First - I am sad to hear of the events of this week, and at the same time I feel like I had excellent timing in my own departure. Second - I appreciate that you make something as serious as the Boeing situation with doors blowing off planes hilariously funny to read about. Third - only you could figure out a way to weave major current event headlines (Boeing's shitshow, Kahneman's passing and What Went down on Wednesday) into a thought provoking and truly engaging read for this week. Whatever you find yourself doing in the coming weeks (spending quality time with family, pursuing other interests and exploring options), I hope this substack remains a constant. I feel incredibly privileged to have worked with you, albeit only briefly and in an at best tangential manner, and I will continue to be an avid fan of Benn.substack, as I have been a loyal subscriber for well over a year and get truly excited every Friday to read this.
Hey Kerry, thanks, that really means a lot. It was great to work with you (brief as it was), and I appreciate you putting up with just as many strained analogies in sales trainings as I impose on people here.
And I'm sure everything will be fine in the end. Such is the circle of life in Silicon Valley. Much of this is just part of the job; you've got to take the weird and unexpected with the good and exciting.
And yeah, the plan is to keep this going. I'm sure it'll evolve in some way or another, though I have no idea how. Given how poorly planned every post was before, I can't imagine I'll get better at doing that now. So whatever it turns into every week, you'll find out a couple days after I do.
There are echos of this same principle in one of Matt Levine’s posts this week about meme stocks.
“With time, I have become more comfortable with the answer to "what are we all doing here?" The answer is "not fundamental analysis." Maybe it is
"having fun online." Maybe it is "playing a complex game of mass psychology." Maybe it is "using our investments as a form of self-expression, buying stocks and cryptocurrencies we identify with and feeling better about ourselves if they go up." The third era is new, and we do not understand the mechanisms here as well as we understand discounted cash flow analysis, but maybe there are mechanisms to discover; maybe in 10 years there will be textbooks on Meme Stock Analysis.”
Oh interesting. Yeah, on a sort of meta level, I think it's easy to develop this sense that there are physical laws to how things work, when those things are actually more just convention. If we all believe a thing (how stocks are priced, what data teams are for, that our business can't sell to Gen Z, whatever) those things can kind of being self-reinforcing. And at some point, it's easy to assume that that's how the world works, and if you deviate from it, at some point, someday, you won't be able to defy the laws of gravity anymore. But a lot of those things aren't gravity; they're just how it used to work. But like Matt Levine says in that post, even if you can understand that academically, man is it unsettling.
My investment thesis(es) has always been focused on "hard" assets (Gold, Copper, Oil, Fertilizers)--the things that make up (or help us meet) the lower level of Maslow's Hierarchy of Needs.
The past couple of years I have embraced the fact that, much of life (for the last 40, 50, and particular 20 or so years) have been focused more on the higher end of Maslow's Hierarchy as technology, machinery and innovations have allowed us to meet these needs with plenty "leftover"--for a certain segment of certain countries.
I think that the outgrowth of meme stocks / cryptocurrency, "speculative assets" is the inevitable result of this process, and something I see continuing to the future.
I love this vision for data analytics. The narrators of the organizational hero's journey. Curators of formal and informal knowledge. Naturalists of bias, heuristic, superstition and dogma. Mappers and mutators of mental, causal models. Tenders of the metric trees.
Drive habits not decisions. Through reflection, discussion, intervention, mediation.
Stop being number crunchers, fact finders and figure suppliers. Sort the narratives, form perspectives, supply stories.
There's a need for this. Always has been. The analytics team is in prime position to take the lead on this.
Here's to hoping it's not just another, more elaborate, scheme.
Thanks, I appreciate that. TBD if my family or any of the people around me will appreciate it though...
And agreed about data teams. I think it takes a shift in character, where the job is more political (and, like, almost...emotional?) than it is scientific and feeling smart. Basically, I think we might be better off acting like user researchers, to be honest.
I made it through thinking fast and slow in… about 6 months. So slow.
Love the water cooler idea. I have never been intentional about collection assumptions like that, but I think it’s a good idea. I should probably even do it with my own ideas from time to time. 😀
I absolutely love the "data analytics is most definitely, 100% NOT a ponzi scheme" walk back when considering returning back to work after taking time off for family stuff.
In regards to the meat of your article, what it sounds like to me is that an analysts best skills could be implemented by: 1. Find and determine the assumptions, pathways, and decision making trees which are currently in operation at a specific organization; 2. Challenge these assumptions / pathways / decision making trees through verifiable data--as stated in the article; 3. Rewrite these assumptions / pathways / decision making trees with new, verifiable data and then implement this new structure through some sort of process, preferably software / code based in order to replicate and codify these implemented and verified decision making 'Algorithms'.
The biggest challenge I can see to that is finding a way to overwrite the CEO's / Vice Presidents / Middle Management's "instincts" for a particular field that they may have been for 30-40 or more years, but I think the best solution would for "Step 1"/Automated decision making to have been directly tested and implemented for an organization.
If I gotta shill for money, I will shill for money.
And I think I'd halfway agree with that prescription? I agree with the broad outline of 1) find assumptions, 2) check assumptions, 3) update assumptions, but don't think it's quite that rigorous of an exercise, nor is it something you can really codify. To me, it's not about figuring out what are the quantifiable inputs to models or whatever; a lot of these assumptions are looser understandings of how the world works - like, "the best thing for us to do today is stuff that doesn't scale." There's no way to validate that scientifically anyway.
To me, it's more about giving people more theories or frameworks about how the world works. Which doesn't mean taking on the CEO's instincts directly, because, to your point, you probably won't change their minds anyway. If they think the right thing to do is stuff that doesn't scale, that's what they're gonna think. Instead, I think you can give people more clarity about how the thing that you're doing that doesn't scale is going. Eg., "there is basically no difference between AI written and carefully researched BDR emails; I have a theory that all that really matters for our prospects is seeing the email, because nobody reads it anyway." That helps adjust the "do things that don't scale" theory without ever saying you're wrong.
Hi Benn. I guess what I had in mind, regarding the "codified" organizations, is something similar to Decentralized Autonomous Organizations (DAO) which were the popular thing in cryptocurrency 5 or 6 years ago. I'm sure you are familiar with the idea, but essentially an organization dominated by "code", I believe, would be brutally effective.
Ah, yeah. I may well be wrong, but I'm pretty skeptical of that sort of idea. I think there can be organizational processes that are run by code, where humans mostly introduce variance and error (eg, high frequency trading), but those things seem pretty rare. For the most part, I'd argue that companies are way to organic for us to run with any sort of deterministic code. (Could AI do it? Eh, maybe? But that's sort of a cheat to me, because that's not really modeling a business, nor is it using a "rule" to make decisions.)
First - I am sad to hear of the events of this week, and at the same time I feel like I had excellent timing in my own departure. Second - I appreciate that you make something as serious as the Boeing situation with doors blowing off planes hilariously funny to read about. Third - only you could figure out a way to weave major current event headlines (Boeing's shitshow, Kahneman's passing and What Went down on Wednesday) into a thought provoking and truly engaging read for this week. Whatever you find yourself doing in the coming weeks (spending quality time with family, pursuing other interests and exploring options), I hope this substack remains a constant. I feel incredibly privileged to have worked with you, albeit only briefly and in an at best tangential manner, and I will continue to be an avid fan of Benn.substack, as I have been a loyal subscriber for well over a year and get truly excited every Friday to read this.
Hey Kerry, thanks, that really means a lot. It was great to work with you (brief as it was), and I appreciate you putting up with just as many strained analogies in sales trainings as I impose on people here.
And I'm sure everything will be fine in the end. Such is the circle of life in Silicon Valley. Much of this is just part of the job; you've got to take the weird and unexpected with the good and exciting.
And yeah, the plan is to keep this going. I'm sure it'll evolve in some way or another, though I have no idea how. Given how poorly planned every post was before, I can't imagine I'll get better at doing that now. So whatever it turns into every week, you'll find out a couple days after I do.
There are echos of this same principle in one of Matt Levine’s posts this week about meme stocks.
“With time, I have become more comfortable with the answer to "what are we all doing here?" The answer is "not fundamental analysis." Maybe it is
"having fun online." Maybe it is "playing a complex game of mass psychology." Maybe it is "using our investments as a form of self-expression, buying stocks and cryptocurrencies we identify with and feeling better about ourselves if they go up." The third era is new, and we do not understand the mechanisms here as well as we understand discounted cash flow analysis, but maybe there are mechanisms to discover; maybe in 10 years there will be textbooks on Meme Stock Analysis.”
Oh interesting. Yeah, on a sort of meta level, I think it's easy to develop this sense that there are physical laws to how things work, when those things are actually more just convention. If we all believe a thing (how stocks are priced, what data teams are for, that our business can't sell to Gen Z, whatever) those things can kind of being self-reinforcing. And at some point, it's easy to assume that that's how the world works, and if you deviate from it, at some point, someday, you won't be able to defy the laws of gravity anymore. But a lot of those things aren't gravity; they're just how it used to work. But like Matt Levine says in that post, even if you can understand that academically, man is it unsettling.
My investment thesis(es) has always been focused on "hard" assets (Gold, Copper, Oil, Fertilizers)--the things that make up (or help us meet) the lower level of Maslow's Hierarchy of Needs.
The past couple of years I have embraced the fact that, much of life (for the last 40, 50, and particular 20 or so years) have been focused more on the higher end of Maslow's Hierarchy as technology, machinery and innovations have allowed us to meet these needs with plenty "leftover"--for a certain segment of certain countries.
I think that the outgrowth of meme stocks / cryptocurrency, "speculative assets" is the inevitable result of this process, and something I see continuing to the future.
I love this vision for data analytics. The narrators of the organizational hero's journey. Curators of formal and informal knowledge. Naturalists of bias, heuristic, superstition and dogma. Mappers and mutators of mental, causal models. Tenders of the metric trees.
Drive habits not decisions. Through reflection, discussion, intervention, mediation.
Stop being number crunchers, fact finders and figure suppliers. Sort the narratives, form perspectives, supply stories.
There's a need for this. Always has been. The analytics team is in prime position to take the lead on this.
Here's to hoping it's not just another, more elaborate, scheme.
And I hope you deeply enjoy your time with family and freedom. Mightily appreciate your writing and can't wait to see what else you choose to create.
Thanks, I appreciate that. TBD if my family or any of the people around me will appreciate it though...
And agreed about data teams. I think it takes a shift in character, where the job is more political (and, like, almost...emotional?) than it is scientific and feeling smart. Basically, I think we might be better off acting like user researchers, to be honest.
Enjoy your well deserved time off & hope to see you soon back in action!
Thanks!
I made it through thinking fast and slow in… about 6 months. So slow.
Love the water cooler idea. I have never been intentional about collection assumptions like that, but I think it’s a good idea. I should probably even do it with my own ideas from time to time. 😀
You're moving faster than me. I started it in 2015 or something and am stuck on chapter 2.
And yeah, that's good post. Probably in my top 5 most recommended.
Hi Benn,
I absolutely love the "data analytics is most definitely, 100% NOT a ponzi scheme" walk back when considering returning back to work after taking time off for family stuff.
In regards to the meat of your article, what it sounds like to me is that an analysts best skills could be implemented by: 1. Find and determine the assumptions, pathways, and decision making trees which are currently in operation at a specific organization; 2. Challenge these assumptions / pathways / decision making trees through verifiable data--as stated in the article; 3. Rewrite these assumptions / pathways / decision making trees with new, verifiable data and then implement this new structure through some sort of process, preferably software / code based in order to replicate and codify these implemented and verified decision making 'Algorithms'.
The biggest challenge I can see to that is finding a way to overwrite the CEO's / Vice Presidents / Middle Management's "instincts" for a particular field that they may have been for 30-40 or more years, but I think the best solution would for "Step 1"/Automated decision making to have been directly tested and implemented for an organization.
If I gotta shill for money, I will shill for money.
And I think I'd halfway agree with that prescription? I agree with the broad outline of 1) find assumptions, 2) check assumptions, 3) update assumptions, but don't think it's quite that rigorous of an exercise, nor is it something you can really codify. To me, it's not about figuring out what are the quantifiable inputs to models or whatever; a lot of these assumptions are looser understandings of how the world works - like, "the best thing for us to do today is stuff that doesn't scale." There's no way to validate that scientifically anyway.
To me, it's more about giving people more theories or frameworks about how the world works. Which doesn't mean taking on the CEO's instincts directly, because, to your point, you probably won't change their minds anyway. If they think the right thing to do is stuff that doesn't scale, that's what they're gonna think. Instead, I think you can give people more clarity about how the thing that you're doing that doesn't scale is going. Eg., "there is basically no difference between AI written and carefully researched BDR emails; I have a theory that all that really matters for our prospects is seeing the email, because nobody reads it anyway." That helps adjust the "do things that don't scale" theory without ever saying you're wrong.
Hi Benn. I guess what I had in mind, regarding the "codified" organizations, is something similar to Decentralized Autonomous Organizations (DAO) which were the popular thing in cryptocurrency 5 or 6 years ago. I'm sure you are familiar with the idea, but essentially an organization dominated by "code", I believe, would be brutally effective.
Ah, yeah. I may well be wrong, but I'm pretty skeptical of that sort of idea. I think there can be organizational processes that are run by code, where humans mostly introduce variance and error (eg, high frequency trading), but those things seem pretty rare. For the most part, I'd argue that companies are way to organic for us to run with any sort of deterministic code. (Could AI do it? Eh, maybe? But that's sort of a cheat to me, because that's not really modeling a business, nor is it using a "rule" to make decisions.)