Improve Market Value
dennyzhang
URL: https://quantcodedenny.com/posts/improve-market-value/
Working deeply in a specialized area like Meta or Google ML infra can easily make your expertise feel too narrow — you solve critical problems, but your broader market value may not fully reflect your capabilities. Often, the challenge isn’t lack of knowledge; it’s being too busy with daily work, leaving no time for high-impact but non-urgent improvements or blindspots.
This post is a living document to help tech leads organically increase market value, stay visible, and continuously improve.
1. Guiding Principles
- Think End-to-End and Prioritize Impact
- Keep a mental map of the full system — data flows, dependencies, upstream/downstream impacts, and business objectives.
- Ask daily: “How does this change affect other teams, user experience, and business metrics?”
- Focus on Important but Not Urgent Work
- Identify blindspots: recurring issues, cross-team dependencies, systemic technical debt.
- Block weekly time for reflection, documentation, and proactive improvement.
- Translate Work into Marketable Achievements
- Frame accomplishments in revenue, cost savings, reliability, or scale.
- Capture not just what was fixed, but why it mattered: “Reduced ad delivery latency by 15%, improving ad revenue predictability by X%.”
- Document, Share, and Learn Across Teams
- Maintain daily notes or internal posts capturing problem, options, tradeoffs, and decisions.
- Summarize patterns weekly to increase visibility and influence.
2. Good Practices
- Frame P0/SEV incidents with business impact
- Log root cause, mitigation, and downstream effect.
- Highlight prevented losses or restored capacity in marketable terms.
- Design for scale and reliability
- Ask: “If traffic/data grows 5–10x, will this design hold?”
- Document assumptions, constraints, and failure modes.
- Share reusable learnings
- Daily: note patterns, gotchas, scripts that saved time.
- Weekly: summarize in central doc or chat to benefit team.
- Ask strategic, end-to-end questions
- Before implementation, ask about downstream impact or hidden dependencies.
- Capture tradeoffs even if the decision is local.
- Allocate deliberate focus time
- Weekly: protect 1–2 hours for reflection, blindspot review, systemic improvements, or skill expansion.
3. Bad Practices
- Local optimization without context
- Fixing minor metrics without measuring system-wide or business impact.
- Mitigation: document downstream consequences.
- Not documenting decisions
- Knowledge gets siloed; team forgets key context.
- Mitigation: 3 bullets per decision — problem, options, chosen path, rationale.
- Avoiding ambiguous or cross-team problems
- Appears reactive or narrow.
- Mitigation: log ambiguous issues and track next steps.
- Over-reliance on internal tech
- Limits marketable skillset.
- Mitigation: note underlying principles, patterns, and reasoning for tech choices.
4. Daily Runbook Approach
- Daily Reflection: Capture what you learned, decisions made, tradeoffs, and why they matter.
- Weekly Focus Block: 1–2 hours for important but non-urgent work (blindspots, systemic improvements, cross-team learning).
- Weekly Market Mapping: Reframe work in industry terms; highlight transferable impact.
- Monthly Patterns Review: Aggregate recurring issues, lessons, systemic improvements.
- Quarterly LLM Update: Use an LLM to summarize, reframe, and extract marketable insights.
5. LLM Prompt Template for Continuous Improvement
Context:
I am a tech lead working in Meta Ads Infra.
I maintain a runbook of learnings, decisions, tradeoffs, and weekly focus work on blindspots and non-urgent improvements.
Task:
1. Summarize key patterns and decisions from the notes.
2. Highlight marketable achievements understandable outside my company.
3. Identify gaps in breadth, end-to-end thinking, or neglected blindspots that could increase market value.
4. Suggest improvements or principles to focus on next.
5. Output in org-mode format with actionable bullets suitable for daily practice.
Notes:
[PASTE YOUR RUNBOOK OR WEEKLY NOTES HERE]