The runway doesn't (Can't) work? Help!

MikeGCT

Registered
This is going to be my probably 4th run at GTD, and all of the previous runs have been unsuccessful. I've read GTD and Making it all work, have soaked up everything I can on this, and buy into the the reasoning behind GTD wholeheartedly. The issue comes when trying to implement in real life.

I know you're only supposed to write down a next action on an action list, not actions that have prerequisite dependencies. So let's say I have "Client Website Project" as a project. I write down a next action on my @work list. Most of what I have to do on this project is linear dependency on the previous action.

In the morning, I see I have to maybe "develop wireframes for approval" This takes 3 hours. This is priority 1. There is a bunch of other next actions on my @work action list related to other projects, but they are lower priority.

I finish the wireframe task. I look at my @work list. There are a ton of next actions on my list. GTD is supposed to make me secure in that I'm doing exactly what I should be. However, there is NOTHING on this list about my project, becuase I just completed the only next action, and every action about this project has a dependency on it.

The lists don't REALLY tell me what the most important next action is, because I can only have the next action for any project on a list. This project is priority 1, and there are no tasks for this project. If I "follow my system" and do other next actions that are on my @work list, I'm actually doing exactly what I SHOULDN'T be doing, because the website project is the highest priority..

Am i meant to constantly be questioning the validity of my system because it doesn't denote the real, highest priority next action I have to take unless I go back every time I complete a task, re-evaluate, and add more tasks? The churn seems unmanageable. I don't see how this fits into real life, and allows you to have confidence in the system.

I hope someone can help me understand or help make my thinking less rigid, because I just don't see how this works. It works for me from 50,000 to 10,000 but I just don't see the runway making any kind of logical, practical sense in the real world. It actually causes me anxiety instead.
 

Oogiem

Registered
I know you're only supposed to write down a next action on an action list, not actions that have prerequisite dependencies. .... However, there is NOTHING on this list about my project, becuase I just completed the only next action, and every action about this project has a dependency on it.

The lists don't REALLY tell me what the most important next action is, because I can only have the next action for any project on a list. This project is priority 1, and there are no tasks for this project. If I "follow my system" and do other next actions that are on my @work list, I'm actually doing exactly what I SHOULDN'T be doing, because the website project is the highest priority..

Am i meant to constantly be questioning the validity of my system because it doesn't denote the real, highest priority next action I have to take unless I go back every time I complete a task, re-evaluate, and add more tasks? The churn seems unmanageable. I don't see how this fits into real life, and allows you to have confidence in the system.

I hope someone can help me understand or help make my thinking less rigid, because I just don't see how this works. It works for me from 50,000 to 10,000 but I just don't see the runway making any kind of logical, practical sense in the real world. It actually causes me anxiety instead.
As someone with huge lists, lots of dependent actions and many time and in some cases life critical projects let me give you my view.

First off NO WHERE does it say that you can only write down 1 next action. Far from it. In fact if you carefully read the requirement is AT LEAST 1 next action. I have projects that I've planned out that might take me a bunch of actions and some sections are dependent. There is no reason you can't put all those actions in at once especially if your GTD list tool allows it and will hide them from you easily.
For example: I have a large AOF working on my LambTracker system. One of the projects is to implement a way to handle embryo transfers. My project looks something like this:

Project: Embryo Transfer sheep are properly assigned to birth and foster dams.
Action: Define database structure for handling genetic dam vs rearing dam.
note: What about frozen embryos where the sire and dam are dead when the lamb is born?​
Action: Modify Create Ewe Breeding Record activity to handle past breedings and support embryo implantation
Action: Modify Lambing Activity to handle rearing dams vs genetic dams.
note: this should also handle the orphan/adoption/foster lamb issue from 2017 with Ginny's lambs​
Action: Update query code to show a rearing dam if it is different from a genetic dam for NSIP EBV analysis

And so on. I know that all those things need to be done and so I go ahead and get them entered in and documented. I use Omnifocus and I keep electronic project support in DEVONThink. OF allows me to set things and hide them until a prerequisite action or set of actions is done. DT allows me to keep lots of support material easily at hand.

Now another issue I see with how you are doing things is that you seem to think that sticking in a context all the time is important. Not necessarily so. While there is a huge mental cost to switching contexts there is also a valid argument made for sometimes taking a single project and working it all through even if you have to change contexts to do it. In the example above I will be using several different software packages to complete that project. Let's say I have some frozen embryos and I'm going to use them this breeding season. That makes getting that working in LambTracker is a high priority project and I might choose to work on it to the exclusion of other projects because the surgery for the recipient ewes might be in a week and I need the code to handle that by then.

Another issue I see is that I suspect you are not reviewing your projects frequently enough. A weekly review is pretty much a requirement but I'v found that with many projects and changing priorities based on weather and other factors that I benefit from a quick 15 minute scan daily.

Another issue is processing incoming work. Even in my life I need about an hour to hour and a half to properly process all the new inputs. If I scrimp on that time I feel it in a vague unease that I ma making the right choices about my actions. The very act of properly processing inputs gives me a chance to see what I have already on my plate and I can adjust. It's not uncommon that a new input that I am adding to my system causes me to decide to put a project on hold to leave space for the new one to be worked on.

And lastly I'd question your contexts. I know the books say start with a simple set @work @phone and so on but you might consider other options some people are very successful with contexts that indicate the amount of brain power needed or a context of tiny actions that take almost no time at all. Play with contexts until you get a set that fits your work. I currently have 34 contexts. It goes up[ and down a bit but I nearly always have at least 30 of them.
 

mcogilvie

Registered
.
I know you're only supposed to write down a next action on an action list, not actions that have prerequisite dependencies. So let's say I have "Client Website Project" as a project. I write down a next action on my @work list. Most of what I have to do on this project is linear dependency on the previous action.

In the morning, I see I have to maybe "develop wireframes for approval" This takes 3 hours. This is priority 1. There is a bunch of other next actions on my @work action list related to other projects, but they are lower priority.

I finish the wireframe task. I look at my @work list. There are a ton of next actions on my list. GTD is supposed to make me secure in that I'm doing exactly what I should be. However, there is NOTHING on this list about my project, becuase I just completed the only next action, and every action about this project has a dependency on it.

The lists don't REALLY tell me what the most important next action is, because I can only have the next action for any project on a list. This project is priority 1, and there are no tasks for this project. If I "follow my system" and do other next actions that are on my @work list, I'm actually doing exactly what I SHOULDN'T be doing, because the website project is the highest priority..

First, to address your immediate question, the best practice is: When you finish a next action, add the next one to your list. The source can be a project plan or other support material, or you can just decide based on your own experience. Projects can have more than one next action, as long as they are not dependent. If you forget to immediately add a next action, you should catch this at your next weekly review. The weekly review is not the only time for project planning, but it is a safety net.

Second, your example next action seems more like a subproject than a next action, both in its wording and the time devoted to it. You would probably benefit from smaller chunks. It gives you more flexibility in doing work, provides a better sense of accomplishment, and opens up little pockets of time where other things which also need to be done can be taken care of.
 

Longstreet

Professor of microbiology and infectious diseases
What @Oogiem and @mcogilvie said....read carefully. I wish to add that it is also perfectly okay to block time on your calendar for important projects. I do this all of the time. This way, you can spend a morning only doing focused work on that important project.
 

MikeGCT

Registered
What @Oogiem and @mcogilvie said....read carefully. I wish to add that it is also perfectly okay to block time on your calendar for important projects. I do this all of the time. This way, you can spend a morning only doing focused work on that important project.

Ok, but I guess my point is this.

I have an important project and a next action. I only want to work on this project today. I finish the next action. Now what? Move onto what i know I have to do for my next action, or go back to my list, add the next action, complete it, check it off, etc.
 

MikeGCT

Registered
First, to address your immediate question, the best practice is: When you finish a next action, add the next one to your list. The source can be a project plan or other support material, or you can just decide based on your own experience. Projects can have more than one next action, as long as they are not dependent. If you forget to immediately add a next action, you should catch this at your next weekly review. The weekly review is not the only time for project planning, but it is a safety net.

Second, your example next action seems more like a subproject than a next action, both in its wording and the time devoted to it. You would probably benefit from smaller chunks. It gives you more flexibility in doing work, provides a better sense of accomplishment, and opens up little pockets of time where other things which also need to be done can be taken care of.

The issue I have with this is I can spend all day breaking down tasks in varying levels of granularity. If everything that's more than one step is a project, I'd have 2394729837492734928 projects on my list. Any software dev project of mine will be defined as a project, then broken down into any number of smaller chunks including epics, features, user stories, and tasks. If every user story was a project, I'd be overloaded and completely unable to move.
 

Longstreet

Professor of microbiology and infectious diseases
The issue I have with this is I can spend all day breaking down tasks in varying levels of granularity. If everything that's more than one step is a project, I'd have 2394729837492734928 projects on my list. Any software dev project of mine will be defined as a project, then broken down into any number of smaller chunks including epics, features, user stories, and tasks. If every user story was a project, I'd be overloaded and completely unable to move.
Then you have to create what you are comfortable with in calling a project. Remember, one has to personalize GTD so that it fits your work style and approach. The basic principles are important, but one can expand. Look at the variety of contexts out there that people have come up with.
 

Longstreet

Professor of microbiology and infectious diseases
Ok, but I guess my point is this.

I have an important project and a next action. I only want to work on this project today. I finish the next action. Now what? Move onto what i know I have to do for my next action, or go back to my list, add the next action, complete it, check it off, etc.
I would just have a few next actions as bookmarks for your project -- they can be sequential -- and then just keep working on the project for a fixed time block. Schedule 2-3 hours in the morning for your project. Work in 50-minute sprints.
 

Oogiem

Registered
Move onto what i know I have to do for my next action, or go back to my list, add the next action, complete it, check it off, etc.
Next actions are bookmarks. Unless you want to there is no requirement to document every little thing you do. So in your case with your example I'd just keep on working until I ran out of steam or otherwise needed to stop on that high priority project. When I did I'd create the next action and add it to my lists so I knew where to pick up again next time.
 

Oogiem

Registered
If everything that's more than one step is a project, I'd have 2394729837492734928 projects on my list. Any software dev project of mine will be defined as a project, then broken down into any number of smaller chunks including epics, features, user stories, and tasks. If every user story was a project, I'd be overloaded and completely unable to move.
It will depend on what you decide works for you. In my case LambTracker is a software development project that is an Area of Focus for me. It's much larger than a single project can handle. I do create separate projects for each major chunk of work. For example I have a project of fix bugs in the Look Up Sheep Activity. Within that are notes about each of the various bugs I've identified. When I am working on that project I actually scan all the bugs I have found and try to correct ones that are nearby in the code so I am not wasting as much time moving back and forth. My code is separated into distinct modules and I will typically have a bunch of separate projects that are contained within that module. Sometimes features are a simple things, one line action of "Add display of Codon 171 genetics to display in Look Up Sheep" That's add a new display field int eh XML code and add the lookup to the query where I get the data to fill the display after scanning and EID tag. I already have the code written I just need to integrate it into the program and test it. Sometimes they are far more complex and end up being their own project like "Add display of latest EBVs for Lambease, Self replacing carcass index, Maternal Weaning Weight, Number lambs weaned and Hogget weight to extended display as superset of Look Up Sheep" That turns out to actually be a whole 'nother module, and thus a whole big project was formed called "Create a genetics display module"

That sort fo granularity is why I typically have between 200-300 active projects at any given time.
 

Gardener

Registered
Assorted comments:

- While I do prefer to have exactly one next action per project, I don't think that's actually a rule. I used to think it was, but I think that almost everyone has more than one action per project. I think that it can be (different for everybody, of course) a mistake to have a long, long list of actions per project, but adding enough actions to get you to the next weekly review seems fine.

- If you do like to have exactly one next action per project, another option is to always add a new action when you finish the last one.

- You mention that you write down the action. Does that mean that you're using a paper system? There's nothing wrong with a paper system, but if you've never tried a software-based system, it may be worth a try. In OmniFocus, for example, I could set a project to be "sequential" and enter five or ten tasks for that project. Only the "top" task would be visible (well, depending on how I organize my views), and when each task is checked off, the next one would automatically float up and be visible. (Edited to add: When I used to do this, the last task in the sequence was often "Add more actions for this project" so that when I finish the last real action, the project doesn't vanish.)

- It's not unusual to do the next action for a project and then just keep working on that project when that action is done. Then, after the work session is done, you add a Next Action to the project so that the project doesn't vanish from your attention. That means that lots of actions are not "documented" as completed in your GTD system, but unless there's a reason why they need to be, that doesn't need to be a problem.

- It may be that your work plan for a project already exists somewhere else, in project support material. In that case, your action for that project might just be "Work on Project Blah", and you may go to another document to choose the tasks.
 

Gardener

Registered
The issue I have with this is I can spend all day breaking down tasks in varying levels of granularity. If everything that's more than one step is a project, I'd have 2394729837492734928 projects on my list. Any software dev project of mine will be defined as a project, then broken down into any number of smaller chunks including epics, features, user stories, and tasks. If every user story was a project, I'd be overloaded and completely unable to move.

My view on this is that an action is something that you can "just do". And that depends on you and your skills.

For example, for an experienced baker, "Make birthday cake for Joe" may be a single action.

For someone who's never baked anything, "Make birthday cake for Joe" is probably a project. It might even be more than one project.
 

mcogilvie

Registered
The issue I have with this is I can spend all day breaking down tasks in varying levels of granularity. If everything that's more than one step is a project, I'd have 2394729837492734928 projects on my list. Any software dev project of mine will be defined as a project, then broken down into any number of smaller chunks including epics, features, user stories, and tasks. If every user story was a project, I'd be overloaded and completely unable to move.

As Oogiem wrote, next actions are bookmarks. You say you want to spend all day working on one project. Fine, but is that really true? Do you take breaks, read email, answer phone calls? What happens if something of higher priority drops in your lap? What I have learned to do is place bookmarks (next actions) on my lists that tell me where I want to start from when I come back to what I was doing. Thank of it as being kind to your future self. It really has little to do with how you plan or don't plan your projects. However, when I look at something I think I need to do, any resistance I feel often is a sign that I haven't really figured out either what exactly I want to do or how I want to start. Just knowing that is very helpful for me.

By the way, you don't actually have 2394729837492734928 projects. The fact that you feel that you do, even in jest, is probably an indication that you can benefit from GTD. But some people don't take to it, and I understand that.
 

MikeGCT

Registered
As someone with huge lists, lots of dependent actions and many time and in some cases life critical projects let me give you my view.

First off NO WHERE does it say that you can only write down 1 next action. Far from it. In fact if you carefully read the requirement is AT LEAST 1 next action. I have projects that I've planned out that might take me a bunch of actions and some sections are dependent. There is no reason you can't put all those actions in at once especially if your GTD list tool allows it and will hide them from you easily.
For example: I have a large AOF working on my LambTracker system. One of the projects is to implement a way to handle embryo transfers. My project looks something like this:

OK so this makes me feel better re: real world implementation.

For example, when I'm scanning my projects and I have a project like:

Implement Helpdesk Report

I'm likely to want to add something like:
  • Get database name and login info from Mike
  • Research database schema
  • Identify key fields between datbase and spreadsheet
  • Create merged report.
I know I have to do all 4. They are the next actions. They are ALSO dependent on each other. But if I don't add them, I have to make 4 trips to the system (I'm using todoist) to add these, when I know full well they all need to be done, even if it's not until the previous one is done.

In todoist, I have a next_action label, (which is the same way contexts are applied) and I'm thinking can tag @next_action to the first task, and leave the rest. So the first task shows up in my Work Next Actions filter. Then if I finish it and am not sure what to do next, I can look at the project list in my tool for that project and see the other "bookmarks" to identify what to work on. IF at that point i came up with 3 more tasks from doing that one task, I can add all 3. At the end of the day, I can probably set a next_action for that project for the next day, or always set the next_action when I finish a task.

I guess this really blurs the lines between having context based lists and project support material, by having lists per project, but I don't know how else it could possibly work. If I have to open a project support folder every time I need a new task to do instead of just looking at my system, there's NO WAY I'll ever do it, it's too much overhead.
 

Gardener

Registered
I guess this really blurs the lines between having context based lists and project support material, by having lists per project, but I don't know how else it could possibly work. If I have to open a project support folder every time I need a new task to do instead of just looking at my system, there's NO WAY I'll ever do it, it's too much overhead.

Lists per project seems totally normal to me. As I see it, you have a basket of tasks that you can "cut" at least two different ways--viewed by project, or viewed by context.

Some GTD tools, such as OmniFocus (the only one I really know, but it only works on the Mac) do that for you. But I think that a modest list of near-term actions for a project, inside your main GTD lists, is perfectly compatible with GTD. I say "modest" list because if it's a long project with dozens of actions, I want most of those actions outside and in project support material--or I don't write them down at all until I'm closer to doing them. But I think that some people do list quite a few actions per project.
 

MikeGCT

Registered
...because if it's a long project with dozens of actions, I want most of those actions outside and in project support material--or I don't write them down at all until I'm closer to doing them. But I think that some people do list quite a few actions per project.
Yes this makes perfect sense to me.
 

aderoy

Registered
Some GTD tools, such as OmniFocus (the only one I really know, but it only works on the Mac) do that for you. But I think that a modest list of near-term actions for a project, inside your main GTD lists, is perfectly compatible with GTD.

EMACS with ORGmode can also handle projects , tasks and task with dependencies via filtering. It is plain text only, yet available on just about any computer platform. i.e. use a Win7 machine at work, yet Mac OSX at home no issues starting on one then continuing on another.
 

bcmyers2112

Registered
I hope someone can help me understand or help make my thinking less rigid, because I just don't see how this works.

I'm going to address this a little more simply (and I apologize if it comes off as snarky, because I'm actually trying to be helpful): if you want to be less rigid in your thinking, then be less rigid in your thinking.

If you complete an action related to your highest priority project and it makes sense to continue on to the next action in the sequence (assuming it's a project where each action is dependent on the previous one, as you've described) without reviewing the other actions in your lists, then do so. It's that simple.

Chances are if the project is as linear as you describe, you already know what the next action is. If you don't and it's hard to find those "future actions" in your lists, streamline your system to make it easier to use.

If the concern is "promoting" a future action to your next actions lists so you have it recorded in case you get interrupted, here's a suggestion: jot something down on a sheet of paper before you allow that interruption to take hold (whether it's answering the phone, talking with a co-worker/family member/whoever who drops by, or what have you) and toss it in your inbox. If you're processing your inbox regularly, it'll be there.

Of course, if "promoting" a future action is that hard -- again, I would also suggest streamlining your system.

Don't approach GTD so rigidly that it becomes a straitjacket rather than an enabler. It's meant to make things easier, not harder.
 

Mike Simms

Registered
There are many good suggestions already; some thoughts:

Suppose you are set to work on a high-priority project for a whole morning, and you have one next action on this project's list. Now suppose you start that next action, then finish it in, say, 1 hour, and at that time you know what to do next to get closer to the outcome of the project, then go ahead and start it, without thinking about the lists. If after finishing this second bit of work, it is again obvious what to do next, then go ahead and start that - even if you are not recording it anywhere on your system. I would call that being in the 'flow'.

If on the other hand you get interrupted and have to stop and do something else, then make sure you at least write down what you were doing at that moment, so that you can pick up next time you get back to this project, whether that's after lunch or in a month's time, with as little time as possible spent figuring out how far you had got. This kind of thing makes the runway work of me. The object is to get closer to the outcome (the "move the needle" as DA says); not to record everything along the way, (assuming you don't need to)

Also, You say "If I have to open a project support folder every time I need a new task to do instead of just looking at my system, there's NO WAY I'll ever do it, it's too much overhead." I use Emacs org-mode for my lists, yet when throwing 3-4 hours at a project I find that a pen and paper is the the best way to keep a note of any 'things for this project that I think of along the way but don't want to do this second but may be worthwhile'. This piece of paper IS my project support material. It may loose thoughts and ideas that are to woolly yet to be next actions but it is a source of possible ideas about what to do next. This piece of paper hangs around in a project support material folder along with similar pages for unrelated projects: what I'm saying is that referring to the top layer of your project support material should be easy.
 

Longstreet

Professor of microbiology and infectious diseases
I'm going to address this a little more simply (and I apologize if it comes off as snarky, because I'm actually trying to be helpful): if you want to be less rigid in your thinking, then be less rigid in your thinking.

If you complete an action related to your highest priority project and it makes sense to continue on to the next action in the sequence (assuming it's a project where each action is dependent on the previous one, as you've described) without reviewing the other actions in your lists, then do so. It's that simple.

Chances are if the project is as linear as you describe, you already know what the next action is. If you don't and it's hard to find those "future actions" in your lists, streamline your system to make it easier to use.

If the concern is "promoting" a future action to your next actions lists so you have it recorded in case you get interrupted, here's a suggestion: jot something down on a sheet of paper before you allow that interruption to take hold (whether it's answering the phone, talking with a co-worker/family member/whoever who drops by, or what have you) and toss it in your inbox. If you're processing your inbox regularly, it'll be there.

Of course, if "promoting" a future action is that hard -- again, I would also suggest streamlining your system.

Don't approach GTD so rigidly that it becomes a straitjacket rather than an enabler. It's meant to make things easier, not harder.
All of the advice has been excellent, but this is really good! Don't become so rigid in your GTD practice that you become a slave of your system and making sure you are following GTD to the letter. The system is wonderful, but you do have to personalize it so that it works for you. As everyone knows here, I am a champion of time blocking. I time block for deep work in the mornings, which usually involves a major project. Remember that next actions are indeed bookmarks. Add as many actions as you need in your project so that you can obtain flow. And then do focused work! When you stop, or are interrupted with something urgent that requires your attention, add a next action as a bookmark so that you know where you left off.
 
Top