STRATEGYWeeks to result

Paired Key Results for Quality Control

Pair every quantity metric with a quality counterpart to prevent damage

Problem it solves

damage

Best for

Any team writing OKRs or setting performance metrics, especially those pursuing aggressive quantitative targets where quality or ethics could be sacrificed

Not ideal for

Purely qualitative work where quantitative pairing doesn't apply, or very early exploration phases where measurement is premature

Overview

Why this framework exists

The Paired Key Results framework is a safeguard against one-dimensional goal pursuit by requiring that every quantitative key result be paired with a qualitative counterpart. This 'measure both effect and counter-effect' principle prevents the well-documented dangers of pursuing numbers without guardrails.

The pattern is simple: when a key result focuses on output (quantity), its paired counterpart stresses quality. Three new features? Pair with fewer than five bugs per feature. $50 million in Q1 sales? Pair with $10 million in maintenance contracts. Ten sales calls? Pair with two new orders. The quantity metric drives volume; the quality metric ensures the volume is meaningful.

This framework draws directly from the cautionary tales of Ford's Pinto (aggressive cost and weight targets with no safety counterpart) and Wells Fargo (ruthless sales targets with no ethical guardrails). In both cases, organizations achieved their quantitative goals spectacularly while destroying value, lives, and reputations. The pairing discipline prevents organizations from optimizing their way into disaster.

Core principles

4 total
  1. Every quantity metric has a shadow cost that must be measured alongside it.
  2. The more ambitious the OKR, the greater the risk of overlooking a vital quality criterion.
  3. What you don't measure, you don't manage -- and what you leave unmeasured can destroy you.
  4. Quality counterparts transform reckless pursuit of numbers into disciplined pursuit of value.

Steps

4 steps
  1. Identify All Quantitative Key Results
    Review your OKRs and flag every key result that is primarily a quantity or volume metric: units shipped, revenue targets, users acquired, features delivered, calls made, etc.
  2. Ask the Shadow Cost Question
    For each quantitative key result, ask: What could go wrong if we hit this number by taking shortcuts? What would be sacrificed? What would an adversarial observer worry about? This reveals the quality dimension that needs pairing.
    Pro tipThink about the Ford Pinto test: if a team achieved its quantity targets perfectly but something terrible happened as a result, what would that terrible thing be? That's your quality metric.
  3. Craft the Quality Counterpart
    Write a specific, measurable quality key result that guards against the shadow cost. Features shipped pairs with bugs per feature. Revenue pairs with customer retention or satisfaction. Calls made pairs with conversion rate or orders closed.
    Pro tipQuality counterparts should be equally specific and measurable. Vague quality goals like 'maintain high standards' provide no real guardrail.
  4. Treat Both Metrics as Equally Important
    Communicate that achieving the quantity metric while failing the quality metric is not success. Both must be tracked with equal rigor in check-ins and scoring. If one fails, the paired OKR as a whole is incomplete.
    Pro tipIn OKR reviews, present quantity and quality pairs together on the same screen or report. Never report one without the other.
    WarningIf quality counterparts are treated as secondary or optional, they won't serve as guardrails. They must be equally weighted.

Checklist

Saved in your browser

Examples

2 cases
Ford Pinto: The Cost of Missing Quality Counterparts

Ford's Pinto was developed under aggressive metrics -- 'under 2,000 pounds and under $2,000' -- enforced with an iron hand. When a crash test found that a one-dollar plastic piece could prevent gas tank punctures, it was rejected as extra cost and weight. The product objectives covered size, weight, cost, and performance. Safety was nowhere on the list.

OutcomeHundreds died in rear-end collisions, thousands were severely injured. Ford recalled 1.5 million vehicles. A simple quality counterpart on safety could have prevented the catastrophe.
Intel Manufacturing Quality Pairs

At Intel, Andy Grove insisted that production volume metrics always be paired with quality counterparts. Vouchers processed in accounts payable were paired with error rates found by auditing. Square feet cleaned by custodial groups were paired with quality ratings from senior managers. This discipline was embedded in the semiconductor manufacturing culture where yields depend on precision.

OutcomeIntel maintained world-class manufacturing quality while scaling production, establishing the discipline that allowed them to fabricate millions of chips with a millionth-of-a-meter accuracy.

Common mistakes

3 traps
Writing Quality Counterparts That Are Vague and Unmeasurable
A quality key result like 'maintain quality standards' provides no real protection. It must be specific and measurable: 'fewer than 5 bugs per feature' or 'customer satisfaction score above 4.5/5.0.' If you can't grade it objectively, it's not a real guardrail.
Adding Quality Counterparts as an Afterthought
If quality metrics are bolted on after quantity goals are already set and resourced, they'll always be underfunded and deprioritized. Design both quantity and quality key results simultaneously as equal components of the same objective.
Pairing Only Financial Metrics
Quality counterparts aren't just for revenue goals. Feature velocity needs code quality counterparts. Hiring speed needs candidate experience counterparts. Marketing reach needs lead quality counterparts. Apply the principle across all domains.

Origin story

How this framework came to be

Andy Grove formulated this principle in High Output Management, drawing from Intel's manufacturing discipline where quality and yield metrics always accompanied production volume targets. He observed that in accounts payable, the number of vouchers processed should be paired with error rates found by auditing. John Doerr elevated this to a standalone framework after analyzing the Ford Pinto disaster and the Wells Fargo scandal, both of which demonstrated the catastrophic consequences of one-dimensional metrics pursued without quality counterparts.

Source

Traced to primary
Source · BOOK
Measure What Matters
John Doerr · 2018
Open source →

Related frameworks

Browse all Strategy →