Organizational Physics Applied: Field Notes — Practical Patterns from Inside the System
Northstar was one company's eighteen months. These are the patterns that undermined it, each usable in your company, each usable alone. This is the closing piece of Organizational Physics Applied.
The arc of Northstar’s story matters most in its ability to show the patterns underneath it, patterns that exist in every company to one extent or another. What follows here is a field guide: each pattern the Northstar arc demonstrated, at length, condensed into what it looks like from inside, why such things go unnoticed or ignored, and how to expose that pattern for what it is.
The patterns that both built and damaged Northstar, and the ones that saved it, are most certainly running in your organization right now, in some form, whether you’ve identified them or not.
Use the one pattern that describes something you’re already half-seeing or suspicious of within your own company and start there.
The Exception With No Edge
The in-the-moment call that, at that time, was the right thing to do, but actually reset the standard and allowed for a new baseline to creep in. It could be as simple as a client calling at 6pm with a valid reason to move the deadline, and the request was entirely reasonable in that isolated moment. But after that moment, nobody documents where that flexibility ends, or that it was an even an exception. Which then allows the next similar request to get a similar, though now slightly easier and expanded exception, because the standard is now able to shift one reasonable approval at a time. And it will continue shifting until the original reference point you’re measuring against has completely changed without any overt decision or intention.
This is the cheapest pattern to catch, and the one most often missed, because every approval that led to it is justifiable, in isolation, along the path. The degradation lives and breathes in the rule and documentation nobody wrote down: the clear line that would have told the next decision where the last exception ended, and that it was, in fact, an exception.
The exposing question to ask:
Identify the most recent exception that appears the most logical and justifiable. Can you answer, in one clear statement, the condition or rule that prevents the next incident from inheriting the same or similar exception? If you can't articulate that, then the exception is most likely already forming into a new standard that nobody approved.
The Average That Reaches the Top
Truth starts at the bottom of the, with those closest to reality, and something drastically different comes out at the top of the org. Every layer on the way up refined and smoothed virtually every bit of information toward resolved, toward managed, toward fine, because everyone doing the summarizing on the way up has learned exactly what their superior rewards or dislikes hearing.
No one in that chain lies, exactly. But each and every one of them curates and adjusts. And the person with the most authority, the one at the end of that chain of small and compounding tweaks, ends up with the least accurate data in the company. And this is a gap that widens the more senior they get, the longer their tenure, and the better the company looks like it’s doing.
Counterintuitively, asking for candor to counteract this actually worsens the curating instead of eliminating it. The information was always being adjusted because curating makes things easier and cheaper. A request for “more truth” runs against that, causing it to carry more risk and more cost because surfacing a problem after the request for candor now reads as admitting something wasn’t handled before. It is a self-preserving instinct.
The exposing question to ask:
Think back to the last time someone brought you a raw, unfiltered problem, with no plan or solution attached yet? Name the date, and be precise. If there is no such date, then it is one of two possibilities: people are actually genuinely solving everything without your involvement, which may be good or bad and bears you checking on, or…or the lack of escalation and the smoothed and refined narrative you have been seeing as a sign of your team’s health is actually a symptom of the curating, and you’ve been reading averages and half-truths for quite a while.
The Temporary Thing Everyone Now Depends On
The thing about workaround is that they often solve a real problem, they work, thus the name. What happens to things that work is…they get re-used, they get built upon. Without any discipline, nobody ever sits down and decides it’s permanent, and they typically don’t sit down and communicate the temporary nature of a workaround either.
A dozen completely reasonable reuse decisions later, which sped up other parts of the company or made people’s days easier, and now that workaround is carrying fundamental weight the organization cannot remove in any way other than a destructive way. The workaround got built into, one project and usage at a time, until built-into was the default, and there was nothing temporary about it.
Caught early on, this costs only a decision or a documnentation step: someone chooses to correct the workaround, or document that it is temporary and cannot be re-used, with a funded timeline to patch it. Caught later on and it has an owner, a budget, headcount, multiple dependencies and tendrils defending it, along with the intrinsic tribal knowledge of its use, and correcting it costs something impacting and real.
The exposing question to ask:
Document everything you still call temporary, whether it be code or processes or even teams or projects. For each one, enumerate what depends on it now. Any nonzero count means you’re carrying unacknowledged infrastructure, the most expensive kind there is, as this kind of “shadow” infrastructure requires the full weight of a permanent system but never got the investment one would have required. A workaround is most often the more acceptable term for a hack.
The Reorg That Pours Concrete on the Degradation
After a rough stretch, a capable leader often restructures by adapting to and building around what the company is now instead of the original goals and standards it drifted from. That coordination layer with a status checkpoint meant to fix slow escalation now becomes the permanent home that slow escalation lives at. That team spun up to keep producing the custom SOW work directly injects the degraded adaptation directly into the org chart. Each and every individual move is defensible at the time of its creation, in what it responds to and tries to solve, and every move after that further ratifies the degradation, because structure is simply behavior made permanent.
This pattern often goes entirely unnoticed, because it works, initially. Coordination improves, the numbers get better, the changes look rational and reasonable and even like the smart play at times. The system reads those improvements as proof the reorg was correct, but the standards beneath it all are still eroding.
The exposing question to ask:
For each box on your latest org chart, look at whether its intent is to produce what you actually want, or merely to keep doing what you drifted into, and does it actually do that? Adding a team to handle one-off custom SOW work just creates a permanent outlet to give more custom SOW work to.
The Standard-Holder Working Twice as Hard
Somewhere in your org, right now, there is a strong performer correcting something about the system that they can affect, and it is working, despite the surrounding system that has continued to degrade and is actively pushing against this localized correction from every edge it can find. That person is very likely working visibly harder than their peers just to hold a line the structure itself keeps attacking. The gap between their effort and their ordinary-looking results is often the defining signature of this pattern, and this it is often the most reliable early warning trigger that you’re about to lose the person you can least afford to lose.
Praise, as a response and a counterforce, is the devious failure hiding inside this pattern. Praising someone for fighting the system’s inclination is cheaper than fixing the system, and it leaves the whole weight of the system sitting on their shoulders until, at some point, leaving is easer than staying.
The exposing question to ask:
Look at your strongest people and find those who seem to be spending an unusual amount of energy to produce results that look, surprisingly, ordinary. It is very likely that person is actively fighting, or holding the line against, the system around them and all that energy is being spent in opposition and correction. And that person is very likely holding the line by themselves, and their exit interview will rarely mention that they were correcting your system for free, alone, for as long as they could stand it.
The Order of the Exits
If you or your company reward adaptation over the standard you profess long enough, the organization eventually stops just shaping behavior, it starts to select for it.That selection will move faster than any policy you write or speech you give. The departures stemming from that selection almost invariably run in a fixed order, with the order being the indictment: those with the most talent and thus the most options go first, followed by the people with the strongest conviction who had nowhere to use it, and then the uncredited and unseen ones who were silently enforcing the standards the formal system only said it cared about. What remains are people who have been filtered, stage by stage, down for adaptation and low friction, those with the least incentive or means to do anything about anything other than take orders.
Every other pattern named here is structural and yet reversible with enough will and force. This one walks out the door, and it will never give back what it took.
The exposing question to ask:
Line up last year's departures by talent and conviction, instead of by date. If your best and most committed people, those who started out pushing and making things better and over time got more frustrated or more silent before they left, cluster toward the top of the list, then you have direct evidence that selection has already been enforcing a lower standard into your company, one defeated exit at a time.
The Correction Cost That Only Goes Up
Every pattern described above is a face of the same underlying number: what it costs to fix the structural fact underneath it, and that cost only grows the longer the fix is unaddressed. A one-sentence correction that is available to you today turns into a multi-quarter rebuild in three years, both aimed at the identical root cause. No dashboard tracks this number or tells you about it, because dashboards surface symptoms, and the structural variable runs hidden deep beneath the metrics.
Early recognition is the entire game here, because this is the one pattern where the price only moves in a single direction.
The exposing question to ask:
Find the single structural fact your other symptoms are all facets of, and determine what fixing it costs today, whether it be budget or headcount or meetings or restructuring. Whatever that price is, it is the cheapest it will ever be, right now. Each and every day without your correction increases the cost until at some point the correction will can only be destructive.
What This Field Guide Is For
None of this is a checklist to complete, and it is not even all-inclusive, these are simply the most evident and most impacting patterns to begin to see. Once you start looking, and know how to look, you will find more. That’s why this is more about a way of looking at your own system, rather than just the literal patterns themselves.
The coaching you’ve paid for, the psychological safety work you’ve done, the culture investment and empathetic leadership you’ve embarked on…none of that was a waste, none of it was wrong.
Each one of those attempts addressed something real. But what none of them could do was actually stick long enough to have an impact because the variable underneath them all was working in direct opposition. It’s the same variable every pattern above names in a different, but at a deeper layer. The incentives that govern behavior, adaptation, and unity. Correct that variable, those incentives, and everything else you’ve already built can finally start having a lasting impact and be self-reinforcing instead of collapsing.
Start with whichever pattern names something you are already half-seeing or suspicious of your own company, the thing that stings, or doesn’t quite make sense. You have the intuition of something right now, but perhaps you simply haven’t had the language or observation shift to see it clearly. Hopefully, now, you do.
Use it.