Scoring Maker's Schedule vs Manager's Schedule Using Herfindahl Index
Maker’s schedule vs manager’s schedule is a way contrasting the type of schedule most productive for someone who’s doing work that requires deep concentration vs someone who’s doing work that requires managing many small discreet tasks and context switches. Most people in tech have read Paul Graham’s essay on the topic, and if they haven’t simply telling them the title is usually enough to convey the idea. The essay names something that everyone already knows and makes it easier to talk about.
Now because programmer’s love precision I’ve had a few debates about if a given schedule is a maker’s or a manager’s. You probably have a feeling for if you have a maker’s schedule or a manager’s schedule, but that’s not quantifiable. Having worked on both schedules I thought there should be a way to score and quantify it. When building something new you want a maker’s schedule and when shipping and selling something you want a manager’s schedule.
You can’t manage what you don’t measure so this is a way to score if you have a Maker’s schedule or a Manager’s schedule.
A ‘good’ way of scoring this would be:
More nuanced than binary, most people work something in between.
Easy to understand and interpret. Ideally scored from 0-100. With 0 being someone who’s whole day is 30 minute sales calls and 100 being a PhD student who should be working on their dissertation.
More meetings give you a lower score. If you’re a manager it might bother you that you should probably have a lower score. If it helps think of it like golf where a lower score is better. Or present it as “I’m 30% individual contributor and 70% manager” instead.
Penalize a meeting that break up a long block of uninterrupted time more than a back to back meeting or one at the start or end of your day.
Two 30 minute meetings are worse than one 60 minute meeting. This is because there is less context switching in the 60 minutes meeting. You’re more likely to be thinking deeply in a 60 minute meeting than a 30 minute meeting.
The final score should be on the whole week. I schedule all of my meetings on Wednesday and Thursday so those days have terrible scores, while the rest of my week is free.
Herfindahl Hirschman Index
In economics there is the concept called Herfindahl Hirschman Index. HHI is best known for measuring market concentration. It quantifies if a given industry is only a few big firms, or even a monopoly vs many small firms fighting for control. It ranges from 0 to 10,000 and takes into account both the size of the firm and the number of firms. It’s computed by summing the squares of each firm’s market share.
So in a monopoly where one firm controls the entire market the HHI is 1002 = 10,000
If two firms each control half the market the HHI is 502 + 502 = 5,000
If one firm controls 50% another firm controls 30% and a third controls 20% then the HHI is 502 + 302 + 202 = 3,800.
Applying this to your schedule
Your day is the entire ‘market’ and every ‘slot’ of time, be it free or a meeting, is a ‘firm’ whose market share is its percentage of your day.
For simplicity imagine you worked a 10 hour day, 600 minutes, from 8 AM to 6 PM. If you had the entire morning free, followed by a meeting at 1 PM, and then the rest of your afternoon free, your schedule would be:
|Slot of Time||Length||% of day|
|8 AM to 1 PM||5 hours||50|
|1 PM to 2 PM||1 hour||10|
|2 PM to 6 PM||4 hours||40|
Then the HHI score would be 502 + 102 + 402 = 4,200.
However you can’t look at a single day’s HHI score in isolation. For example I try to schedule all of my meetings on Wednesday. This means that my score for Wednesday looks terrible but the rest of my week is relatively free. By averaging a work week’s HHI scores you can get a better view of the person’s overall day.
Examples from a week of 10 hour days
If you have “No meeting Monday” then Monday has a perfect score of 10,000 (100%). Not that interesting but useful as a point of comparison.
If Tuesday you have a single 30 minute meeting at the start (or end) of your day your score only drops to 9,050 (90.50%).
If on Wednesday however that meetings is two hours later in your day score drops drastically to 6,050 (60.5%). This is because the meeting now cuts your day into two halves.
If on Thursday you have ten 30 minute meetings back to back in the afternoon your score drops to 2,750 (27.5%). In the original essay it is mentioned that if you need to split your day between maker and manager you can put all of your meetings in half the day and block the other half off to do maker work. Coincidentally this is also fairly close to 2500 which the DOJ considers the cutoff for a highly concentrated industry that may need to be regulated. Of interest is that even if you only had five 30 minute meetings but with 30 minutes in between each your score would be the same.
30 minutes is only long enough to check email
If on Friday your 10 hour day is back to back 30 min blocks of time you score a comically low 500 (5%). I doubt anyone with a white collar job actually has that type of schedule, but it’s good to have as a point of comparison.
I you average the entire week you end up with a score of 57%, somewhere between maker and manager. This intuitively makes sense as Monday and Tuesday are clearly maker days and even Wednesday isn’t too bad. But Thursday and Friday both dramatically bring down the average for the week.
Proof of Concept
Once I saw that the scores made sense I turned the script into a Chrome Extension. The proof of concept currently works only in Google Calendar. But the algorithm is straightforward enough to be implemented for any calendaring system.
Ways to Improve your score.
The point of writing a scoring system is to have a more nuanced way to talk about our schedule. If you need to shift between being a maker schedule and a manager schedule it’s much easier to see which you are on with numbers. If you are about to launch something then you should probably ignore your score and focus on what needs to be done. But if you’re building something you should be shooting for a higher score.
Ways to raise your score.
Say no to meetings. Unless they are obviously worth going to.
Cancel meetings. I have a weekly status meeting that gets cancelled 1/3 of the time if no one has anything to talk about.
Schedule your meetings back to back. This way you get them all done in a shorter period of time
Put blocks in your schedule (I usually like to do either the entire morning or afternoon)