Using NLP to enable flatter organizations
middle-out.io
middle-out.io
Currently: I have a product idea, I sell it to my manager, my manager sells it to others managers, those managers sell it to their teams. My manager deals with politics in a well defined group of people they are familiar with.
Proposed Future: I have a product idea, software chooses 10 people who are stakeholders and will vote In it, I now need to sell my idea to those 10 people. I deal with politics in an undefined group of people that constantly changes.
If _nobody_ is politicking between arms of an org, everyone will just ask for the moon and over-sell whatever their team in particular is focussing on. Resources will be drained, and likely the many projects would diverge into an unsustainable mess.
Corporate politics isn't inherently bad, in that a ship needs steering.
* one of the major problems at corporations is lack of openness. People BS all the time. Worse, and I've seen several of these in my job, people believe that if you do not manage-up and BS then you're a chump. Part of the management is to cut this BS out. The difference between you and me at this point in time is that it's an unchecked assumption for you, but not for me.
* Cross functional coordination is essential to proper corporate governance. In particular, manager's damn well drill into internal service suppliers they'd better be customer driven and not make whatever they want - see also Conway's law. A lot BS at companies is half-assesd internal suppliers who underperform because they get management cover 'cause we have to use them ... and they cannot be essentially fired without extraordinary cause. Cross functional coordination is essential to getting and keeping people on the same sheet of music.
I'd read "Human Element" (Dr. Schutz) and "Total Quality Management the Japanese Way" (K. Ishikawa [might be sp there but title is good]).
In my experience manufacturing knows this is true, even they don't always practice it. But most service companies are clueless.
Yes, if you don't have a management approach (TQM the Japanese way), you don't have a solid framework to human organizations and teams (Human Element), you don't have cross functional coordination, you don't have customer driven teams, and management tolerates the BS ... then you don't in fact have management and you do have politics whence I refer you to my original statement.
Your point is well taken and I'll add those books to the read list.
I don't quite agree that "People BS all the time" really invalidates my point that you still need people directing a company. Flaws in the system, or people gaming it, doesn't mean the desired mechanics of the system aren't there.
I've always wondered what would happen if every company/corporation had a hard cap on its size and could not have more than 150 (a.k.a Dunbar's number) of people. Forget subsidiaries, conglomerates, etc. No individual could sit on the board of more than one incorporated entity, and any entity could not exert control over any other.
What would happen when a company started reaching this size? Would it start to focus on increasing productivity of the existing employees, or would they outsource absolutely everything that is not part of their core profit center and keep itself as lean as possible? How many layers of management would it have? What would happen with investment in R&D? Would this lead to companies being ultra-competitive for local markets or would competing companies form cartels and balkanize its customers in order to guarantee mutual survival?
Would tech companies still be able to follow the strategy of "grow first, find out how to make money later." My guess is they wouldn't, but would that be a bad thing?
I get why you picked that number, but (like most people) you've overlooked that most individuals have folks outside of their current place of work occupying a lot of those slots: family, friends, former co-workers, teachers and classmates, members of your community, professionals in various capacities (your doctor, your kids' teachers, etc.), And so on.
So, if you want everyone in your org to know each other without intermediaries, you have to pick a number
> Forget subsidiaries, conglomerates, etc. No individual could sit on the board of more than one incorporated entity, and any entity could not exert control over any other.
It is not unheard-of for people to occupy board seats in order to serve as someone else's proxy. By limiting the number of boards a person can sit on, all you've really done is raise the expense of using such a proxy.
I see your point though. Do you think that a company of 150 is still too big for each associate to care about the individuals? What would be a better number, then? Half of that? A third?
> all you've really done is raise the expense of using such a proxy
Part of it, yes. But I think it's not just that. By putting a cap on the size of the corporation, you also would be limiting the leverage that any one leader of a corporation would have over the others.
Anyway. If you expect everyone in an organization to have strong interpersonal relationships with each other, you'd probably top out at something like a dozen at most, and that's really pushing it. Eight or nine is a lot more reasonable, which, coincidentally, is about the size of a two-pizza team.
If you use the "two-pizza team" as an organizing principle (assuming a balanced tree), you can extend this to eight teams of eight, plus their eight managers, plus a ceo, with a total of 73 people, which is probably the point where relationship fatigue sets in and you can't claim to know everyone in the org even casually, unless you start dropping external relationships and the company is your social circle.
[0] If any. Just because two people both know you doesn't imply that they know each other. But if they do, you would probably know that and know what their relationship was. If everyone in the group had non-trivial connections to everyone else, the number of relationships in a group of 150 is over 11k.
I'll repeat. I am not saying that the idea would be to have a company where everyone had to have strong personal relationships. My idea was to think of a company that is no bigger than the "number of people that the human mind can keep track of in terms of their identity and relationship to you" (your words) as way to have a company where everyone can at least make a distinction between the individual and the corporation (in my words, "care" about the people).
You are right that we should not expect "strong personal relationships" with more than a handful of people. But using Dunbar's number was just a way to get an first approximate sense of dimension of a company size that is not so big to the point where individuals become faceless.
It's more or less like life in a megalopolis vs a small town. Living in a small town does not have a personal connection with everyone, but there is a sense of a "everybody knows each other" even if the town has more than 150 people. less The question would then be: how big can a village or a town be until it loses the "everyone knows each other" feeling?
People in a megalopolis don't necessarily have more or fewer connections per-se, but their connections are more widely distributed (often even close family members are globally scattered) and may not overlap at all with their neighbors'.
Just a title change? Noticed the subtitle and graphics were similar. IMHO, the “genghis” title was much better.