Tracking the rate of perceived exertion (RPE) of our individuals and teams can help us to make better decisions about changes in velocity and act as a warning system of the deteriorating health of a team.
RP what now?
RPE originated in endurance training and was developed by Swedish psychologist Gunnar Borg. The original scales asks you to rate your perception of the effort a physical task took on a scale from 6-20 which could then be multiplied by ten to give a fairly accurate estimation of the heart rate that you would have experienced when doing the task
Mike Tuchscherer adapted this idea for use in powerlifting but converted it to a less unwieldy 1-10 scale. He (and subsequently large parts of the community) use it in programming to make sure that lifters work at the appropriate intensity needed to drive progress while taking into account stress that can affect performance but comes from factors outside of the gym. Poor sleep, nutrition or stress from other major life events as well as natural fluctuations in strength mean that on a given day you may feel the exact same measure (weight on the bar in this case) to require very different levels of effort.
|8||Moderately difficult. Could have done at most two more reps.|
|9||Could maybe have done one more rep. Really difficult.|
|10||Bone on bone grinder. Could have done no more reps or failed the set.|
RPE as it applies to weight training
The use of RPE has expanded into other areas of fitness (like cardio-vascular training) and can also be used to track the effort needed to complete training sessions in team or individual sports.
|6||More boring than difficult.|
|7||You can talk in short sentences but you can’t sing|
|8||Feels difficult. Can’t talk. Hard to sustain for longer than a few minutes|
|9||Really intense. Hard to sustain longer than 20-30 seconds|
|10||All out sprint. Unsustainable for longer than a few seconds|
RPE as it applies to cardio-vascular training
It can also be useful to more accurately track other metrics. Lifting a set at 100kg in one session and 110kg in the next is not necessarily indicative of an increase in strength if we don’t take into account the exertion needed to complete the reps. If session one was 5 reps at RPE 6 and session two was 5 reps at RPE 8 then all we can say is that the second session was more stressful.
Initial stab at an RPE scale for knowledge work
Imagine you are at the end of the workday and choose the sentence that best describes how hard it felt.
|1||Leave day: I didn’t have to work at all 🙂|
|…||I’m skipping 2-4 as they don’t represent levels of effort that are reasonable expectations at work.|
|5||I completed the work for the day and I had to go looking for more to do.|
|6||I was able to complete the day and direct some of my energy toward longer-terms pursuits (getting better, learning)|
|7||I’ve had a great productive day and I’m definitely ready to take some time off this evening.|
|8||I had to prioritise the use of my energy (am I going to meet up with friends or get my workout in?) in order to complete the work.|
|9||It took everything I had to complete the work for today. I didn’t have energy for anything else.|
|10||All I had energy for today was work and I still couldn’t get through enough.|
A proposed RPE scale for work.
The goal would be that teams and individuals stay for the most part in the 5-7 range. Occasional dips into higher intensities are not necessarily problematic but they are unlikely to be sustainable for long periods of time.
Team health metric
In theory this should allow us to see if we can make real and sustainable improvements in velocity. For example if we increase our velocity but the teams average RPE climbs out of ideal ranges then we haven’t gotten faster, we are just putting in more effort. On the other hand if our velocity holds constant but our RPE is dropping that may mean we are able to push a little harder but remain well within our ability to recover.
I’m unsure how frequently this metric would need to be checked to be useful. Too infrequently and it could easily be too influenced by recency bias but you don’t want to put to much tracking burden if it’s ultimately not helpful. The two most obvious approaches would be to track once per sprint at the sprint retro or to track daily (perhaps at standup) and calculate the average for the sprint.
I’m focused on software development but I don’t see any barriers to this being applicable more broadly to knowledge workers as a whole.
It may also be possible to get to the point where it is useful prescriptively. For example in planning you could keep it in mind that you are aiming for a sprint at RPE 6.
So far, all of this is just an interesting idea. My next step is going to be tracking this for my team on my next project and using the results to plan a next step.