Companies do NOT know how to measure productivity - So play the game and win! - No AI

Core claim: “Productivity” is not measurable in most corporate knowledge-work roles

• In white-collar/corporate environments, organizations cannot directly measure true productivity.
• Instead, they measure proxies such as:
  • hours and activity levels
  • tool activity
  • email/ticket volume
• This leads to work becoming a “game”: employees optimize what gets counted rather than what creates profitability.

Business lens: what companies actually optimize

• The primary organizational objective is profitability (the only thing a corporation ultimately cares about).
• Productivity is treated as a proxy for profitability—not the real driver.
• As a result, performance systems optimize for what can be observed, not what is most economically valuable.

Framework / playbook implied: “Proxy-filling” strategy

The actionable “how to win” embedded in the talk is: identify which proxy metrics your company uses and satisfy them.

1. Determine the proxy for productivity your manager/company tracks
   • Examples mentioned: hours/activity, reporting frequency/timeliness, monitoring/keyboard activity, lines-per-code, “shippable output.”
2. Optimize behavior to “fill the proxy”
   • Even if real value is created faster or elsewhere.
3. Avoid outcomes that trigger quota/targets ratcheting or resource cuts
4. If automation is discovered in your role, expect retaliation
   • via process changes, training requests, or headcount reduction

Concrete examples (and what they illustrate)

1) Sales quotas: the “103–105%” game to maximize payout without ratcheting targets

Sales is treated as closer to profitability because more sales → more revenue → (often) more profit.

But compensation systems create perverse incentives:

• If you miss quota: job risk escalates after a few quarters/months.
• If you exceed quota slightly (e.g., 103%–105%): you likely get a bonus/kicker.
• If you exceed too much (e.g., 150%–200%): next period’s quota increases, reducing future upside.

Google example (as referenced):

• Policy described: around 107%+ triggers quota increase next quarter.
• 102–105% triggers compensation with a higher share of total sales—while avoiding the quota ratchet.

Actionable takeaway: don’t maximize for the company; maximize for comp mechanics—target the band where you get the bonus without triggering higher quotas.

---

2) Finance automation (Sally vs. Barbara): proxies punish real efficiency

• Sally (high activity):

  • Works long hours (e.g., 7:00am–6:00pm).
  • A packed calendar and visible busyness are interpreted as “hard work.”
  • Reports/projects may be late, but still perceived as valuable.

• Barbara (high output via automation):

  • Uses BI/automation to complete ~95% of her work quickly.
  • May need only 15 minutes to 1 hour/day of real effort.
  • Produces better deliverables (presentations/reports generated automatically).

Even though Barbara is more effective:

• Managers measure activity proxy, not productivity.
• If automation is discovered, manager reaction escalates:
  • Barbara may be asked to train the department for ~3 months
  • Layoff risk: automation reduces required headcount (speaker suggests 1 out of 20 roles might become unnecessary if 95% is automated)
• The political/protection angle: Barbara’s competence can threaten others, so she is treated as a problem.

Actionable takeaway for knowledge work: if you automate, be prepared to either (a) conceal true efficiency by “looking busy” to satisfy activity metrics, or (b) work in environments whose incentives reward results.

---

3) Software engineering quality vs. output metrics (Bob vs. Frank): short-term metrics beat long-term value

• Bob (fast, sloppy):

  • Ships quickly (e.g., “few hours”), but code is buggy/hard to maintain.
  • Issues appear later (testing/users downstream).

• Frank (slow, robust):

  • Takes longer (e.g., “two or three days”), but produces clean, maintainable, bug-free code.

However, the organization is said to favor proxy measures such as:

• lines of code per hour
• simplistic throughput metrics like “shippable feature by person-month/person-week”

Result:

• Bob looks more productive short-term and gets promotion/raises.
• Frank looks worse, even if long-term maintainability is better for the company.

Actionable takeaway: when evaluation metrics are misaligned, teams can systematically select for short-term throughput over long-term quality.

Organizational/leadership tactics discussed (incentives design problems)

Compensation design: “Reward for A hoping for B” failure mode

• The speaker argues motivation/compensation is hard because organizations must reward behavior that creates outcomes.
• The failure pattern:
  • organizations compensate for A (quota, activity)
  • while hoping for B (better profitability/productivity)
• What gets rewarded is what happens, which often diverges from what leadership truly wants.

Sales managers’ strategic resistance to quota increases

The speaker claims top sales leaders push aggressively during quota-setting not because they can’t hit numbers, but because they want lower quotas so that smaller overachievement (e.g., 103–105%) still generates bonuses.

Metrics / KPI-like targets explicitly mentioned

These are not presented as formal company KPIs, but as cited thresholds and the proxy logic used:

• Sales quota banding / thresholds (examples):

  • 103%–105%: “good” zone for personal bonus/kicker
  • 107%+ (Google described): triggers quota increase next quarter
  • higher overachievement (e.g., 150%–200%): escalates quotas and reduces future upside

• Automation level (finance example):

  • ~95% of work automated
  • ~15 minutes to 1 hour/day of real effort

• Automation-driven headcount implication (finance example):

  • if 95% is automated, speaker estimates roughly 1 out of 20 roles may be needed

• Time/activity proxies:

  • “packed calendar,” “working from 7am to 6pm,” and “monitoring” are described as the mechanism for activity-based evaluation

Practical recommendations implied by the speaker

• If you work under proxy-based evaluation:
  • make sure your observed activity matches what managers track (i.e., “fill the proxies / check the boxes”).
• If you want to win long-term in a healthy culture:
  • seek or promote teams where evaluation aligns to outcomes/results rather than activity.
• During compensation/quota design:
  • be cautious of reward systems that create quota ratcheting or incentivize “stalling” to stay inside the bonus band.
• If you automate work:
  • expect political/operational consequences (training requests, headcount reductions) and plan accordingly.

Presenter / sources

• Presenter: The unnamed speaker in the subtitles (the video appears to be commentary/personal experience; no additional external sources are cited).