Trying to quantify leadership or empathy or patience with a 1-10 score is the fastest way to ruin them. Tracking them qualitatively via evidence is different — and it works.
Evidence over scores
Instead of "leadership: 6/10," track moments. "This week I sat quietly in a meeting where I wanted to jump in and the junior engineer figured it out." Five moments like that per month is a skill developing.
Let the AI do the attribution
Tagging every journal entry with "leadership" is the fast way to quit. Auto-attribution by embedding similarity is the sustainable way — describe the skill once, the system finds relevant entries forever.
Weekly, not daily
Soft skills show up in bursts. Daily tracking produces 5 days of noise for every signal. Weekly review is the right cadence.
Goodhart's law applies
Goodhart's law — "when a measure becomes a target, it ceases to be a good measure." If you're tracking "how many times I practiced leadership this week," you're not practicing leadership. You're performing.
The coach as sparring partner
The real use of evidence tracking is to have the AI coach surface patterns you missed. "You wrote about difficult conversations four times this month — what's the pattern?" is the kind of observation a calendar-trigger reminder can't make.
About Sovereign — A privacy-first AI personal assistant that runs entirely on your iPhone. On-device LLM, zero-knowledge encryption, and a coach that learns from your own words. See how it works or visit the homepage.