Productivity Metrics
As part of the ProductivityImprovement project, it makes sense to keep some sort of metrics to measure whether productivity is being improved (or not), and if so, how much.
criteria
- metrics must encourage better productivity, and must NOT encourage worse productivity. (obvious) This may sound like a tautology, but is specifically meant to address the fact that often what is measurable does not reflect what qualitatively matters. If/when a metric appears to be encouraging less productive behavior, the metric should be critically analyzed and deconstructed until it is fixed to no longer encourage worse behavior.
- keeping metrics should take minimal time, effort, and cognitive load. The keeping of metrics itself is not directly contributing to getting things done, but rather, only applying indirect pressure to get things done. Thus any time spent on such meta-productivity should be considered a "tax" on productivity and thus minimized.
checkpoints
Checkpoints are the coarsest of measurements, and yet, they may be better than nothing. A key advantage of checkpoints is that they are (or certainly should be) relatively easy to evaluate (thereby helping to satisfy criteria #2).
The following checkpoints specifically apply to Getting Things Done (GTD) terms and concepts. Yes answers indicate better performance.
daily checkpoint metrics
- Were you able to log metrics for the previous day? (added and logged 2008-12-15 on) A metametric, that is, perhaps the most important aspect of keeping metrics is being diligent about recording them promptly. If metrics weren't recorded at all, it likely indicates that non-trivial portion of the GTD process for the day failed, since recording metrics for the previous day is one of the first things one should do each day. Thus if that didn't get done, then likely most things following it didn't get done either.
- Were you able to be on time to or otherwise successfully complete on time all date and time dependent actions today?
- ...including all items placed in the calendar for that day? (not a separate metric, but requirement for the above)
- the intent of this detail is to discourage putting things in the calendar that should simply be in the next actions list.
- Gaining sufficient prediction ability and control over your time is a key rough indicator of how in control you are over your own time, in contrast to others' controlling your time.
- Were you at least 5 minutes early to all time dependent items?
- Being a little early to appointments and meetings shows not only even better prediction ability and control, but also the ability to consistently provide a buffer and thus reduce risk of being late due to low probability interruptions or obstacles delays.
- Were you able to process at least one item (if there are any) in a physical inbox? (added 2008-12-23, start logging 2009)
- By encouraging yourself to process at least one item, once you get into that processing mode, you will likely process a few more, and thus hopefully make a dent in remaining inbox items. Also, encourages the habit of doing at least *some* processing per day.
- Were you able to process all physical inbox(es) down to empty sometime during the day? (added 2008-12-23, start logging 2009)
- ... including at least one item of incremental net long term ClutterReduction in at least one room? (not a separate metric, but requirement for the above, added 2009-163 to raise the bar on physical inboxes 0)
- the intent of this detail is to make continuous forward progress on living space ClutterReduction. For at least one room per day: pick up / neaten up everything to the previously achieved minimal clutter level, and then improve upon that by reducing the number of objects in view, until the room can be considered free of clutter by reasonable outside observers.
- Were you able to process at least one item in an electronic inbox? (added 2008-12-23, start logging 2009)
- Reducing physical clutter can help clear your head and increase the rate at which you process additional items. Keeping physical inboxes empty helps with this.
- Were you able to process all your inbox(es) down to empty sometime during the day?
- Why not empty at end of day?
- Since per GTD it is possible to collect into inbox(es) at any point during the day (or night), and yet, it may make sense to only spend one focused chunk of time per day "processing" those inbox(es) down to empty, it is only reasonable to reward reaching "empty" at least once per day, rather than stressing about processing to empty at the end of the day when it may be difficult to do so (mental exhaustion related cognitive impairment), or when you may have higher priority tasks or projects to tend to, or both.
weekly checkpoint metrics
(have not started tracking yet. consider rescoping to 5 day weeks per NewCalendar before tracking)
- Were you able to log metrics for the previous 7 days? (added 2009-054)
- Were you able to complete a review sometime in the past 7 days?
- Perhaps start with a monthly review, or measure how often you are able to complete a weekly review, and then set a goal to do it slightly more frequently until you reach the rhythm of doing a weekly review.
monthly checkpoint metrics
(have not started tracking yet.)
- Were you able to log metrics for the previous 30 days? (added 2009-054)
- Were you able to complete a review sometime in the past 30 days?
summary of checkpoints metric
Without having a specific methodology for valuing any of the above checkpoints more than the others, if we assign each a 1 point value, then we can assign a discrete point value to each day. The more points being the better.
- number of yes checkpoints for each day
checkpoint insufficiency
The above metrics appear to be necessary but insufficient to measure actual productivity, as there is no measure of what is being produced.
See Matthew Levine's metrics for additional possible metrics.
Perhaps the metrics should be divided into two categories (and thus summaries)
- metrics of being "on system" (has upperbound of precisely following the system)
- perhaps call this: GTD checkpoint metric
- metrics of productivity output (has no upperbound)
- perhaps call this: production metric (measuring production, what is actually produced)
measurement requirement
Why? You must measure metrics in order to be at least somewhat scientific about determining progress.
When? Before processing/doing any tasks in a day, measure/document the metrics for the previous day.
others metrics
It may help to consider others' thoughts on productivity metrics as well in developing better metrics.
related external pages
references
Return to ProductivityImprovement \ FrontPage.
Comments (0)
You don't have permission to comment on this page.