> Measure productivity by all means—I’m all for accountability—ideally as tangible business impact… proxy business metrics are fine too… Just don’t try to measure the individual contribution of a unit in a complex adaptive system, because the premise of the question is flawed.
Great article, but I disagree with this conclusion and don’t think it reflects its own reasoning.
If proxy metrics are acceptable, then measure by proxy the effort and impact of the activity. You are using stories and have story points. You should also have increasing velocity and capability metrics to show the levelling up from mentoring or happier pairing versus the hourly salaries involved and opportunity costs management gets to evaluate… Put those mentoring stories in and point them, or proportion those points across paired members, and measure them over time. The analytics will show things.
I find unwillingness to subject engineering activity to metrics highly suspect, and often based of flawed business application of business accounting or process monitoring. Proxy metrics are used for this exact reason, but to arbitrarily stop quantifying and assert subjective opinion as a conclusion is literal vibe management. “It feels like X, and we measure only 1/2 of Q, therefore X… which coincidentally was our pre-conclusion, our pre-judice, and preferred-conclusion.”
Here we have a teammate pairing for 100% of a senior developer position? … counter assertions we could make into the total lack of data about that: the juniors were stifled and would progress far faster self-directed, the seniors were annoyed by habits they can’t talk about politically with higher turnover and lower productivity, the manager is normalizing just-show do-nothing non-accounted jobs which is a corruption gravy train, there’s an entitled Socratic unicorn getting salary refusing to account for his time, and comparable teams are seeing more effective mentorship and aggregate team output improvement per senior-pairing hour (we think … but could never actually compare, since one party isn’t accounting).
Now I bet author is fully right and this super dev was an accelerant and used well. I disagree that this can’t be measured and feelings match a proper accounting. If the author was being misled by a bullshit artist you have a 0-point 100% salary that should be visible from orbit being granted based on personal favours. Not a great look for anyone involved. Not something I’d pay for.
15
2