You're improving as a PM when your decisions get better calibrated, your reflection turns specific instead of vague, you frame ambiguous problems faster, and the same category of mistake stops repeating — not when your title changes. Promotions and tenure are lagging, noisy signals; real growth lives in decision quality.

Quick Answer: Real PM growth shows up as better-calibrated decisions, deeper reflection, faster problem framing, and fewer repeated mistakes — not in your job title, team size, or years of tenure. Track those four leading indicators quarterly instead.

Most ambitious PMs ask "am I improving?" once a year, right before a review, when the honest answer is already too late to act on. The better version of the question is quarterly, self-administered, and answerable from evidence you already have lying around — old decisions, old journal entries, old postmortems. This piece walks through what to measure instead of title and tenure, why rereading your own past writing is the most honest check you have, and a lightweight scorecard you can actually keep running.

Why Promotions and Tenure Lie to You About Your Growth

Titles, scope, and years-in-role feel like proof of growth because they're visible, comparable, and show up on LinkedIn — but they're lagging indicators shaped as much by budget cycles, reorgs, and manager advocacy as by your actual judgment. A PM promoted after one lucky launch isn't necessarily sharper than a peer passed over twice during a hiring freeze.

This isn't a cynical take on your company; it's a measurement problem. Eric Ries popularized the idea of a "vanity metric" for products — a number that goes up and to the right but doesn't tell you anything actionable. Titles and tenure are the personal-growth equivalent: satisfying to watch climb, but they tell you nothing about whether you'd make a better call today than you did last year.

Three reasons these signals mislead you specifically:

  • They're gated by other people's decisions. A promotion requires a leveling cycle, a budget line, and a manager willing to advocate — none of which you fully control, and none of which measure judgment directly.
  • They reward visibility, not accuracy. Shipping something loud and defensible gets rewarded faster than making a quietly correct call to not build something.
  • They're slow. Tenure and title changes update annually at best. A signal that moves quarter to quarter has to be one you measure yourself.
Vanity SignalWhat It Actually MeasuresLeading Indicator to Watch Instead
Job title / levelLeveling-cycle timing, manager advocacy, budgetDecision calibration (confidence vs. outcome)
Team or scope sizeReorgs, attrition, executive sponsorshipTime to a workable problem framing
Years of tenureRetention, not skill acquisitionDepth and specificity of reflection
Launches shippedRoadmap velocity, not launch qualityRepeat-mistake rate across postmortems

Read the table as a substitution list: wherever you'd normally reach for a vanity metric to reassure yourself, there's a slower, noisier proxy on the left and a faster, more honest one on the right. The right column is what the rest of this article is about.

Leveling data aggregators like levels.fyi and compensation surveys such as Radford consistently show PM promotion cycles stretching anywhere from about a year to several years, varying more by company size and hiring freezes than by individual skill. That range alone should tell you a promotion is a lagging, noisy confirmation at best — not a real-time reading of your judgment.

The Four Leading Indicators That Actually Track PM Growth

Four signals reliably track whether your judgment as a PM is improving: decision calibration (does your confidence match your outcomes), reflection depth (how specific and self-critical your retros get), framing speed (how fast you turn ambiguity into a workable problem), and a falling repeat-mistake rate. Track these quarterly and the trend line beats any promotion cycle as evidence.

Decision Calibration

Calibration is the gap between how confident you were and how right you turned out to be. Philip Tetlock's forecasting research, later popularized in Superforecasting, found that most people's stated confidence barely predicts their actual accuracy — until they start explicitly recording predictions and scoring them against outcomes, at which point calibration measurably improves over repeated cycles.

The PM version of this: before a launch, write down what you expect and how sure you are — a rough percentage is enough. After the outcome lands, score yourself. A decision log that captures this loop, and forces you to revisit it later instead of rewriting history in hindsight, is one of the highest-leverage habits in the craft. We go deep on building one, and on the hindsight bias that quietly corrupts it, in our guide to decision logs and hindsight bias.

Growth here looks like your rough Brier score — or even an informal "was I roughly right" tally — tightening over quarters, with fewer wild swings between overconfidence and surprise.

Picture two versions of the same PM a year apart. Version one says "I'm 90% sure this pricing change lifts conversion" and it barely moves the needle — a confident miss with nothing recorded to learn from. Version two says "I'm 60% sure, and here's the one segment I expect to disagree," checks back after the launch, and updates the next prediction accordingly. The second version isn't smarter; they're just closing a loop the first version never opened.

Reflection Depth

Depth is about specificity, not volume. A shallow reflection says "the launch went okay, could've communicated better." A deep one names the exact assumption that was wrong, the moment it became visible, and what would have surfaced it two weeks earlier.

Amy Edmondson's research on team learning found that groups who treat setbacks as data to examine outperform groups who treat them as something to move past quickly. The same distinction applies to how you write about your own work. If your journal entries from a year ago and today read at the same level of vagueness, that's a growth signal too — just not the one you wanted.

Building a consistent capture habit is the subject of our piece on the friction-journal habit and product sense, and the broader discipline sits inside our complete guide to PM craft and reflection.

Framing Speed

Framing speed is how quickly you turn a vague stakeholder request into a workable problem statement. A junior PM handed "customers are churning" starts building dashboards. A PM who's grown reaches for a structured lens almost immediately — pulling apart the underlying jobs with a jobs-to-be-done analysis or mapping the emotional dips in a customer journey — and has a testable hypothesis within a day instead of a week.

Gary Klein's research on experienced decision-makers under time pressure — firefighters, ICU nurses — found that expertise shows up as faster, more accurate pattern recognition, not more raw analysis. The same holds for PMs: growth means recognizing "I've seen this shape before" sooner, not working harder once you recognize it.

A concrete test: time yourself, honestly, from the moment a vague request lands to the moment you have a one-paragraph problem statement you'd defend in front of a skeptical VP. If that gap has shrunk from a week to an afternoon over the last year, that's framing speed improving — regardless of what your job title says.

Fewer Repeated Mistakes

This is the simplest indicator to track and the easiest to fake yourself out on. The bar isn't "zero mistakes" — it's the same category of mistake not recurring. Shipping without a rollback plan twice in one quarter is a repeat; two unrelated launches failing for two unrelated reasons is just ordinary PM work.

Anders Ericsson's research on deliberate practice found that raw repetition plateaus quickly without a tight feedback loop that catches errors while they're still correctable — experience alone doesn't build skill, deliberately reviewed experience does. That's why a postmortem you never reread is close to worthless: the feedback loop only closes when you go back and check whether last quarter's lesson actually changed this quarter's behavior. Our guide to deliberate practice for PMs goes deeper on building that loop on purpose.

What Rereading Your Old Journal Entries Actually Reveals

Rereading journal entries from a year or two ago is the fastest way to see if you've actually grown, because it removes the self-serving narrative you'd otherwise tell about your own progress. Either the old entries look naive in a way you can name, or they read almost identically to today's — and that second outcome matters more.

Growth you can't see in your own old writing is growth you're asserting, not growth you're demonstrating.

Look for three specific things when you go back and reread:

  1. Vocabulary shift. Do you now name the actual mechanism — "we conflated activation with retention" — instead of the vague symptom, "engagement felt off"?
  2. Ownership shift. Does old writing externalize blame ("stakeholders kept changing scope") where recent writing locates your own contribution ("I didn't push back on scope with data soon enough")?
  3. Speed shift. Did the old entry take three paragraphs of throat-clearing to reach a problem statement that a recent entry reaches in one?

This only works if the old entries actually exist, are dated, and are comparable — a journal you abandoned after two weeks gives you nothing to reread. It's also why this works better as a structured habit than an occasional notes-app page.

If you're using something like Prodinja, whose Journal is designed to capture your real dated entries, including voice notes, alongside a growing Library of past situations, pulling up entries from a year ago next to this quarter's gives you a concrete before-and-after read on how your judgment has actually matured — not a self-report you're writing about yourself today.

A Quarterly Self-Scorecard You Can Actually Keep

A useful growth scorecard is short enough to finish in twenty minutes, scored the same way every quarter, and stored somewhere you'll actually reopen. Score each dimension 1-5, write one sentence of evidence per score, and compare the sheet to the one from three months earlier — the trend line matters more than any single quarter's number.

DimensionWhat You're ScoringEvidence to Cite
Decision calibrationDid confidence match outcome on your biggest calls?Your decision log entries and their actual results
Reflection depthAre recent entries more specific and self-critical than old ones?A side-by-side reread of two entries, a year apart
Framing speedHow fast did you reach a testable problem statement this quarter?Time from request to hypothesis on your two hardest asks
Repeat-mistake rateDid last quarter's lesson change this quarter's behavior?One named mistake you did not repeat
Stakeholder read accuracyDid you predict reactions and objections correctly before the room did?One meeting where your prediction beat your gut going in

Run it on a fixed cadence — the same week every quarter, not "whenever I remember" — or the comparison loses its power. A few rules keep it honest:

  • Score against evidence, not vibes. If you can't cite a specific entry or outcome, don't award the point yet.
  • Let a dimension go down. A quarter with a new team, a new domain, or a crisis often should score lower on framing speed — that's information, not failure.
  • Keep the scorecard next to your journal, so the comparison is one click away instead of a hunt through old files and folders.

Building the Review Into a Habit, Not a Performance Review

The scorecard only works if it stays a private diagnostic tool, not a dress rehearsal for a manager conversation — the moment it's written for an audience, you'll unconsciously round every score up. Keep it separate from your actual performance review, even if some of the same evidence ends up informing both.

Treat a low score as a targeting exercise, not a verdict. If framing speed is your weak dimension this quarter, next quarter's deliberate practice has a specific target instead of a vague resolution to "get better at ambiguity." That specificity is itself a sign the system is working — vague resolutions are a symptom of the vanity-metric mindset this whole exercise exists to replace.

Two failure modes are common enough to name directly:

  • Score inflation, where every dimension creeps toward 4-5 because you're comparing yourself to a bad month instead of your best quarter. Fix it by rereading last year's scorecard before you fill out this one, not after.
  • Abandonment after one bad quarter, where a rough scorecard becomes a reason to stop measuring instead of a reason to keep measuring. One weak quarter is a data point; a habit only becomes a trend line after four or five of them.

If you're starting from zero — no old journal, no decision log, nothing to reread yet — don't wait for a "clean" starting point. Write this quarter's entry today, score yourself honestly even with thin evidence, and treat the next four quarters as the baseline you wish you'd started a year ago. The cost of starting late is smaller than the cost of never starting.

Key Takeaways

  • Titles, scope, and tenure are lagging, political signals — they're gated by budget cycles and manager advocacy, not by the quality of your judgment.
  • Decision calibration — tracking your stated confidence against real outcomes in a decision log — is the most direct leading indicator of better judgment.
  • Reflection depth matters more than reflection frequency; a specific, self-critical entry beats ten vague ones.
  • Framing speed shows up as reaching for a structured lens, like JTBD or a customer journey map, within minutes instead of jumping straight to solutions.
  • A falling repeat-mistake rate, not a zero-mistake record, is the realistic bar — the goal is that last quarter's lesson visibly changes this quarter's behavior.
  • Rereading your own old journal entries is the most honest growth check available, because it bypasses the story you'd otherwise tell yourself.
  • A short quarterly scorecard, scored against cited evidence and kept separate from performance reviews, turns all of the above into a repeatable habit instead of a one-time insight.

Frequently Asked Questions

How do I know if I'm actually improving as a product manager?

You're improving if your decisions get better calibrated, your reflections get more specific and self-critical, you frame ambiguous problems faster, and you stop repeating the same category of mistake. Compare a decision log and journal entries from a year ago against this quarter's for concrete evidence either way.

What are leading indicators of PM growth, versus lagging ones?

Leading indicators are things you can measure this quarter that predict future performance — decision calibration, reflection depth, framing speed, and repeat-mistake rate. Lagging indicators, like title, promotion, tenure, and team size, only update annually and reflect politics and budget as much as skill.

How often should I run a PM self-assessment?

Quarterly is frequent enough to catch a trend and infrequent enough to avoid vanity fluctuations from a single good or bad month. Run it on a fixed calendar date rather than "whenever I remember," so each quarter's scorecard is genuinely comparable to the last one.

Does getting promoted mean I'm actually a better PM now?

Not necessarily — a promotion depends on budget, timing, and a manager's advocacy as much as on judgment, and it can lag or precede real skill growth by a year or more. Use it as a lagging confirmation at most, never as your primary evidence of improvement.

What should I actually write in a PM journal to track growth?

Write dated entries that name the specific assumption you held, the moment it was confirmed or broken, and what you'd do differently — not a general summary of what happened. Entries built for rereading a year later, not just for venting in the moment, are what make the comparison meaningful.