Professional Growth You Can See in the Work

“Be more strategic” can survive for years in a performance review.

It sounds important. It also gives the person, their manager, and the next reviewer no shared way to tell whether anything changed.

A course can be completed. A stretch project can be survived. A new title can be granted. None of those events shows that someone now handles the work differently.

Professional growth becomes visible when a person changes how they handle a recurring decision or repeatable piece of work—and can reproduce that change.

That definition is deliberately narrow. It moves development out of aspiration and into work that can be inspected, challenged, and tried again.

The unit of development is not a career level. It is a bounded capability experiment.

Turn the aspiration into a change in behaviour

Start with language that appears in review notes: become more strategic, improve stakeholder management, show stronger judgement, or develop executive presence.

Do not debate the label yet. Ask what someone doing this well would do differently in a real situation.

Define five things:

• the recurring situation in which the capability matters;
• the decision or action the person is responsible for;
• the behaviour visible today;
• the behaviour worth practising;
• the work product in which the difference should appear.

“Improve strategic thinking” is not testable.

“When proposing an investment, compare credible alternatives, expose the decisive assumptions, and recommend one trade-off” is.

The second version names a situation, an action, and evidence. It can be attempted without pretending that “strategy” is one indivisible trait.

Keep authority in view. A product manager may improve the quality of an investment recommendation without owning the investment decision.

Development should not smuggle a larger role into an informal exercise.

Establish a baseline from recent work

Self-assessment is useful for choosing where to look. It is weak evidence of what already happens.

Choose a recent artefact from the recurring situation: a decision memo, discovery readout, prioritisation proposal, experiment review, or stakeholder update.

Inspect the artefact and the decision around it.

• Which alternatives were considered?
• Which evidence changed the recommendation?
• Which assumption carried the most risk?
• Which trade-off remained hidden?
• What did the audience need but not receive?
• Where did the person need rescue or escalation?

The baseline is not a verdict on the person. It is a trace of one performance in a named context.

Record contextual limits. A rushed memo during an incident should not become a permanent diagnosis of someone’s judgement.

Use the baseline to select one behaviour, not to assemble a catalogue of faults.

Practise one component, not the whole identity

Normal work creates experience. It does not automatically isolate the part of performance that needs to improve.

The 1993 Ericsson, Krampe, and Tesch-Römer paper distinguished purposeful, effortful practice from simply performing a familiar activity.

That distinction is useful. It suggests designing a repeatable attempt around one component and obtaining information that can shape the next attempt.

It does not justify importing the popular “10,000-hour rule” into product work.

The original paper combined a theoretical model with small, selected samples of musicians. Much of its practice history was retrospective, and the relationships were correlational.

A later meta-analysis by Macnamara, Hambrick, and Oswald reinforces the need for restraint.

After a 2018 correction, its average correlation across 88 studies was .38. Deliberate practice accounted for 14% of performance variance on average, with high heterogeneity at I² = 88.54.

The corrected domain estimates ranged from 24% in games to 1% in professions. Explained variance is an association in these studies, not a causal share of professional performance.

Practice may matter. It is not a complete theory of capability, and hours are not the unit that matters here.

For a product practitioner, a useful practice target might be:

• writing two viable alternatives before recommending one;
• separating an observation from an inference in a research readout;
• naming the decision that a metric is meant to inform;
• stating the cost of a stakeholder exception;
• identifying what evidence would reverse a product bet.

Choose one. Rehearse it in work that genuinely recurs.

Find a challenge with stakes and a boundary

A development exercise needs enough consequence to expose judgement. It does not need maximum risk.

DeRue and Wellman’s 2009 field study examined 225 work experiences nested within data from 60 managers.

They found diminishing returns between developmental challenge and leadership skill development. Access to feedback moderated that pattern at high levels of challenge.

This was an observational leadership-development study, not a product-management experiment. It does not show that assigning a stretch project causes growth.

It does challenge the idea that more difficulty is always better.

Design the practice opportunity with four boundaries:

1. Real stake: the work informs an actual decision or commitment.
2. Reversible scope: failure can be corrected without exposing customers or the business to avoidable harm.
3. Support: a named person can observe, question, or intervene.
4. Escalation point: conditions requiring a pause or a more experienced decision-maker are explicit.

The right challenge is not the largest available project. It is the smallest real situation that makes the target behaviour necessary.

If the capability concerns regulated claims, pricing, security, employment, or irreversible commitments, narrow the exercise or add qualified oversight.

Development is not permission to spend organisational risk for personal learning.

Ask for feedback on the work

“Any feedback?” invites reassurance, taste, or a list of everything the reviewer noticed.

Ask the reviewer to inspect the chosen behaviour.

• Which alternative did I dismiss without enough examination?
• Where did the recommendation outrun the evidence?
• Which trade-off would you make explicit before deciding?
• What would make this useful to the decision-maker?
• At what point should I have escalated?

Name the feedback owner before the attempt. Choose someone who can observe the work and understands the decision context.

A manager may be right. A peer, engineer, researcher, commercial lead, or customer-facing colleague may see a more relevant part of the behaviour.

When the gap needs sustained outside perspective, a bounded mentoring relationship may help.

When the work is helping another person think rather than supplying the answer, use coaching practices.

Feedback is evidence from a perspective. It is not an objective score and should not be accepted without context.

Look for specific observations, competing interpretations, and a change to try next time.

Separate experience from learning

A full calendar proves exposure, not learning.

After the attempt, reconstruct what happened while the decision and its context are still available.

Use four prompts:

1. What was I trying to do?
2. What actually happened?
3. Where did my reading of the situation fail?
4. What will I change in the next attempt?

Di Stefano, Gino, Pisano, and Staats tested short written reflection against additional practice in an HBS working paper.

In its field study, 101 Wipro agents arrived in batches of 10–25. Each batch, rather than each individual, was assigned to reflection or practice.

Fifty-six agents spent 15 minutes reflecting on each of ten days; 45 used the same time for additional practice.

The paper also reports a randomised online task with 453 participants and a third study in which 256 participants chose reflection or more practice.

Reflection produced better near-term task results in those settings. In the third study, however, participants selected their condition rather than being randomised.

The evidence has narrow boundaries: call-centre training and short online tasks. The advantage on the field study’s highest customer-rating category disappeared after the first month.

The paper is a draft for discussion, not a settled rule that reflection beats practice in product management.

The practical case is smaller. A short reconstruction can stop repeated experience from passing without inspection.

Do not turn reflection into a polished success story. Preserve the mistaken assumption, the contrary signal, and the uncertainty that remains.

Test whether the change transfers

One improved memo may reflect the topic, the reviewer, or unusually generous preparation time.

Repeat the same target behaviour under a meaningful change in context.

Change one dimension:

• a different investment decision;
• a different stakeholder group;
• a different product area;
• weaker or more ambiguous evidence;
• less help from the original reviewer.

The transfer test asks whether the person can recognise when the behaviour applies and reproduce it without the original scaffolding.

Do not demand identical execution. The point is to preserve the underlying judgement while adapting to the situation.

If the behaviour disappears, the experiment still produced information. The capability may be context-bound, the prompt may be doing the work, or the target may be too broad.

Keep a minimal evidence trail

Retain enough evidence to compare attempts without creating a shadow performance-management system.

Keep:

• the baseline artefact;
• the revised or later artefact;
• the feedback source and observation;
• the person’s reflection;
• the transfer context;
• the limits of any conclusion.

Do not convert this record into a claim of mastery.

Teaching a concept, receiving praise, or presenting a strong artefact can support a judgement. None proves that the capability is reliable across situations.

This evidence trail is for choosing the next development move. It is not yet a career portfolio or a personal-brand asset.

For a move into PM, Build Evidence Before the Title shows how to expose what transfers without passing practice off as experience.

Decide what happens after the experiment

End with a disposition, not a ceremonial “completed.”

• Repeat: the behaviour changed but is not yet reliable.
• Narrow: the capability was too broad to practise coherently.
• Increase difficulty: the behaviour transfers and needs a harder context.
• Add support: the person needs closer feedback, instruction, or modelling.
• Change the environment: the work does not provide a fair practice opportunity.
• Stop: the capability is not material enough to justify further effort now.

A mentor is not automatically the answer. Neither is another course, a public commitment, or a larger project.

Choose the intervention that matches the observed constraint.

If the experiment concerns a move into people leadership, leadership readiness is a separate question with different work and risks.

A fictional capability experiment

Consider an explicitly fictional senior product manager whose review says, “be more strategic.”

Their recent investment memos describe one preferred initiative in detail. Alternatives appear as short objections, and assumptions are mixed with facts.

The development target becomes specific: for three real investment decisions, present at least two viable alternatives, their decisive assumptions, their trade-offs, and one recommendation.

The baseline is the most recent memo. The practice behaviour is writing the alternatives before building the case for a preferred option.

The feedback owner is a product director who will challenge whether the alternatives are genuinely viable. A finance partner will inspect the economic assumptions.

The risk boundary is clear. The product manager owns the recommendation, not the final investment decision. Legal, security, and contractual commitments remain outside the exercise.

After each decision, the product manager records which assumption they misread and what changed in the discussion.

The transfer test uses a decision in another product area without the director helping to frame the alternatives.

No promotion, commercial result, or improved decision outcome is claimed. The example shows only how vague feedback can become observable practice.

Use a capability experiment card

Capability
Recurring decision
Baseline artefact
One behaviour to practise
Practice opportunity
Feedback owner
Risk boundary
Reflection
Transfer test
Review condition

Keep the card close to the work. If it needs a scoring model, a competency library, and a quarterly programme to function, the experiment is probably too large.

Professional growth should leave a trace in the next decision, not only in a development plan.

Sources

Ericsson, Krampe, and Tesch-Römer, 1993

The Role of Deliberate Practice in the Acquisition of Expert Performance is a peer-reviewed Psychological Review article.

Study 1 examined ten violin students nominated as the “best,” ten matched “good” violin students, and ten matched music-education students the paper called “music teachers.”

Ten middle-aged professional violinists also supplied developmental-history data.

Methods included interviews, activity ratings, one-week diaries, retrospective practice estimates, and indicators of musical performance.

Study 2 compared 12 young expert pianists with 12 young amateur pianists through diaries, retrospective estimates, and musical and non-musical performance tasks.

Early-practice estimates were also collected from 12 older experts and 12 older amateurs.

The samples were small and highly selected. Much of the practice history was retrospective, and the relationships were correlational. The study does not establish a 10,000-hour rule or direct transfer to product work.

DeRue and Wellman, 2009

Developing Leaders via Experience is a peer-reviewed Journal of Applied Psychology article.

It analysed 225 on-the-job experiences nested within data from 60 managers. The challenge relationship showed diminishing returns, while feedback availability moderated high challenge.

It was an observational leadership-development study with a small manager sample. It does not show that challenge or feedback causes professional growth in product teams.

Di Stefano, Gino, Pisano, and Staats, Working Paper 14-093

Making Experience Count is an HBS working paper distributed for comment and discussion.

Its studies included 101 Wipro agents, 453 randomly assigned online participants, and 256 online participants who chose between reflection and practice.

The Wipro intervention lasted 15 minutes per day for ten days. Results concern short-term performance in call-centre training and bounded online tasks.

The top-category customer-rating difference disappeared after one month. Study 3 is vulnerable to self-selection, and the working-paper status limits claims of settled evidence.

Macnamara, Hambrick, and Oswald, 2014 and 2018

The 2014 meta-analysis synthesised 88 studies across games, music, sport, education, and professions.

The 2018 correction revised the average correlation to .38 and overall explained variance to 14%.

Corrected domain estimates were 24% for games, 23% for music, 20% for sport, 5% for education, and 1% for professions. Heterogeneity was high at I² = 88.54.

The professions estimate was not statistically significant (r = .09, p = .377).

Definitions and measures varied, and much of the underlying evidence was observational. Explained variance must not be read as the proportion of performance caused by deliberate practice.

Read next

• Mentorship Programs: Build the Learning Relationship

• Building Strong Teams with Coaching Practices

• How to Develop Leadership Readiness

• PM Career Paths: Choose the Work Before the Ladder

• A Product Manager’s Evidence Portfolio: Record Judgement Before Memory Edits It