Knowing how long tasks take is essential

Focused
21

I will say the best time tracking and adjustment loops are in the context of doing the work itself and ideally automatic. E.g.

  • Task systems in PM software running timers when you pull items off the shelf, that feed into reports on similar cascades of cards/templates/projects
  • Browser/window title tracking software that spots trends (I used to send reports from this straight to Freshbooks billing for freelance)
  • Time block planning (as popularized by Cal Newport) only asks you to develop better heuristics over time, shaped by seeing your right and left columns diverge often enough
  • David and Rosemary Orchard track everything but automate most of it (with their bread-and-butter skills, too)

There are a lot of specialized tools along these lines.

In some old studies we found manual time tracking and estimate correction to add a huge overhead, up to 20%, and caused burnout if time wasn’t allotted for it. And any time tracking that wasn’t obviously used for a good (or at least understandable) purpose was alienating.

22

Hear hear! That pretty much sums up what I’ve been thinking but unable to express nearly as well. I’ve used time blocking, not tracking, for years. I guestimate or assign time to work on deep tasks during the week. All other times are reserved for standing meetings or unstructured “people time.” What I don’t do is use an app to track how I spend my time meticulously. My calendar essentially does that for me. I would find using a time tracking app to add stress, not reduce it, and it would take time to track time, which I don’t believe would result in more productive use of time.

I’ve probably shared this before, but I have a calendar “template” that pre-schedules my time before my EA begins to add appointments. This way, I protect my deep work time as much as feasible, recognizing that interruptions will always be unavoidable.

In short, I pre-plan my work and then get to work. No tracking is needed.

Below is the “template” calendar that starts each year (I operate on an academic year, not a calendar year). Below that is how I create a rough outline of my “big rock” goals for the year and then I outline key projects per month in Apple Notes. The detailed tasks with dates are kept in Reminders.

September 11, 2024
September 11, 20242732×1629 480 KB

Priorities for the Year and by Month
Priorities for the Year and by Month2732×2048 252 KB

Our strategic plan provides the roadmap for my priorities and how I block my time.

3 Likes
23

Absolutely. Being able to accurately judge how long a task will take you to complete is essential it a lot of areas, and as a programmer I consider it a key part of being successful.

Without this, it is very hard to judge which tasks are worth doing and which are not, or which items on the todo list should be done first, and which will be prohibitively expensive to implement.

And people (especially programmers?) are astoundingly bad at it. Over and over they estimate how long a job will take and then find it takes much longer.

So as you note, honestly evaluating this is essential. If nothing else, when doing a task, estimate how long it will take and then note how far out you were, and then next time you make an estimate, adjust by that percentage, and repeat until such time as you are able to more accurately estimate.

2 Likes
24

Just wanted to add that, apparently Todoist is working on task durations.

https://twitter.com/amix3k/status/1679047472729997312

Which is interesting, because, combined with start dates I believe they are on course to extend the “Upcoming” view to a full calendar --which, IMHO is critical for time blocking. I may switch to Todoist just for this!

25

I think time-tracking can be really powerful for evaluating where your time REALLY goes, rather than where you think it goes.

My example: as an academic, my work roughly falls into three categories: teaching, research, and admin. My employers create ‘workload allocation models’ that put percentages to these three buckets. So how much time do I spend on any of these, on average, over the year? Clearly research is up in the summer; and takes a nose-dive when marking season opens. Seasonality across the year makes it hard to get real world data, when all your have to go on is your own feelings in the moment.

How good am I (are we as a group of professionals) at assessing our time spent working on these buckets from memory? Not very, in my view. There is research, I think, that suggests we routinely overestimate our working hours. We also (academics, I mean) routinely complain that admin and bureaucracy runs our lives (it does); traps us in the weeds of administering our jobs (it does), keeps us from the ivory towers where the REAL WORK is meant to be happening. I am overegging it here – you get the picture.

So I’ve time-tracked over the last 3 years or so (Toggl) to get an honest account of where my time goes. There were some surprises - I spend way less time marking than I think I did (not the eternities one feels when in the middle of it!). I spend way more on email and unquantifiable admin. I have fewer hours of deep thinking in me in any one day than I would like, but I know my mornings are best for this.

These insights, in part accrued through tracking time and otherwise staying mindful about how I work, allowed me to change my habits where possible; hold my employers accountable (hmmm, well, maybe) when I over-invest on non-essential work; say no to longer-term commitments; set priorities in ways that seems at least a little more achievable over time. Would I like to review another grant application? Let me see: how long did that take me that last 3 times I said yes … ok - so where will I find 8-10 hours over the next four weeks? And so on.

So for me, at least, it’s been essential to learn about my own working practices. Should I still be doing it? Not sure, to be honest, but it’s become a habit. NB: I find the comments above on overload and stress very interesting. I don’t think I have concern about the first, but agree that micro-managing time can cause me more stress. Food for thought.

Right – now I better log the 20 minutes procrastination time when I should have been writing a report.

4 Likes
26

This probably merits its own thread.

I work in tech consulting, coming from an engineering background. I do strategy work (working with our own MBAs and client Cxx stakeholders) but some times I’m also involved in delivery itself (working with engineers). In both aspects, trying to come up with correct estimates is impossible as there are usually a lot of assumptions over uncertain requirements.

But while with consulting work you can absorb it with crunch time (the occassional all-nighter with a deck or a spreadsheet), on the other hand within the context of an engineering project this is something you cannot do with any team and expect acceptable quality on the final solution. I would assume that academic research work is similar to the later scenario. That’s why a lot of the “strategic” overwork in my case involves trying to come up with realistic estimates for the “engineering” work, where a project overrun can destroy your bottom line for the quarter.

PS: Don’t get me started on Agile! XD

1 Like
27

They’re also a lot more expensive than they used to be and have removed the non-profit discount once offered and don’t appear to offer student discounts either. £50 a year for a task manager is a bit much when there’s reminders which is native.

28

I finally quit for Things 3 - I got so mad when they removed their educator discount.

1 Like
29

Is this the current price for new subs? I paid £36 last time I renewed. Regardless it’s not cheap, and I agree with @ERJ_T that it’s a shame they removed the edu discount. They said at the time they’d review and perhaps reintroduce it at a later point but clearly that hasn’t happened. Edu discounts make my subscriptions a lot more palatable!

1 Like
30

And I think developers should take note - teachers talk and share the tools that help them do their jobs. Craft, for awhile, was giving “free” edu licenses (I’m sure they’ll end at some point), and I was able to try it out and turn a few colleagues onto Craft. They’ll probably turn into paying customers later down the line.

1 Like
31

Your last paragraph has been my practice and it’s so nice (after a few years of doing that) to have a VERY good idea of how long things take me. It’s been a game changer for being predictably productive.

1 Like
32

I am sure you and your wife have done time tracking in some form or another.

It is hard to fathom anyone passing medical school without even rudimentary time tracking - when to revise, what to revise, time until exam, etc. And micro time tracking in an exam - 1.5 minutes per MCQ, etc.

And you run a “busy practice” and, presumably, don’t track meetings or appointments.

Your wife, assuming she is in a speciality that sees patients would have an appointment book - the ultimate time tracking tool.

33

Back when I had projects to manage the length of tasks was pretty much guess work. One could collect data and perform something like a least-squares regression in the same way that software development estimation was done with CoCoMo but that requires a different guess as to the number of lines of code involved. I solved the LOC problem but front ending my CoCoMo spreadsheet with a function point sheet. FPs were calibrated by (someone else) lines of code needed in specific languages.

I no longer work in software development so the spreadsheet is long gone but it was a simple transciption of the descriptions of each method from a book on software engineering (and permission from the CoCoMo developer). The book also included details of interruptions based on work by Esterling which helped to refine the CoCoMo output on “mythical man hours” into real time.

Used this sequence of spreadsheets in anger many times and vividly recall the day I sat with a customer to tell them how long their project would take and reminded them of as I walked out to the day I predicted it would all be wrapped up.

TL;DR If there is no historic data to work from then how long a task will take is entirely guess work. Think it was Newton who said something like “improvement only happens after measurement”.

34

Agreed, but some guesses are more educated than others. There’s often a parallel or analogous task on which you do have “historic data” (or at least reasonable experience). The real problems are from hopelessly unrealistic estimates.

35

You mean those made by managers who have no practical experience of the issues/tasks involved. Sadly had to deal with those, typically exhibiting the Peter Principle. They think that because they held their wetted finger up to test the wind that their guess is written in stone and unamendable. Of course they never end up having to carry the can for late delievery.

EDIT: Oops typos. Most egregious now corrected.

1 Like
36

Time tracking surely is a retrospective exercise? I think it’s important to draw the distinction between time tracking and planning your time, they’re entirely different.

I, like most, plan my time. When do I have meetings, when do I want to complete a task. I do not track my time in any meaningful way.

And to be clear, estimating that a task will take 2 hours is NOT time tracking.

37

Yes time tracking is a retrospective activity that is used for planning. Even rudimentary time tracking is used by students. How long did it take me to read that chapter of West’s Respiratory Physiology last week? Half an hour? Okay , I’ll use that to plan my time reading the next chapter.
Time tracking does not need to track every nanosecond of what you’ve done, but even if this is “done in your head” and not formalised with an app, it’s still time tracking.

Oh, and you use time tracking with your appointments. You know using retrospective data how long a procedure will take and use this data to plan your appointments. You know which patients are more complex and need double appointments.

Time tracking is more nuanced than looking at what you’ve done that day/week. Anyone that says they don’t time track in one form or another isn’t really thinking about it.

38

So true. However this thread reminded me of the best, most accurate way I’ve ever seen for calculating how long programming tasks will take. Evidence Based Scheduling

1 Like