3 ms·

I've used OKR for 7 years, on a quarterly basis. I had simple markdown files at first, then vnl-log files, and now R notebooks (https://github.com/dkogan/vnlog

I've used OKR for 7 years, on a quarterly basis. I had simple markdown files at first, then vnl-log files, and now R notebooks (https://github.com/dkogan/vnlog https://github.com/dkogan/vnlog) to read / plot. It may seem like overhead, and there's some snark in this thread about how it's project / team management without the project and team. I completely disagree. If you set up your KR's so they are 1) quantitative, 2) daily measurable, 3) simple to log ( a few keystrokes while journalling) and 4) completely under your control to achieve. At the end of the day, I mark down my progress on all my OKRs. I can quickly plot them, look back at progress, and look back at goals and concerns by seeing the types of objectives I had. It's a 10,000 foot journal that I otherwise wouldn't have. There's more to this than simply quantifying yourself. We like #'s because they are representations of complex systems. The self and your personal history are absolutely a complex system worth tracking. Looking back at my OKRs when I was dating my (now) wife, comparing the ways I put effort into our relationship and our changing priorities. Seeing over time my running distances, weight lifting activity, meditation record, and seeing how I consistently attempt to over-achieve by setting KR values too high ... Having those points of reference has made today more enjoyable, and been a constant reminder that progress comes slowly and missing on any particular attempt at something is irrelevant. It's so completely a part of my life now that I can't imagine setting goals or daily priorities without it. Think of it like quantitative journalling.

sh_123 6y ago
Can you share more about your process? I'm very interested in learning more. How do you track your daily data points?
- tuatoru 6y ago
  There are a lot of these comments on HN, that baldly ask for more information without contributing anything or providing any context. So many that I suspect a bot.
- jvanderbot 6y ago
  I have a text file in a known location. It's backed up on git. I can open and edit that file quickly using bash. From within VIM I can quickly commit to reduce the # keystrokes. More details: https://github.com/jodavaho/bashlog#git-integration https://github.com/jodavaho/bashlog#git-integration (Really though, you can just save as an environment variable your file location) The text file is structured with space-delimited text, like this: https://github.com/dkogan/vnlog https://github.com/dkogan/vnlog My current structure for tracking a contribution to an OKR item (e.g. "meditate") by date (e.g., 2020-01-15) and amount (e.g., "1") is simply: "# date item amount" Easy to open, easy to plop down today's date and the item / amount. Save multiple date/item entries and add them inscripts ... Easy to save. Done. The work came in building stuff to plot it: I have other files that link "items" to goals or other metadata like categories, etc. Like a relational database, but easily edited in text. For example, to make a link from daily items to Key Results "# item okr" The OKR categories are usually Fighting (I box), Tech, Mental health, and Social, but YMMV Another example is "workout-mapping.vnl" with structure: "# item muscle-group factor" etc etc. vnlog has nice vnl-join commands that quickly build the table and all linked metadata and output formatted text (one command). Then reading that in R is easy ... ``` data <- read.table('../2020-Q4-OKR.vnl',skip=2,stringsAsFactors = FALSE) ``` Aggregating by date is easy: ``` data.agg<- aggregate(x=data$Amount,by=list(data$Date),FUN=sum) ``` Then plot using ggplot2. Examples from 2020 Q4 (just ran the numbers yesterday!) https://josh.vanderhook.info/media/okr.png https://josh.vanderhook.info/media/okr.png https://josh.vanderhook.info/media/workouts.png https://josh.vanderhook.info/media/workouts.png https://josh.vanderhook.info/media/mood.png https://josh.vanderhook.info/media/mood.png
  - thelittleone 6y ago
    Thanks for sharing. Would you consider writing a blog post on your process?
demadog 6y ago
Any books you recommend along these lines?

sh_123 6y ago

Can you share more about your process? I'm very interested in learning more. How do you track your daily data points?

tuatoru 6y ago
There are a lot of these comments on HN, that baldly ask for more information without contributing anything or providing any context. So many that I suspect a bot.
jvanderbot 6y ago
I have a text file in a known location. It's backed up on git. I can open and edit that file quickly using bash. From within VIM I can quickly commit to reduce the # keystrokes. More details: https://github.com/jodavaho/bashlog#git-integration https://github.com/jodavaho/bashlog#git-integration (Really though, you can just save as an environment variable your file location) The text file is structured with space-delimited text, like this: https://github.com/dkogan/vnlog https://github.com/dkogan/vnlog My current structure for tracking a contribution to an OKR item (e.g. "meditate") by date (e.g., 2020-01-15) and amount (e.g., "1") is simply: "# date item amount" Easy to open, easy to plop down today's date and the item / amount. Save multiple date/item entries and add them inscripts ... Easy to save. Done. The work came in building stuff to plot it: I have other files that link "items" to goals or other metadata like categories, etc. Like a relational database, but easily edited in text. For example, to make a link from daily items to Key Results "# item okr" The OKR categories are usually Fighting (I box), Tech, Mental health, and Social, but YMMV Another example is "workout-mapping.vnl" with structure: "# item muscle-group factor" etc etc. vnlog has nice vnl-join commands that quickly build the table and all linked metadata and output formatted text (one command). Then reading that in R is easy ... ``` data <- read.table('../2020-Q4-OKR.vnl',skip=2,stringsAsFactors = FALSE) ``` Aggregating by date is easy: ``` data.agg<- aggregate(x=data$Amount,by=list(data$Date),FUN=sum) ``` Then plot using ggplot2. Examples from 2020 Q4 (just ran the numbers yesterday!) https://josh.vanderhook.info/media/okr.png https://josh.vanderhook.info/media/okr.png https://josh.vanderhook.info/media/workouts.png https://josh.vanderhook.info/media/workouts.png https://josh.vanderhook.info/media/mood.png https://josh.vanderhook.info/media/mood.png
- thelittleone 6y ago
  Thanks for sharing. Would you consider writing a blog post on your process?

demadog 6y ago

Any books you recommend along these lines?