How to Keep Your Company Alive – Observe, Orient, Decide and Act | Hacker News Reader

How to Keep Your Company Alive – Observe, Orient, Decide and Act

steveblank.com

219 pointsweinzierl75 comments

The title quote "observe, orient, decide, act" is called the OODA loop, and it turns out the OODA loop is significantly more valuable than the article lets on.

https://en.wikipedia.org/wiki/OODA_loop

The OODA loop was developed by military strategist John Boyd, to explain how to direct one's energies to defeat an adversary and survive. Boyd emphasized that "the loop" is actually a set of interacting loops that are to be kept in continuous operation during combat. Boyd also indicated that the phase of the battle has an important bearing on the ideal allocation of one's energies. Boyd's views on the OODA loop are much deeper, richer, and more comprehensive than the common interpretation of the "rapid OODA loop" idea.

A great book on applying Boyd's ideas to business is Certain to Win by Chet Richards.

Steve Blank is well aware of that. It seems it was used in the title but not explicitly in the body. He's written about it in the past though:

https://steveblank.com/2015/03/31/hacking-for-defense-in-sil...

I highly recommend the biography "Boyd" about this man. I think I got the initial reco from this site.

Agree with this recommendation. Excellent book, great story. Very interesting career.

Boyd was an amazing man and his biography is a must read for anyone who wants to accomplish the impossible in life.

https://www.amazon.com/Boyd-Fighter-Pilot-Who-Changed/dp/B01...

The five items under orient are a strange choice.

They refer to one’s internal predispositions and biases as well as those of an organization attempting to respond to a situation. Depending on the situation, genetics could be in reference to height, G-tolerance, etc.

As far as I can tell the most common realization of the OODA loop is:

Observe (the concept of the OODA loop as if you are the only person to have ever heard of it)

Orient (yourself in front of your computer)

Decide (to log into any random discussion site)

Act (by explaining the OODA loop in a non-concrete pseudophilosophical way to random people on the internet)

Am I the only one who reflexively eye-rolls when people in the startup world co-opt military jargon?

I thought the OODA concept was pretty cool when I first read about it and John Boyd. And it's fun to say. OODA Loop, OODA Loop, OODA Loop. But it has been seriously overused, to the point where nowadays it's just a shibboleth that "thought leaders" use to show their "edginess" or whatever.

The problem is that "orient" has no real meaning in most decision-making contexts, and watching people stretch it to fit is almost painful. Sure, if you adopt exactly the right point in the word's semantic field, and carefully distinguish it from its neighbors (Observe and Decide), and squint a bit, it can make sense, but good metaphors or acronyms don't work like that. They don't require a High Priest to provide interpretation. A better metaphor, which I learned in a talk on autonomous computers at a UC Berkeley CS retreat in ~2001, is:

  * Measure

  * Analyze

  * Plan

  * Execute

I might not have remembered the terminology exactly, let alone the acronym, but I think it's close. Steps 1/3/4 are roughly equivalent to the same in OODA, but step 2 is quite different. More about accuracy/completeness than about rapidity, though of course both are always important. I suppose one could substitute different words to make a catchier acronym like MAKE or MACE. If increasing dilettante appeal seems important, feel free. For me, the important thing is that we're not all fighter pilots and it's OK to have a qualitatively different decision process. Come up with your own metaphor, your own acronym, that fits your situation intuitively without such arduous adaptation.

Identifying the system and in particular the closed feedback loops is a great learning from Boyd's OODA model; what is critical though is identifying and optimising out (well, minimising to the best of your ability) the entropy that inhibits iteration and degrades overall system performance.

OODA is certainly overused but "Orient" is hardly a mystery or metaphor – Boyd described it as the continuous re-evaluation of your "Cultural Traditions", "Genetic Heritage", "Previous Experience", "Analysis/Synthesis", and "New Information", and explained each of those in detail.

This hasn't explained what "Orient" means at all, you've just piled new concepts that need explaining on top of it and said it's a summation of them.

I'm completely new to this acronym but I'm grokking orient as "account for your bayesian priors."

That's about perfect

"Orient" is a fancy work for "think" the obvious step between Observe and Decide

In computing it's

Input

Compute

Return

Effect (output controls device drivers)

There's the pain I was talking about. Ow ow ow.

I found it harder to take OODA seriously after I came across "Observe, Overreact, Deny, Apologise".

HAHAHA! Good one. Thank you.

OODA is not executed sequentially like Measure, Analyze, Plan, and Execute and it can’t really be used without considering its interaction with an opposing process. It’s really something different. I do agree that misunderstandings of it are too popular.

That's a really good point about an opposing process. Of course, there are opposing processes in business too, but I guess not the same - multiple, not always directly or continuously opposed, vs. a single implacable enemy. Do you think OODA still applies to business, or does overcoming these differences change it into something else? Not trying to be snarky or refute what you said here. I'm genuinely curious.

Yes, the multiple adversaries and unending situation can make it difficult to apply to business. Business also has a lot of specialized operation methods that take up all the capacity for meta discussion in many organizations. In many cases this is good, e.g., most of the benefits of Boyd’s thinking as it pertains to individual sales performance are already found in good training material. In other cases, such as the cycle of SWOT analyses followed by initiative rollouts, Boyd is nearly absent. There have been some good efforts to reframe marketing as cost-efficiently understanding and anticipating market trends in order to reliably surprise the market which are closer to Boyd’s teaching of how to gain energy advantage than to the misunderstanding of OODA as outworking the problem. In short, it just depends on the business aspect and what kind of thinking is possible by the people working on it.

Edit: a simpler way to say all that might be that Boyd was too wonky for the military, so it’s even harder to apply to business. The principals can be repackaged and taught to people who don’t need to know OODA. The resentment of the people who write about OODA, in my opinion, partially comes from the knowledge that they would be unable to directly apply OODA in a real organization because they can’t teach the term without losing the full meaning. But it is possible to teach the meaning in business.

Your sequential cycle is more common known as the Shewhart cycle, Deming cycle, or PDSA:

- Plan

- Do

- Study

- Adjust

First plan an action through which you will learn something about reality, then perform that action. Study the results, and adjust your operation accordingly.

This is a sequential, iterative process, though, and sufficiently different from the original OODA concept for it to be worth not conflating them.

That doesn't quite seem the same. PDSA describes two phases - Plan/Do and Study/Adjust - resulting in two sets of changes to a system. They differ in the sense of big vs little, forward looking vs. backward looking, hence four total. The methodology I described results in one set of changes to the system, with a very different distinction between measurement vs. analysis vs. planning (which in PDSA subsumes the other two in the first phase). I'd say that PDSA is very different from OODA, but that's why it shouldn't even be part of this discussion.

You asserted it, but what makes your metaphor better? What results have you gotten out of it that everyone can learn from?

It's not really about my metaphor being better, though I apply this approach to problem-solving just about every day and it has played a part in a pretty successful career. It's not even about any particular one metaphor being better. The real point is that finding a metaphor/slogan/mnemonic that can be applied usefully to a particular situation is a perilous exercise. OODA is just an example of trying too hard to "borrow success" from a completely different kind of endeavor. It sounds smart at first, but doesn't hold up to serious scrutiny.

We've had 10 years of frothiness in venture markets. Many employees under 30 don't even know what a downturn looks like. They're used to high pay, constant calls from recruiters, and venture capital flowing into their companies. Even the companies who flop, no big deal, just go down the street to the next company. I worry that that generation is going to have the most difficulty with these changes. They'll wonder why things are contracting, why difficult decisions need to be made, can't we just wait this one out...?

I've never seen anything like COVID-19. I lived through the dot-com bubble, watching friends who want from an IPO party to crying and packing up their car back to Iowa within 6-months. Tech contracted and survived. Then there was the 2008 financial crisis. In that case, the damage was spread out beyond tech centers. Capital markets dried up and life seemed very uncertain for awhile.

With COVID-19, we went from extreme prosperity to an unprecedented shutdown of global economies with massive uncertainty as to how long this will last. Every business will be affected, the vast majority of them negatively.

I really wonder whether everyone is prepared for what it means to live in scarcity instead of abundance.

“We've had 10 years of frothiness in venture markets. Many employees under 30 don't even know what a downturn looks like. They're used to high pay, constant calls from recruiters, and venture capital flowing into their companies. Even the companies who flop, no big deal, just go down the street to the next company. I worry that that generation is going to have the most difficulty with these changes. They'll wonder why things are contracting, why difficult decisions need to be made, can't we just wait this one out...?“

I went through this in 2002. It took me a long time to accept that the good life in the years before wasn’t because I was so smart but because of a favorable market. It was quite a shock to actually having trouble finding something new and also to accept that my pay was significantly lower. Somebody told me at that time “now you see how job search always has been for most of us”.

We've been trying to hire since last year and have really struggled because rates devs have been looking for on the market are well above what we could see them returning value-wise for their experience level (that's just the nature of the market, I get it). When you have people getting £130k for doing React at a fintech company... well, I just hope people are prepared for the market changes that are underway.

Honestly, I'd rather this was all over and things went back to normal, but I do think that there's going to be a fairly hard reset of salaries for people that previously had a really sweet ride.

Consumer debt will make this harder.

> When you have people getting £130k for doing React at a fintech company...

How much is that as a fraction of the value of an average family home in your location?

If there's a big fall in salaries, there's also going to be a big fall in property values.

“If there's a big fall in salaries, there's also going to be a big fall in property values.“

I don’t think so. Rents will go up and housing ownership will be concentrated even more on wealthy people and big companies.

The destruction of AirBnB may push rents down. Hard to see how rent can go up beyond what people can actually afford?

Rent, health care, education and housing cost can go up to a level that it eats up most income for average people. That seems to be the trend.

I don't know about the national level, but within major cities where tech salaries and Airbnb are both prominent, I'm curious to see what happens to the local real estate markets. I know many people in Austin and SF who can afford their home because they A. have a large salary working in tech and B. offset their mortgage by Airbnbing.

£130k for a react job (or any software engineer job) in the UK? Where?

London

How many times have you seen this salary?

OP did say ‘rates devs’ - guessing they mean contractors who are typically £550-600 per day which works out to around £130k p/a (before NI, taxes, etc).

I'm not on £130k but I'm in London and also doing some development aa part of my job. Most of the knowledge I gained over the last few years will be rendered useless in a few months from now because there will be hordes of 10 times more experienced desperate for jobs. So be it. I'll pull out my plumbing tools, I haven't used for years, and start providing different services. If that won't work,will have to figure out something else. This will be a massive readjustment for everybody..

It's an artificial scarcity, though. We have the most productive technologies ever - we are just not very good to put a better part of them flow towards common good.

The amount of senseless applications and waste is staggering.

Good points. Plus the demand for tech is higher than ever due to the productivity advantage. There are pockets of that common-good advantage being demonstrated here and there, so maybe that can be patternized and better demonstrated to the whole. I'm sure there are immediately applicable ways to redirect a lot of the waste.

"...watching friends who want from an IPO party to crying and packing up their car back to Iowa within 6-months." LOL

"I really wonder whether everyone is prepared for what it means to live in scarcity instead of abundance."

Everyone? Not a chance. You and I who have previously experienced and survived the closest to it? Maybe.

For a lot of people who hang out here, this amounts to baseless FUD.

Technology is not going to contract any time soon. Quite the opposite. It's also really hard to build and distribute technology products well. Good software engineers and tech sector employees will continue to be in high demand for quite a while.

There might be an interim period of depressed income but that will also likely correlate with reductions in cost of living, as a consequence of the same macro factors. The key is that a good technology employee making wise career choices will continue to be able to live somewhere towards the front of the standard-of-living curve.

What might be lost for a while are some of the more extreme tail outcomes (in the many millions of $s) that had become a bit too easily realized in this latest iteration of frothy tech bubbles. Not a big deal. Most of us won't notice any difference.

You don’t need to worry about people under 30. Every older generation thinks every younger generation isn’t going to handle things well. And every single time, they are wrong.

There's no doubt action (and speed of adaptability) is important, but I'd also highlight a 2nd virtue - patience.

Action for the sake of action is called panic. People who like to control things (executives, leaders, me, etc) often feel like they need to be doing something to have a sense of control in the midst of chaos.

The only thing that needs action right now is triage - cost-cutting. Cash flow to survive and keep payroll going for the next 12-18 months. Once the safety net is in place, the mood changes from frantic survival to calm.

At that point, the ratio of thinking:acting time should arguably double. Recessions take plenty of time to unwind. Take 30 days (not 2) to gather information from across the org and evaluate trending opportunities. “Measure twice, cut once.” applies to pivots, too.

Re-developing new models for recognizing opportunities and a new decision framework for when to act upon those opportunities takes time. The folks who spend more time thinking about where the puck is going to be vs. reacting to its every jitter are the ones who will come out ahead.

This is such an important point. However, sticking to the "panic only as much as needed" idea is difficult because, sometimes, incentives are skewed to reward panic. Think about all the laws that are passed without regard to the side-effects because "someone ought to do something about this". In peacetime, prepare for war. Before the crisis hits, you should take the time to think through how you and the people around you need to behave and what is likely to trip you up. Don't assume you will do the right thing. Plan to do the right thing.

Or as Deming liked to say, "Don't just do something -- stand there!"

A similar title for this article could be: "How to bodysurf your way out of a tsunami"

Good luck, you're screwed.

If you're a cash flow negative startup even slightly tangentially related to an industry impacted severely by COVID, you're done. This economic shutdown is end of days for a huge number of small businesses.

The media is hyping shutdown until at least May, but I'm seeing more "June and on". If that comes to pass, there will be very few small businesses left standing in the heavily impacted industries. It's going to kill a lot of otherwise healthy businesses too.

The stimulus is mostly a feel good measure for politicians. How many of you have gotten your checks? How many have applied for EIDL loans and received them? How's the PPP going?

Oh I'm sure they'll iron out the kinks, but it'll be too little too late. Hope ya'll are up to date on your Steinbeck, because that's what's coming.

A similar model taught to pilots is DECIDE:

  - Detect that change has occurred
  - Estimate the need to counter or react to the change
  - Choose a desirable outcome
  - Identify actions to achieve the outcome
  - Do the necessary action
  - Evaluate the effects of the action

It suffers a bit from being a backronym, and I don't know anyone who explicitly applies it "in real life," but I do emphasize the last item with students because it's so often overlooked. A lot of people will see a need for a change, do something to make the change, and then fail to notice that the desired outcome wasn't achieved.

I see this omission in software systems a lot too, which manifests as a lack of monitoring, control, verification, and/or alerting. An ETL process that doesn't verify the correctness of its output, for example.

I imagine in many cases, what they chose to do in the past matters more than what they are doing now...in terms of survival.

Cash reserves, assets you can get loans on, percentage of workforce that are contractors or professional services, flexible contracts, reasonable real estate costs, for example.

Some changes in the business environment seem non-intuitive. My cloud migration consulting work (which I thankfully do remotely) has seen a quick uptick. A lot of customers want to migrate to cheaper clouds or reduce their cloud footprint. I was expecting not have much work as many of my startup clients see revenue pressure. But in the short term a few months the opposite may happen.

> “ Notice that the word speed appears twice. This is not the time for committees, study groups or widespread consensus building. Even with imperfect information, the future of your company depends on your ability to make rapid decisions and start acting.”

Yikes. I guess principled decisions have to go? Executive authority!

It reminds me of the president in The Simpsons Movie... “I was elected to lead, not to read.”

A few issues: 1. A decision isn't automatically more principled because it is made by a committee/consensus. Often the opposite is the case. 2. The purpose of the firm is to make money/value for shareholders, not principled decisions. If making principled decisions gets in the way of the goal, principled decisions are what should slip. 3. At least in the US, companies are generally not organized as democracies, so even if executives sometimes choose to delegate their authority to committees, that arrangement is always subject to revocation in extraordinary times.

The OODA loop was invented by Col. John Boyd during the Vietnam War and is being adapted here for the business environment.

The point is for some hard decisions, you will have to act on imperfect or incomplete information. A good leader will make a decision, because doing nothing and hoping you get more info will in fact be worse than making the wrong decision. As you get more information, you can then alter or tailor your decision. This is especially true in a military environment.

For an example in a popular movie, look at 1st Lieutenant Norman S. Dike Jr. in Band of Brothers.

Isn't the point of getting inside the opponents OODA loop to force them to make decisions faster than they are able to sensibly and ergo force wrong decisions. If you get inside your own loop by hurrying your decision making forcing wrong decisions how is that sensible?

It also seems obviously untrue that acting on a wrong decision is better than doing nothing. The results of a wrong decision could of course be worse than the results of not making a decision in haste.

If you're already at the point of being forced to make decisions then you've already screwed up? You're being reactive.

>Isn't the point of getting inside the opponents OODA loop to force them to make decisions faster than they are able to sensibly and ergo force wrong decisions. If you get inside your own loop by hurrying your decision making forcing wrong decisions how is that sensible?

Nope. The point is if I am inside their OODA loop, I can make decisions and react quicker than the adversary. Now they are reacting to my decisions and I have the advantage.

> It also seems obviously untrue that acting on a wrong decision is better than doing nothing. The results of a wrong decision could of course be worse than the results of not making a decision in haste.

The keyword there is "could be". That's what makes leadership hard. You have to know when is enough information to make a decision, and when you need to wait. Or if you made the wrong decision, what can you do now?

EDIT: For an example, there's a small brewery down the street that cannot can all of their beer. Making beer takes 3 weeks at least. So there's predictions of COVID-19 ending in May to Sept. So what is the right choice? Do I hope that it ends early and brew beer in a couple of weeks, or do I wait until Sept? What about all of the perishable supplies I have now?

It does me a lot more harm to wait until the "all clear" and then order supplies. Why? Everybody else will do the same thing, and now I am out of beer for three weeks (and longer for other beer). That is a lot of revenue left on the table lost.

What about making it now? Well if I am too hasty, now I have beer that will skunk because I cannot sell it fast enough.

So the brewery owner needs to see what info they have now (obverse, orient), and decide and act. What will cost me more money?

> Nope. The point is if I am inside their OODA loop, I can make decisions and react quicker than the adversary. Now they are reacting to my decisions and I have the advantage.

Yes you'll have the advantage because they'll be making bad decisions as the situation will have changed by the time they've decided to act on older information. It's not sufficient to simply act faster you have to be making good decisions as well.

> The keyword there is "could be". That's what makes leadership hard. You have to know when is enough information to make a decision, and when you need to wait. Or if you made the wrong decision, what can you do now?

Yes exactly so acting isn't axiomatically the thing to do. Although making a great folly seems like a great way to get lionized in military history. :D

> Yes exactly so acting isn't axiomatically the thing to do. Although making a great folly seems like a great way to get lionized in military history. :D

Usually when battles are studied in military history, they try to look at it based on what the leader knew then versus what we know now. Hindsight is always 20/20. There are examples of famous battles where even studied now, military strategists say that the losing commander did the right thing, based on the information available.

For sure you can make good decisions based on the info available and still lose. You can also make bad decisions and still win. Sometimes decisions that seem vital are totally pointless etc. etc. etc.

...what?

I think you completely missed the point of th OODA loop.

I understand the OODA loop thanks.

I was replying directly to your point that military historians found that some military commanders were making good decisions based on the info available even if they caused grand follys. I was pointing out that doesn’t really say much as most people try to make good decisions but won’t always be able to or won’t be making decisions that actually matter as external forces overwhelm no matter what decision is made.

The best laid plans of mice and men often go askew to paraphrase Burns.

There is a french saying I first heard back during the 2008 crisis: "It's urgent to wait".

I liked it, consicously waiting might be exaclty what one should do now. Don't rush, wait, and move once you have to.

Like the line in the movie "Shooter": "Slow is precise and precise is fast". Just rushing things now because of reasons can do more demage than good.

EDIT: Typos

Yeah the classic phrase I've heard is "slow is smooth, smooth is fast, speed is in the efficiency of motion" which is a reminder that by trying to go fast we often get sloppy and it can be faster to be smooth and precise than quick and sloppy.

> If you're already at the point of being forced to make decisions then you've already screwed up? You're being reactive.

In the case of a global pandemic never seen before in any of our lifetimes, everyone has to be reactive.

And we have already seen a lot of failures from failing to act during this crisis, more so than leaders taking actions that turned out to not be necessary.

We’ve definitely seen a lot of bad decision making and I’d agree a lot of that has been due to people being caught off foot and having to be reactive.

"Your revenue plans are no longer valid."

Probably my favorite line in the script :)

If they were in the first place, that is!

While OODA is an interesting and increasingly popular concept, it is really only useful when applied to very short intervals of time. You can and should expect that an opponent with "a reasonable amount" of time will make the correct response. In business that timeframe is much longer than the few minutes or seconds a modern aerial duel might be concluded in.

The general idea of OODA is that if you can "play fast" you can beat someone who can otherwise "play better". Think about how blitz chess differs from traditional multi-hour games, even though the rules of play are the same: for example it is often possible to exert time pressure by playing "confusing moves". This concept of "fast play" translates into the business timeframe more as the concept of "agility" (to be contrasted with large-company "inertia"), whereas businesses only try to confuse one another on television or by accident :-)

Concepts similar to OODA but more applicable to a business context and timeframe would include PDCA [1] and DMAIC [2].

PDCA a.k.a. the "Deming cycle" stands for "Plan, Do, Check, Adjust" and is closely related to the concept of "kaizen" [3] (or "continuous improvement"). PDCA happens in timeframes that can best be described as a "short project" (or a "kaizen event").

DMAIC ("Define, Measure, Analyse, Improve, Control") on the other hand is a framework commonly used for larger business interventions where the stakes are higher, more project members are needed, more structure is required and so on. It is a key feature of the Six Sigma [4] approach as well as Lean Six Sigma [5].

Note: I'm sure there are other valid approaches too. I'm just mentioning the ones I am familiar with in case anyone else is interested in methods to help businesses improve and respond to change.

[1] https://en.wikipedia.org/wiki/PDCA

[2] https://en.wikipedia.org/wiki/DMAIC

[3] https://en.wikipedia.org/wiki/Kaizen

[4] https://en.wikipedia.org/wiki/Six_Sigma

[5] https://en.wikipedia.org/wiki/Lean_Six_Sigma

Unemployed Engineer here. I can also sew. Not sure what I should be doing. Should I make non government certified masks?

>Not sure what I should be doing.

Set up a schedule to modernize your skills and start applying to remote first companies.

>Should I make non government certified masks?

Yes, in the unscheduled time.

Not a single reference to the actual creator of this model, John Boyd, a fighter pilot.

In Akamai the mantra is - Stabilize, Adapt, Emerge

...that and money

Reply on news.ycombinator.com