I have another Agri-Fintech issue pending but the company I am profiling have some interesting developments in the background so I want to let those ripen.
But do stay tuned for that one as it gives me a chance to consider the Full stack versus Ecosystem strategies in the context of a large Agritech company's decision making. 👀👀
In the meantime, I wanted to put an article out that has been on my mind for some time: Agricultural Data as an Asset.
✅ The Bayer and Microsoft announcement has really shaken things up in the data space - or at least promises to;
✅ Precedents of paying users for data are being set - and gaining traction - in others sectors such as advertising and are creeping into agriculture;
✅ With soil carbon, I would argue that the first "digital asset in agriculture" is already being produced and sold.
What do you think?
Below I have drafted some notes, which is certainly no expert view, but, I hope, an up to date piece on where we are, considering:
1️⃣ Data as an Asset
2️⃣ Agricultural Data
3️⃣ Data Value and Exchanges
4️⃣ Some thoughts and questions
Data as an Asset
Data is ubiquitous in Agriculture.
But data as an asset with some economic profile attached to it? That is not so common.
I think we can all agree data is a valuable commodity that users of digital services produce. Take Google for example. Google process 72,000 searches per second or 6.2 billion per day. Youtube, another Google company, has 77,000 videos played in any second, which gives off even more data, that feeds the Google algorithms and fuels their gigantic advertising business.
Brave Browser is offering an alternative to users by stripping ads from its service and replacing these with its own ads, which it pays users to see. I installed this in June and have accrued over $15 of tokens - I'm definitely not retiring any time soon, but the fact this is up and running is pretty sweet!
In agriculture, you can consider that any time a digital service is used some data is logged - it is not necessarily used, analysed or sold, but at least logged. So maybe something similar to Brave is possible?
The Data value chain
I also like to think of data akin to any other agricultural commodity and love using Corn as an example.
Can you imagine, growing just 50kg of corn (equivalent to about 2 bushels in the US) and then taking this to market? The value will be quite low.
That is the same with a small volume of data. But aggregate, grade and clean it and it starts to add more value the more processing and aggregation that occurs.
I capture this dynamic below - the more processed data there is and the more of it there is, then the more value it can capture.
Oh - and we provide most of the input data for free to the Googles, Facebooks, Amazons etc, who now make up the most valuable companies in the world 😀.
For example, revenue for the data economy in Europe was estimated at €65bn in 2017. It has basically been free money for those who captured the early network effects. The European Commission estimate the data economy could be worth a staggering €829bn by 2025 😳.
Tick this box to agree to our promiscuous data policy. ☑️
👆This is not something we are likely to see in Food and Agriculture.
Privacy is normally a feature of most Agritech offerings but similarly companies are moving to become data businesses and not just an Ag business. As early as 2016, the former Monsanto CTO, Robert Fraley, is on record:
Monsanto executives are seeking to reposition the company as a business built on data science and services, as well as its traditional chemicals, seeds and genetic traits operations
This captures the essence of the Data era of Agriculture and it can certainly be seen in the current Bayer announcement (who acquired Monsanto in 2016).
Dr Terry Griffin [@SpacePlowboy on Twitter] is one leading thinker in this space and the summary paper from 2018, along with his academic partners "Big Data in Agriculture: A Challenge for the future" is highly informative and where I have sourced the Monsanto quote from.
One key concept I loved from this piece is the idea of "Small Data" in agriculture, naturally, the direct opposite of the too familiar "Big Data". This small data is typically the data that in-field agriculture produces (soil sample data, yield data, aerial images, weather data as examples) but unfortunately becomes "isolated to the fields where the data originated" writes the article. Naturally, pooling these "Small Data" sets can create huge beneficial insights when aggregated.
But data is Intellectual Property. So naturally, that leads to a question of incentives and value.
Data Value and Data Exchanges
It's probably time to introduce some of the companies in this space to frame this discussion.
Farmobile - "Trust is the new currency in Agriculture"
Farmobile are probably the most widely known example of a data exchange in Food and Agriculture and have been in operation for some time in North America. In 2016, they offered a group of farmers $2 per acre for data and followed this up in 2017, by offering $1 per acre for their data, which is an impressive stamp of intent.
I would classify Farmobile as selling "Small Data" as above. They operate the 'DataStore Exchange' to facilitate trade of data between farmers who can share licensed and certified copies of their agronomic data to approved buyers. They envision
a futures market for data, wherein buyers can ask sellers for standardised data sets such as commodity types, planting dates, harvest dates or total acres planted.
The company has developed several patents for blockchain related infrastructure for food and agriculture in both the US and Canada according to their site. Altogether, this sounds fantastic.
However, the company acknowledges the capacity for conflicts of interest where multiple roles are played within the ag data value chain.
The data access lines become fuzzy in situations where the trusted advisor also serves as a retailer selling inputs that require farmers to use proprietary data collection devices in order for retailers to receive hundreds of thousands — even in some cases millions of dollars — a year in “rebate programs”.
Hmm. I think this is critical. And obviously so do Farmobile. There is an excellent post on their site, titled "Trust is the new currency in Agriculture" which is worth a read.
Independence and governance go hand in hand with success in sensitive markets and must be demonstrated in more regulated financial markets.
Data will be no different, especially if the platform moves towards a futures style market.
Earlier in 2021, Farmobile were fully acquired by AGI, the listed Canadian Agribusiness which is steeped in both physical and digital infrastructure. The Farmobile combination was seen as moving AGI "into the data verification space required by the rapidly developing carbon markets."
When I spoke to Farmobile to find out more, they seemed keen to pursue a broader approach to market data, but admitted the growth in the business due to carbon and other sustainability initiatives had caught them by surprise. This seemed to be the immediate focus, but they were still keen to develop further use cases.
Agrimetrics - "realising value trapped in Data"
Agrimetrics are a UK based business also with a data marketplace offering.
They have published various datasets - which I would classify as "Big Data" as referred to above - and some analytics that have already been done on top of the data.
When I spoke with them they clarified their offering as;
- Private data exchanges for some actors, including AHDB (Agriculture and Horticulture Development Board in the UK) in the UK, which facilitates internal data sharing;
- Services which allow better yield and harvest prediction modelling; and
- model hosting services, for users who want to develop best in class models for internal or external consumption.
Their key value proposition is linked data. Within their data sets they have already joined together livestock data with field data and weather data for example. This is impressive as not only is the data standardised but it also suggests the adjacent datasets for users - see a depiction of the Livestock example below.
According to the company, their user base mostly consists of the large agri-chemical companies, Agronomy advisors /consultants and Governments.
Interestingly, they are also moving to target financial users such as the large grain traders with their prediction model services and they predict much more interaction with sustainability programmes across the agriculture and financial industries.
The other reference point I have is an entity called Ocean Protocol - a Web 3 protocol for trading data assets.
I won't cover Ocean in much detail here, but if you are interested in this topic, I would recommend checking out Ocean and exploring their mechanics.
Even if you are not an avid crypto fan, they publish fascinating content on the data economy and on a monthly basis publish proposals they receive as use cases. Maybe we will even see some agri projects on there soon?
Some thoughts and questions
I think there are a few things which will be interesting to watch in the next few years as this develops - Governance, Carbon Markets and Regulation.
This looks quite favourable actually.
Open data is not a fully developed theme yet, but if we think about where we currently are:
- Restrictions on data use in Europe (GPDR - General Data Protection Regulation) and in some parts of the US (e.g. California Consumer Privacy Act) are already heralding a change in the landscape.
- Open Banking regulation is already mandated across Europe, UK, Australia and is in early phases of adoption across the US, Brazil and India. Open Banking refers to an open protocol for permissioned sharing of bank account data across financial institutions to encourage more competition in the financial landscape.
- There are predictions already that the emphasis will move to 'Open Finance' i.e. sharing other financial information and not just account data and even a full move towards 'Open Data' across all platforms - utilities, telecoms, and possibly Agricultural Data?
Open Agricultural Data - Outlandish? Unlikely? 😀
In July, President Biden in Executive Order 15069 encouraged the chair of the FTC to examine "unfair data collection and surveillance practices that may damage competition" and even referring specifically to agriculture:
"unfair anticompetitive restrictions on third-party repair or self-repair of items, such as the restrictions imposed by powerful manufacturers that prevent farmers from repairing their own equipment"
Similarly in Europe, one major funding programme, Horizon 2021 has also offered generous funding of up to €10m for projects "mapping and improving the data economy for food systems" for example. It will be interesting to see what it backs and what happens afterwards commercially.
One of my key takeaways is the issue of trust and governance.
The line from Farmobile above keeps echoing in my head 👉 "lines become fuzzy in situations where the trusted advisor also serves as a retailer selling inputs".
This is where the actors within the data exchange ecosystem become extremely important.
For example, how might dispute resolution work on a platform, if the dispute involved the platform owner or a major shareholder? This can obviously be managed but generally raises eyes brows if not full suspicions.
I think the market needs to be independent and obviously buyers vetted. The market rules and regulations will need to mirror those of other markets such as regulated financial markets.
3️⃣ Link to carbon markets
Are data markets too closely tied to carbon markets in Agriculture?
Both companies above agreed that current demand from sustainability projects was high and predicted demand for data for sustainability purposes would continue.
Scaling of carbon markets will rely not only data from formal soil carbon projects, but also data from those not participating in carbon projects. Developers and marketplaces have an incentive to gather this data and maybe they are natural buyers of this data.
The value of this data may create a price floor on all data assets while carbon markets scale. Over time, one would expect this to decrease as carbon models start to reach maturity.
But what about the other touted use cases of Manufacturer's, Financial Services, Insurance or Retail? Will we see these use cases materialise and scale?
This is where use cases on Ocean Protocol could be interesting to watch, to garner which ones take off and which don't. 🚀🧨
The key question still remains,
Will this usher in an era of 'Dollars for Data' for producers of that data?
I think it is clear, somebody will monetise it and hopefully this accrues across the value chain.
For me, this is where the importance of Governance really comes in - the mechanism with the best governance and trust will win.
What do you think?