the real danger in assumptions
Prime Future 156: the newsletter for innovators in livestock, meat, and dairy
My reaction to the above analysis of the indoor vertical farming category was to wonder, what assumptions led to so much capital being invested?
Was the assumption that energy costs would be cheaper than they are or that the system would be conducive to higher-value crops? Or that farmland prices would increase by more than they have? Or something else entirely, like subsidies for locally produced food for urban areas?
There had to be assumptions that made these investments make sense.
I recently found a voice memo from 2 years ago when I was deciding whether or not to take my current role. 2023 Janette heard 2021 Janette describe what I was hoping to learn by taking this role and was able to quickly see where my assumptions were very right and where they were very wrong.
Clearly today we’re talking about assumptions, both the skill of articulating assumptions and the habit of documenting assumptions…and the biggest assuming danger most of us don’t think about enough.
Miss getting weekly Prime Future content? That’s easy to solve:
The 'curse of knowledge' is the idea that we tend to forget the time before we knew something, or when we thought differently. It's part of why it's often hard to remember things we were wrong about, because we forget what we thought when or what we knew when. We don’t remember clearly our assumptions before a decision, so its hard to evaluate our decision making process relative to outcomes.
And that can create a drag on learning velocity 💀👎🏼
Especially since most learning is gradual rather than an 'event' with a 'where were you when you learned xyz' type moment. Learning tends to happen incrementally and over time.
So there are the big decisions like whether to invest, start a company, or make a career move where assumptions play a starring role, but perhaps even more important are the assumptions we use in the frequent decisions that we make day in day out.
On that note, I used to work with a SaaS company selling to companies that made decisions that were tied to commodity markets. Think meat packers buying live cattle (when to buy, how much to pay, who to buy from, how to structure the buy, etc) and selling boxed beef (when to sell, at what price, how to structure the sell, etc).
The software included one component of forecasting commodity markets and another component that made recommendations on what to do with those commodity market forecasts. Predictive analytics yada yada.
But my favorite concept from the product was the idea that the software allowed a user to capture their assumptions about what would happen in a given market so that the decision-maker could later compare what actually happened with what they expected to happen, and hopefully learn in the process.
This documentation of assumptions was something new for most users who made assumptions every hour of every workday about what the market would do but lacked a way to document those assumptions. And if the assumptions are not documented, it is much harder to learn from them.
(See also: decision-makers with twenty years of experience vs decision makers with one year of experience repeated twenty times.)
Being able to make assumptions explicit is a multiplier-type skill, and documenting them is a multiplier-type habit.
It's true for investors creating a thesis about a space.
It's true for managers building 1, 3, 5+ year strategic plans.
It's true for founders building new products and new markets.
I wish documenting assumptions and sharing assumptions were a stronger part of the common vernacular. Why? Because even when we don't identify them, they are there.
Even when we can't articulate our assumptions, we have them.
Even when assumptions are hidden, they influence our decisions.
The more we can make assumptions visible, the more easily we can find ways to test them.
And the faster we can test assumptions, aka hypotheses, the faster we can prove them right or wrong.
The costliest assumptions are the ones that are invisible, that we don't realize we're operating from. Why?
Because it's usually after they are proven wrong that we realize we held them. 😑
We all know the saying about what happens when you assume…
But the danger is not in assuming, it's in not being clear what we're assuming.
Sometimes identifying our assumptions is difficult, like a fish trying to identify the water its swimming in. Here are some things that help me in the continual pursuit. of getting better at this skill:
Set aside time to identify assumptions. Intent is half the battle here.
Target 3 key assumptions for a given decision.
Use the question “what has to be true?” as a way to identify assumptions.
Borrow the 6 Sigma framework of the 5 Why’s as a way to identify assumptions.
To take this to the next level beyond merely documenting assumptions, borrow Annie Duke’s tactic from “Thinking in Bets” to actually call out the percentage probability that the assumption is true. This forces you to get more critical of the assumptions you’ve called out.
The thing about learning out loud is that I have shared a lot of assumptions that will undoubtedly be proven wrong in the coming years, but hey they’re documented so we’ll all know where I got it wrong 🙃
Here are a few of those core assumptions:
The gap will get bigger between mega large-scale efficient commodity production and mega high-value niche market production.
That bifurcation of value chains will create distinctly different gaps, needs, and market opportunities for technology and new business models (both in livestock production and tech solutions for livestock producers).
For commodity producers, climate-smart solutions will be adopted when they improve emissions profile AND improve efficiency.
Carbon markets will fade away over the next 10 years, particularly those focused on offsets.
Plant-based meat will be nothing more than a fraction of the meat case and will face increasing pricing pressure. Cell-based meat companies will produce valuable insights for other applications, but cell-based meat will not be commercially viable in my lifetime.
Generative AI will change the operational side of producing, processing, and merchandising meat, livestock, and dairy more than any other technology wave of the last 50 years.
I know a few folks who have worked or are working on cell based meat. Based on what I've heard, your observation that this will not replace meat meat any time soon is spot on. We must admit that fail, learn, adapt , e.g. building intelligence, in food takes a while and beware of "magic" that solves all problems. Food industry definitely benefits from folks like you bringing these important issues to discussion. Thank you!
Outstanding post as usual. This is near and dear to my heart because I invested in a vertical farm in southern California...that closed. Funny thing about those assumptions, i.e. water will be biggest expense; no bugs or pestilence because it's inside; the $3.00 head of lettuce can sell for $7.00 - All wrong. What we learned is that whatever happens within traditional agriculture, up to and including weather (humidity attracts all kinds of problems even in dry so. Cal), happens inside...except the sun is free; electricity costs more and more. Technology will NOT change this. We should all beware that humans have grown and processed food a certain way - think baking, brewing, raising/slaughtering cattle, etc. - for good reason - we can feed alot of people. Yes, we can be more efficient, But disruptive? We want to be really careful. We cannot survive without food and water.