Beeminder Forum

How to build a habit of finishing things?

This is more ‘general productivity’ than beeminder specific, though ideally once I have a bit of a plan I’d like to beemind it.

I have a common problem — things I’ve started but not finished. But most of the advice I’ve found for these scenarios focuses on finding ways to put in more time or to rediscover the enjoyment of the project. I actually do put in what I think is enough time, for the things that don’t seem to be going anywhere — I just work on the wrong things.

My #1 goal for 2018 was to revise and submit a specific journal article for which I had a complete first draft. I submitted five other articles instead. It’s easy to spot why those five were easier: they were collaborative, so I was accountable to my co-writers, AND they had firm deadlines. Not the case for the work I set out to publish.

I tried to give myself collaborators and a deadline by starting a weekly writing group, working our way through the book “Writing Your Journal Article In Twelve Weeks,” which was fun and improved the article immensely but which did not lead to actual completion for me.

When I sit down to work on this article, or any of the others in a similar state, I do whatever seems most fun rather than whatever’s most important to completion. Most counterproductively, I keep inventing new research to do, instead of polishing the research I already have. The research is done! I just need to write it up so someone else can understand it!

I’d like to try tackling this from the angle of building up my “intrinsic drive to completion” muscles. I’m thinking about some kind of guided freewriting, “what can I do next that will get this paper closest to done,” on a regular schedule. And hopefully over time I’ll get used to working on those tasks first… Does anyone have advice or reading recommendations?


That sounds very familiar to me!
I beeminded time I spend on my Master’s project for uni and that led to adding more and more features rather than tying it all up and finishing it.
I then threw Trello into the mix. I wrote out all that was left to do as cards (= tasks) and beeminded doing so and so many of them a day.
That worked great for a week or so but eventually I started adding more cards to it and doing those. So I was somewhat back to where I started. But in the time it worked for me it provided me with a strong feel of focus.

I have suggested adding a “wittle down” type to the trello integration basically mandating that you do all remaining tasks until a certain deadline.
By adding more and more cards you just increase the burden on you rather than replacing it with something you’d rather do. And with beeminder breaking this bursen down onto individual days this should incentivise not doing so.
So I think it’s best if you poke @dreev so it gets implemented :wink:
Currently it’s only a “do more” type goal which rewards you for adding (and then doing) more (supposedly fun to do) tasks rather than the ones you already specified.


You might find these discussions about habits interesting:

My general strategy for any kind of project is to beemind both input (effort) and output (outcome). Input in this case would be amount of hours spent on the project. Output would be number of words written. Both can be automatically beeminded. Tackling it from both sides allows you to target for results (output) but also get some motivational rewards when you put the effort but the output isn’t coming.

That said, neglecting projects to work on other projects that seem more fun or even inventing new projects is typicall procrastination. My favourite resource is Dr Timothy Pychyl, his book is here and I’ve collected some highlights here
He also has a backlog of podcasts on the topic going back to 2006 (on and iTunes (“iProcrastinate podcast”))

I would also look at mindfull meditation, at it helps on practicing focus and more importantly coming back to your point of focus when you wander of.


How do you auto-beemind hours spent on the project?

I use Toggl time-tracking to beemind the hours spent on this project (and a wide range of others).

Thank you, I think this is really what is tripping me up! I have reached the total number of words written that I need. (In fact, I am a little too long, and need to find places to cut or be more concise.) Tracking words worked great to generate the rough draft, but your framework helps me define one of my problems as “right now I don’t have an output to track”.

Percentage-completion of a to-do list might be the output I am looking for, here… thank you! I could do a whittle-down goal with manual data, if I produced a concrete to-do list of the tasks that actually need to be taken care of before I move on with my draft.


With RescueTime, you can assign different applications to a sub-category of your choice. You can also create custom sub-categories (“Project X”). Then you can beemind time spent on the specific sub-category.

Sounds silly, but I actually use three different text editors to track time spent coding on three major projects :slight_smile:

Any kind of to-do app that integrates with Beemind will do the trick as long as you don’t add any items on the list. For example, one item per “page/section that is fully edited” or something to that effect
As you noted, just find a metric that represents progress to you.

By the way, for some reason whittle down goals (“pages left to be edited”) feel more empowering to me rather do more (“pages fully edited”).

This is exactly what I want to avoid :smiley: Since I now use VS Code as much as possible I can’t easily throw that into RescueTime because to that it’s just a black box of productivity (I feel like this could be a D&D item). It won’t know if I edited LaTeX in there or programmed this or that or anything.
Which is one of the reasons I got “Make a Beeminder VSCode extension” on my list so VSCode can track my time more accurately.

But right now I just prop up my Smartphone, let the Beeminder app run with the Stopwatch on and… hope it does not forget for how long it ran (because the iOS app totally does this all the time and it is driving me nuts).


If you’re still looking for completion of a to-do list you might find this helpful: Turns out with custom goals you can already make a whittle-down trello goal! See here:

Thank you, I might try that next time! For now I just made a little spreadsheet of tasks that makes a pie chart for me and calculates how many to-dos remain, and I use that for a manual goal. Since I already did all that work, no reason to switch to something else.

Going through a concrete to-do list definitely feels like it’s pushing me towards completion – it forces me to do more than just noodle on the parts that are already fine as-is. I’ve made SOME progress so far!


omg I love it! I know what you mean! To me it gives this feeling of urgency, focus, a certain clarity. I would not want to get rid of tracking the “input” (read: the hours put into it) though.
In my experience tracking both the input and the output (there’s a recent post about this somewhere) really goes well together. I need a constant base velocity (the former) and a direction (the latter) and then it’s a really smooth experience.


Can you elaborate? How does your manual goal work and how does your spreadsheet calculate?

This is what the spreadsheet looks like:

It updates the “done” and “todo” totals based on how many I’ve checked off.

The goal is a whittle-down goal with manual data entry. When I make progress, I enter the total number of todos remaining. Does that make sense?


Agreed!! It’s crucial. But also pretty difficult, for this goal, to work out a trackable concrete outcome: just making the to-do lists of editing tasks took more than one work session.

So I guess for the meta-goal of “finishing things” I am still not sure what outputs or inputs to track. (Well, I track the output of “finished journal articles” but these days it’s “finished dissertation chapters” that are most urgent…).

For an individual project at the mature editing stage, the to-do list seems to be working, but I’m not sure how to generalize to the meta-skill applying to multiple projects. And even for other projects I’m not sure how I can replicate this structure – I don’t think I could make a very good to-do list for my dissertation chapter. What works well with the gothic-motifs article is that each todo is literally a single sentence that needs some kind of revision. Some of those revisions involve me doing a day of research just to write six words (like the quote I read in a book five years ago and now need to actually find), but they’re still fully ‘known’ tasks.

1 Like

Yes, love it! Pretty graphs :slight_smile:

@phi @oulfis @apolyton

My problem with tracking outputs is that they’re not under your direct control and I don’t want to get charged for not finishing an output when it wasn’t under my control. Why do you say it’s crucial to track both?

1 Like

Well, they are at least under your indirect control! For me, tracking an output is how I evaluate whether the inputs I track are really working. My most “outputty” goal is definitely my “publications on my CV” goal, which you might notice has a totally flat road for two entire years, because every few months I’d say, “well, I’m not on track to get a point here in the next few months,” and add more flat to the road. Having to look at that completely flat graph was extremely helpful to me in debugging my various writing goals.

I think that’s the only “output,” actually, that I’d say I don’t have direct control over (since some of those points rely on my work being accepted). The others are all things like sending my weekly newsletter (for which I beemind sections-written and newsletters-sent), the current wordcount of my diss (a number which DISTRESSINGLY must sometmes go down!), the size of grading backlog… it might take an unpredictable amount of effort to meet these goals, but I can change my own actions in order to meet them, especially if I’m looking at them with enough lead time. Maybe others track more hardcore outputs? But for me at least, these kinds of metrics are important to keep the more open-ended “hours spent” goals pointed in a useful direction.


I’ve convinced myself that if, by “finishing things” I mean “finishing my dissertation” (which I largely do! there are other unfinished projects lurking around but this is the most important one), then I should beemind more outputs. And thus, a goal for completing chapter drafts. It’s set to the perhaps-ambitious slope that would indicate “on time completion”… my department’s average time to graduation is 7 years but I haven’t let go of the dream of doing it in 5.


It’s quite to easy for me to get side-tracked in working on a project but not focusing on the important stuff. So some form of output needs to be beeminded as well.


Is there an input you could track instead?