density, NDVI and overall risk (which might consider variety, row spacing etc). This has shortcomings compared to herbicide see and spray because the lever to pull on for fungicides to save money for farmers is around going from a high or standard fungicide rate to simply spraying at the lowest registered rate. There is an on off component as well, but not applicable every year or in every field.

They may be planning to incorporate some weather models to help predict disease as well, which could help the decision too. How variable rate fungicide is typically done today is by using NDVI imagery to evaluate high risk areas based on how robust the crop is and then evaluated physically by an agronomist in the field and then a zone map is created and uploaded to the sprayer, so still a lot of steps that Deere could eliminate. (If anyone at John Deere wants to bring me up to speed on their view I would appreciate it!)

Insects have similar challenges, where insects move, can hide under leaves and sometimes will not be able to be directly contacted, so an insecticide needs to coat a plant to be systemically active. There still may be some opportunities to leverage a better signal from the plants, such as my previous comments around connecting with InnerPlant.

The nitrogen focus is one that is interesting as Exact Rate or See and Spray, as Deere wants to decrease nitrogen use by 20%. They are not planning to go after this for a few years. While the reason driving it down as priority #4 in the sense and act priority list likely has to do with nitrogen being very complex, it likely also has to do with the why’s behind nitrogen applications.

As alluded to during their presentation, nitrogen in many regions tends to be applied to the soil prior to planting. Sometimes in the fall after harvest and sometimes in spring before planting. This lowers nitrogen use efficiency and is one reason we do see nitrogen efficiency rates no higher than 50%.

There is rationale to this though. Nitrogen is bulky, slowing down operations at planting significantly when time is the biggest constraint, so it gets applied prior to planting. It also has the highest demand at planting time, meaning prices are higher.

One of the big opportunities at Deere is autonomy, which decouples some of the time constraints at planting, allowing for more openness to apply N at planting. On top, if they could apply N in a more targeted fashion and increase nitrogen use efficiency through this, a farm would have to use less which may allow them to buy on farm storage to manage, or help digest the increased N prices they may have to take on. So while they figure out how to get their models trained for N and other nutrients, it also allows them to have a higher install base of autonomy which will help them have an increased use case for their N sense and act products.

The sense and act capabilities are what I consider the “killer app”. The sensing acts as data acquisition for the entire farm, and Deere’s entire business. It drives the entire fly wheel for Deere to continuously add value to the farmer and getting ahead of their competitors. In particular, the sprayer units are at the centre of this. The capability unlocks value and locks Deere into the centre of the farmers business.

Other interesting aspects:

The last sense and act capability is combine advisor, which Deere states can save 1% of yield through mitigating harvest losses.
Deere also did say they would have further retrofit solutions.
Deere is starting with see and spray in corn, soybean and cotton. However, they state that when the machine is used in conventional mode across other crops, it helps them acquire more data to train their models on other crops.
They plan to extend the geography from North America to LatAm and Europe as well in time.

Image of an as applied map from see and spray technology:

Share Upstream Ag Insights

Autonomy

I dove into autonomy in January when Deere had their autonomy announcement, so I won’t get too deep into it here. What was notable to me was the fact they want to sell all of their tractors in the future as autonomy enabled. This allows farmers to be able to “turn on” autonomy. This is not only beneficial to grow adoption, it also plays into their business models. I highlighted last week Deere comments around potentially charging by pass, which means for farmers that want to manage their equipment differently and just autonomy only in some areas, could pay a lower amount.

Deere had stated they want to deliver a fully autonomous corn cropping system available by 2030.

What else came up during the presentations was this model: Freemium.

One wouldn’t traditionally think of the same business model used by iPhone games like Angry Birds would apply to state of the art agricultural tractors, but there are likely some opportunities to try it to just get farmers using the autonomy to decrease that hesitancy and then monetize once the farmers derive value or are hooked into that capability.

Sustainability

John Deere emphasized their benefits to the environment at the event as well. They began priming this message on what specific value their solutions provide per acre and from a CO2 reduction perspective in their 2021 annual report:"

While Deere emphasizes SaaS business models and CO2 reduction, they do have the ability to be a driving force behind anything that comes out of sustainability and carbon programs. Empowering farmers with their data to plug that into a carbon program or biofuel program that allows the farmer to access a new revenue stream or increased revenue amounts could further

Given the above it becomes apparent that Deere can be the key point of integration for any and all companies looking to be aggregators in carbon programs, so I expect more to come from John Deere on this front, even just looking at the last points on the above image there is a CO2 emission reduction numbers that is likely strategic to their long term initiatives.

I don’t anticipate them being a carbon aggregating company themselves - too competitive, too much risk and not enough upside and not their core competency but they can be an enabler for their customers to seamlessly connect into any program (sustainability, IP, biofuel etc) they want with their data making John Deere better off focusing on collaborating with larger organizations. Essentially, they can play a pivotal role in MRV.

They also talked about “soil health sensing”. I actually read an interesting study this week around soil health vs. physical soil indicators that identified that physical soil indicators were much more highly correlated to yield and quality in corn. So while they may be emphasizing soil health, really being able to understand chemical and physical attributes of soil in real time would be incredibly valuable to their customers

Talent

I loved the fact that Deere dedicated an entire segment to their talent strategy. As a company trying to reinforce their image as a technology company, it was fascinating to listen to how they think about talent, how they attract and why that’s important.

While unrelated at first glance, it actually is incredibly smart when it comes to the talent strategy: Who Deere invites to events like their Leaps Unlocked day.

There are of course the banking analysts, mainstream media and agriculture media. But, if you look at Twitter, the primary people tweeting about the event are technology analysts. Individuals that would traditionally being looking at consumer technology products or thinking through the business of big tech companies like Microsoft. This at a surface level likely supports Wall Street further, but it also creates awareness to young technology savvy individuals reading their newsletters or blogs who might not have been attracted to Deere otherwise as developers or technology strategists.

This talent strategy obviously builds great products, but it also builds better distribution.

To dig into the talent and distribution connection, I’ll highlight insights from three friends of Upstream: Janette Barnard, Matt Coutts and Patrick Hancoop.

This week Janette did an exceptional job of highlighting and building on insights Matt brought up on a podcast breaking down the John Deere business.

The boring key that unlocks super powers? Distribution - Prime Future

In this edition Janette establishes a clear breakdown of what distribution is:

Distribution is sometimes reduced to the delivery of the physical product, but there are three relevant functions tied (in varying degrees) to distribution:
Customer acquisition & demand creation.
Physical movement & delivery of a product or service.
Customer relationship to support anything post-sale.

Deere has one of the best distribution networks in all of agriculture as Matt points out in his appearance on the Business Breakdowns podcast.

Where they shine specifically is in what Janette has listed in her 3rd point: customer relationship to support post sale. Tractors break down. Time is often the biggest constraint of farmers. Support post sale is a huge differentiator. What support will look like in the future is going to be more and more technology driven. Now, it isn’t Deere delivering the support, it’s their dealers. But what Patrick, an industry equipment expert, has mentioned to me is that Deere puts the expectations on their dealers to have knowledge to be able to support the technology more emphatically than other companies traditionally do.

Deere doesn’t just expect it though, they lead by example in attracting and training a talented work force and support their dealer network with access to these experts. This allows their network to be differentiated even further.

This gets into what makes Deere extremely successful; they stack the deck in their favour well beyond distribution.

This is apparent when we look at my favourite way to assess strategic advantage: The 7 Powers Framework

There are 7 Powers according to Hamilton Helmer that deliver differentiated returns and defendable business:

Switching Costs
Cornered Resources
Process Power
Branding
Counter Positioning
Network Economies
Scale

And Deere meaningfully hits on every single one in some way!

I put a Deere logo in every “power” they have, eventually illustrating they angle in on every single one which shows why a company like Deere has been able to be successful for over 185 years.

At a basic level:

Through Operations Centre they increase switching costs making it more challenging to go to another colour brand because the customer experience and insights are second to none, they have worked on cornering tech talent from a resource perspective, they’ve built best in class operations to produce their machines through process power, their brand is one of the most powerful in all of business (not just agriculture), they have begun working towards a counter position by establishing a new business model (SaaS), through their network of dealers and their connected machines they have network effects and they have the scale to keep their manufacturing and distribution costs low. This all leads to better margins, easier customer acquisition, better customer retention and a deeper understanding of their customers to build further new and innovative products.

Deere has been ahead in many ways (look at revenue growth, margins, stock price etc) compared to their biggest competitors and because of their foresight to dive head first into technology they have actually positioned themselves to move even more meaningfully ahead because it will in fact speed up their flywheel and the opportunities in their business.

I didn’t even get to touch on everything from John Deere, which I think illustrates just how much is going on at Deere and what to expect in the coming 5 - 7 years from John Deere, which is exciting for farmers and the industry overall.

There will be lots to watch in how Deere executes, how their main competitors react and how companies like crop input manufacturers take action.

Note: I didn’t touch on their electrification strategy, but it is very consistent with what I wrote on the electric tractors at the end of last year surrounding an emphasis in the small tractor market - for specifics, highly recommend watching the event link.

Share Upstream Ag Insights