Personal OKRs for Success
mohitkhare.me
mohitkhare.me
Analogy taken from Alan Watts - https://youtu.be/rBpaUICxEhk
I'd also posit that quantitative approach to business is also harmful. This path is taken by companies under stress of scaling - trying to find out a self organizing simple structure. The end result always turns out to be a siloed mess run by micromanagers and people "cheating" the system to satisfy some set of arbitrary metrics and the organization turning into a social laboratory full of instances of Goodhart's law and Cobra Effects.
https://en.m.wikipedia.org/wiki/Goodhart%27s_law https://en.m.wikipedia.org/wiki/Cobra_effect
I have yet to seen a successful implementation of it. Usually a Lean and System Thinking common sense look at the organization is sufficient to run a successful team or a company, but only so few are interested when quantitative (micromanaging) approach is so popular and sexy.
Lean and Functional Programming: https://youtu.be/5s55LA2Renc
If I was training for a marathon I would set numerical goals for how much I would run a week coming into it. If I am recovering from an injury that's even more the case.
Lately my son and I have been learning a martial arts form that has about 80 steps in it. There is a lot new about it for us: I've never done a two-person form or a weapons form before, and he's never trained in a form at all. You learn this kind of thing by breaking it up into pieces: so we have a burndown chart on the wall for it, which I think is a good motivation.
Most professional and highly adapted athletes are extremely quantitative about their training programs, much more so at the advanced stages than in the early stages. They just progress much slower because it takes more training for their bodies to continue adapting.
Athletics is incredibly quantitative- even for the most advanced and experienced athletes.
I try to go for macro metrics that at least let people understand what’s important, bur also not directly rewarding specific values or targets. An example is that I’m trying to promote data hygiene in my organization so we have some guidelines for how to clean data, where to store it, etc. We a general metric for “Who is reading our guidance” to at least see what pages are read. We have another for “who is versioning their data.”
I don’t have goals tied to specific numbers of readers or data repos, but do use it to at least make sure the material is available.
I’m still testing to see, but the problem with no metrics, I think, is that sometimes people will get astray without an outside source to help test and prioritize.
It makes life as boring as work.
These methods are meant for teams working towards a common objective, where anything extraneous to the OKRs is considered a distraction. This has its shortcomings (I've met way too many project managers who seem to be so absorbed by metrics they fail to see what is right in front of their noses). It's most definitely useful is this context, but not in life.
Consider: the author has set himself to read 20 books per year. Suppose this is double of his standard. What he will do to fulfill his OKR? He's not going to choose a difficult read such as Ulysses, or an existential crisis inducing book such as Crime and Punishment, he will go for what is more superficial and easier to read, because his goal is not to be transformed by what he reads, his goal is to read 20 books.
The crucial layer of meaning is lost when we project-manage ourselves. Life is more akin to a poem than it is to a project.
In paper you have accomplished your OKR, but in practice you are further away from your true goal.
The problem that most of us seem to have is that we're only tracking easy things to measure. Everything valuable is difficult to measure.
Reading a book is worthless. The purpose of reading a book is to gain something (learn something, read faster, practice public speaking, etc.). That's what we should be tracking.
Also, this doesn't really apply to emotional relationships for me. Personal OKRs are basically for self improvement which helps me in reaching something I aim for future.
I explored this throughly in 2014:
https://medium.com/@duopixel/the-qualified-self-7f69c6b23623
I have not updated this post, but I did this for years without finding satisfaction, because what I had to address was judgement about myself, not the shortcomings that I perceived I had at the time.
But the exercise of trying to measure the subjective is very instructive in itself. I ignore if you counter my point from experience, if not, I encourage you to register subjective judgement. And if you already do it, I encourage you to drop it and sense how it feels.
Many paths of personal growth require you to do opposite things at different times, so you can maximize your surface of experience.
The problem with this I see is the KR itself, not Objective. In 2019 I choose similar Objective which pushed me towards finishing books I didn't enjoy or valuable, so I decided to set the KR as number of pages read or just 30 minutes a day.
If you don't set a goal, you might just procastinate over it.
> Consider: the author has set himself to read 20 books per year. Suppose this is double of his standard. What he will do to fulfill his OKR? He's not going to choose a difficult read such as Ulysses, or an existential crisis inducing book such as Crime and Punishment, he will go for what is more superficial and easier to read, because his goal is not to be transformed by what he reads, his goal is to read 20 books.
- It is okay to read 10 books or X number of pages. Now, you actually started reading right! One's aim should not be to complete the KR for the sake of completion, but rather pushing yourself to do more, slowly improvements will be visible.
I do keep a list of books I’ve read because I like seeing it grow, but there’s not really a goal besides “read books” and I’m not giving myself an annual performance review.
That would be an obvious misstep. Both in terms of objective setting and in measurable key results. This blog post misses part of the value of OKRs: they teach us to set high level objectives and repeatedly decompose them. Here, objectives are instead just categories that provide no focus. It's like that antipattern where managers ask their directs for projects then try to combine them into OKRs. What, pray-tell, does launching an open source project have to do with boosting knowledge? And like, why 20 books instead of 10? And why specify quantity rather than quality or subject? Does reading a book mean you've boosted knowledge if you barely remember anything you read?
A better OKR setup would likely be something like (a bit modified to suit my own circumstances):
Objective: Boost Knowledge
Key result: learn more about statistics
etcWhich is decomposed into:
Objective: Learn more about statistics
Key Result: Work through the Freedman, Pisani and Purves Statistics textbook
Key Result: Create a hundred Anki cards on statistics
Key Result: Make the front page of HN with a stats blog post
And even those can be subdivided Objective: Make the front page of HN with a stats blog post
Key Result: Generate ten blog post ideas
Key Result: Research and analyze five blog post ideas
Key Result: Publish and promote all five blog post ideas
But you're absolutely right that meaning can be lost. Moving passion projects into a todo list misses the 'passion' part entirely, and is best reserved for the drudge parts of life necessary to enable the sublime. By all means, set OKRs for retirement, taxes, and career development.Try to measure the outcomes rather than the activities :)
> My wife and I are competing our annual reviews of each other. One of the challenges I’m facing this year is we didn’t agree on OKRs and I have a lot of qualitative feedback but don’t like how the KPIs look. I’m worried her annual review of me will be similar. We both aligned that we’d max out 401ks, move for promotions and end a car lease. But her career moved faster and we don’t have joint accounts so I failed to enter a savings goal. I’m going to suggest she didn’t save enough to see if I can get insight. Also the household itself did well this year we deceased our order in rate by 15% while moving to a good mix of organics and non frozen items +20%.
https://www.teamblind.com/post/My-wife-and-I-write-each-othe...
Not to mention all the "guides" seem to be dedicated to companies that are B2C or already heavily use the word "engagement" or depend on ad clicks or only make sense for the sales/marketing team.
1. Contrary to the article, I would say that OKR's need not be binary (complete or incomplete). We use a 0-1 scale but we consider >=0.7 to be considered "done enough". The reason we do this is to encourage our teams to strive for ambitious yet realistic goals.
2. Essential to the idea of OKRs is the distinction between outcomes and outputs. An output is the actual work you're going to do, whereas the outcome is the resulting change/effect you want to have. A KR should track the outcome, not the output. When devs first start out using OKRs, they write in KRs like "decrease lines of code by 25%". This is actually an output. An outcome would be "customer spends 50% less time waiting for page load". Reducing the lines of code would contribute to that outcome, but so would many other things. Those are outputs.
Here's a great book on the subject:
https://www.amazon.com/Outcomes-Over-Output-customer-behavio...
Sure, it’s good to get things done. But its also important to remember that we are human beings. Quite literally animals living on a rock that is flying through space.
Not every minute of our life needs to be tracked and judged and stressed over. It’s perfectly heathy to just do nothing sometimes. It is ok to have an unproductive day, or week, or period in your life.
On the other hand, some people have too little structure, and are miserable because of it. I realized this when I had kids: structure didn’t make them miserable; in fact, it gave them a sort of calm, and reassurance, and clear expectations.
There’s a balance. The “key” part of OKR is important — find the most impactful, highest leverage thing to measure, and focus on that. If you’re a student, that could mean getting a good grade in a course that’s very important to you. Too many goals can be as stressful as too few.
When I was younger I was perhaps too far on the unstructured side. There were creative benefits of that, but the detriment was ratholing on things that ultimately weren’t very impactful.
Ultimately, you have to keep yourself in check. It’s important to have open ended, exploratory time with no clear purpose. It can also be helpful to have some clear, key results as an anchor.
I think this use of personal OKRs might work for some people and might be a great tool for focus. This feels like a technique to manage individual projects, in that analogy.
However, I think it would be very easy to find yourself over-quantifying your life and find yourself feeling adrift even though you're succeeding on the numbers.
One thing that is very powerful about the regret minimization framework (https://www.youtube.com/watch?v=YlgkfOr_GLY - Bezos, famously) is that if forcing you to think about what you want to accomplish with your life and what makes you truly satisfied. If you know those, then quantifying your progress towards those goals makes sense. If you don't, then it's probably premature to have OKRs. OKRs measure progress towards a larger goal. Don't forget the lesson of The Hitchhiker's Guide, and make sure that you are working towards good goals first.
But to each their own, whatever makes you happy. I just know this would make me miserable. These key results look like a way to make your personal life just more training for your work life.
KRs should not be binary. They should be stretch goals. If you're hitting 100% on your key results you need to be setting your sights higher.
KRs should be things you can "influence" and not things you "do". It's the difference between "Get 100 subscribers to my newsletter" and "Write 100 blog posts". Or to use the example in the post, the difference between "Earn a X% return on my investments" vs "Read 3 books on investing". When you create key results around things you do, you run the risk of spending your time focusing on things that do not move the needle on your objectives. Last thing you want is to reach the end of the quarter and find your efforts have not had a tangible benefit. This is the classic "Outcomes" vs "Output".
But your actual life is a series of explorations - sure push yourself, but more important than anything is time to build relationships. especially with your kids if you have them.
I do this both for my own work, and when consulting for clients (separate lists). For clients the retrospective part is very easy to do, since I track my time there.
This has two major effects:
1. You sort of already know what to do when otherwise there'd be an urge to procractinate because you "don't know what to do".
2. You see your actual productivity, which, if you actually sit down and work, tends to be substantial, but doesn't _feel_ substantial. To me it always feels like I haven't done much, until I write down what I have actually done.
Then I got burned out and took a 3 month sabbatical in November last year. And just when I hit the desk again in February, the pandemic stuck. All those elaborate plans and spreadsheets - out of the window.
All this has made me realize that's it's futile to try and optimize existence. It actively kills creativity and joy. And things completely outside your control will toy with your elaborate plans anyway.
I saw them used at a struggling startup, where it seemed we had just one real goal (e.g. "find product-market fit") rather than a large number of objectives. We were already zig-zagging a lot, and we just didn't need the distraction of creating and maintaining OKRs, never mind yet another tool to create confusion about what was essential and what was not.
My first thought when I heard about it was the book "The Goal" where the protagonist's friend immediately says that a business in crisis has one goal: quantitative methods are important to find and remove the bottlenecks for that goal. (In that a startup hasn't established a profitable business at first, it shares quite a bit with a failing car factory or restaurant.)
Going back to Pareto, social scientists have established that you really can't find a total ordering for the world's utility function. A world where everything else is the same and I am $1 richer is a better world (partial ordering), but you can't compare that objectively to a world where you got $1 and I didn't.
This connects to a group of results (such as the Arrow Impossibility Theorem) that show that social science is limited in the way that mathematics is limited by Godel's theorem, the halting problem, etc. There is no voting system, economic system, etc. which provably produces "just" or "correct" results and in institution design we are stuck with limiting the damage that can be done by bad people.
A world where I hit 10 out of 10 OKRs is better than any in which I fail OKRs, in the real world that I hit 5 out of 10 it is hard to compare that to other 5 out of 10 outcomes that I could have had if I had made different choices.
There are some jobs that you can break down into pieces, say a supermarket cashier, the store has a model of what makes a good cashier, they have control charts at the front or in the break room for each of them on two metrics, one of which is the scan time (tied to the profitability of the business), another is the % of times you hit the "No Sale" button to open the cash drawer between customers, stop the scan clock, and cheat on the first metric.
That's profitable thinking because somebody thought it up a long time ago and the store manager knows it and the checkout manager who trained you knows it, and I thought it was fun when as a teen I could try but never beat certain middle-aged women who reminded me of my mother.
To expect every person to make a quantitative model of their work and have their manager review it, etc. is ridiculous. Many people don't think that way but do great at their jobs. People who have PhD's in quantitative thinking can also go into the weeds (e.g. either their first response is what I gave you above, they will pick one thing to maximize, present a Pareto Frontier of choices that is somebody elses problem, or just will go in circles until they get lucky, or learn the first answer the hard way.)
If they can pass as a neurotypical they'll get control of their systematizing urge and bullshit it the same way the sales person and art director do and get back to doing work done.