Experience Curve Review¶
Diagnostic review — instantiates Compounding Advantage Flywheel Design
Certifies whether cost or quality is genuinely improving through learning-by-doing as cumulative production grows — and captures the lessons that drive it — separating a real experience effect from ordinary scale or price moves.
An Experience Curve Review certifies a cause: that unit cost or defect rate is falling because the organization is learning by doing as cumulative output accumulates, and not merely because volume is up or input prices dropped. Its defining move is to test the learning-by-doing driver against cumulative production — does cost fall a consistent fraction each time total units produced doubles? — and, crucially, to harvest and store the lessons that produce the fall, so the experience becomes a retained asset rather than tacit know-how that walks out the door. It is not a reading of the outcome curve's shape; it is an audit of whether learning is the real engine and a mechanism for capturing what was learned.
Example¶
A lithium-battery-cell manufacturer sees its cost per kilowatt-hour dropping steadily and wants to know whether that is a durable learning advantage it should press, or just cheap lithium and bigger plants that a competitor could match tomorrow. The Experience Curve Review plots cost against cumulative cells produced (not calendar time or annual volume) and finds a stable slope: each doubling of lifetime output cuts unit cost by a consistent percentage — the signature of a genuine experience curve.[1] Then it decomposes the fall: how much came from input prices, how much from plant scale, and how much from process learning — yield improvements, faster line changeovers, operator technique. The learning share is large and specific.
The review's second half is the repository. It debriefs the line engineers and files what actually drove each yield jump — the electrode-coating tweak, the humidity-control change — into a searchable process-learning store, tagged to the step it improved. When the firm builds its next plant, that captured experience transfers instead of being re-learned from scratch. The output is a certification plus an asset: "roughly two-thirds of the cost fall is real learning-by-doing, the driver is durable, and here is the documented know-how that produced it."
How it works¶
- Plot cost/quality against cumulative output. Use lifetime units produced as the axis, because the experience effect is a function of accumulated doing, not of a calendar or a single year's volume.
- Decompose the improvement. Attribute the fall across input prices, scale economies, and process learning, so the learning-driven part is isolated from effects a rival could buy.
- Test the slope for stability. A real experience curve shows a roughly constant fractional improvement per doubling; a wandering slope suggests the driver is something other than learning.
- Capture the lessons. Debrief the people doing the work and file the specific changes that produced each improvement into a structured repository, tagged so the know-how transfers to the next unit, line, or site.
Tuning parameters¶
- Attribution rigor — how hard you work to separate learning from price and scale. More rigor yields a trustworthy certification but costs analyst time and clean data.
- Learning-unit definition — what "one unit of experience" is (cells, batches, shifts). Finer units resolve the curve better but need more granular records.
- Capture depth — how much of each lesson you document, from a one-line note to a full process spec. Deeper capture transfers better but burdens the people doing the work.
- Transfer scope — whether captured learning stays with the team, the plant, or the whole firm. Wider sharing compounds faster but raises confidentiality and coordination cost.
- Refresh cadence — how often the review re-fits the slope and re-harvests lessons as output accumulates.
When it helps, and when it misleads¶
Its strength is that it stops a firm from mistaking a temporary tailwind for a durable advantage — and, by capturing the lessons, converts fragile tacit experience into a transferable stock so the learning is not lost to turnover or forgotten between projects.
Its failure mode is over-attribution: teams love to credit their own learning for a cost fall that was really cheaper inputs or a bigger plant, and an experience curve fitted loosely enough will "confirm" a driver that isn't there. Captured lessons also go stale — a documented technique for last year's process can mislead on this year's. The classic misuse is treating the experience-curve slope as a forecast — assuming cost will keep falling at the historical rate forever — when learning saturates and the curve eventually flattens. The discipline that guards against this is to decompose honestly before crediting learning, date and prune the lesson repository, and treat the certified driver as currently real rather than permanently guaranteed.
How it implements the components¶
increasing_return_driver— it certifies the specific driver learning-by-doing: it tests, against cumulative output, that process learning (not price or scale) is what makes the next unit cheaper or better.learning_capture_repository— the debrief-and-file step is this repository: it converts the tacit lessons behind each improvement into a structured, transferable store tagged to the process it improves.
It does not measure the shape of the return curve or flag its saturation (marginal_return_curve, saturation_and_reversal_monitor) — that is its nearest twin, Compounding Curve Review; this review certifies the learning cause and captures its lessons, while that one reads the outcome curve.
Related¶
- Instantiates: Compounding Advantage Flywheel Design — it certifies the increasing-return driver and banks the experience the loop runs on.
- Sibling mechanisms: Compounding Curve Review · Bubble and Lock-In Red Team · Cumulative Reputation System · Data Flywheel Dashboard · Fixed-Cost Amortization Plan · Open Standard or Portability Rule · Platform Seeding Program · Reinvestment Cadence
Editorial Notes¶
Form Classification¶
Form family: Assessment, Review & Assurance
Rationale: Experience Curve Review operates as a bounded evaluation of existing evidence or work that produces a finding or disposition because it certifies whether cost or quality is genuinely improving through learning-by-doing as cumulative production grows — and captures the lessons that drive it — separating a real experience effect from ordinary scale or price moves.
Independent corroboration: The frozen evidence defines Experience Curve Review as 'Certifies whether cost or quality is genuinely improving through learning-by-doing as cumulative production grows — and captures the lessons that drive it — separating a real experience effect from ordinary scale or price moves', so its operative form is Assessment, Review & Assurance.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Organizational & Management Science
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Multi-domain
Rationale: Auditing whether scale improvements reflect organizational learning-by-doing is rooted in operations and strategy's experience-curve tradition.
Related originating lineages:
- Aviation & Aeronautics — Aircraft manufacturing supplied the original empirical learning-curve lineage.
- Economics & Finance — Economic production theory materially distinguishes learning effects from price and scale effects. Econometric learning-by-doing analysis materially supplies controls for scale, input prices, and selection effects.
Review resolution: Both reviewers agree that organizational_management is primary. I retain aviation_aeronautics, economics_finance only as formative origin lineages; cross_disciplinary_synthesis is appropriate because the final form materially combines the agreed primary with the retained formative lineages. Reach is multi_domain because the structure transfers across several fields but is not a near-universal human pattern, an applicability judgment kept separate from provenance. Encyclopedia synthesis is true because the exact generalized artifact is an encyclopedia-authored combination or refinement. No unresolved historical ambiguity remains after reconciling the secondary fields.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Reconciled after independent review; high confidence.
References¶
[1] The experience curve / Wright's Law — Theodore Wright observed in 1936 that aircraft labor cost fell a roughly constant fraction each time cumulative production doubled, a regularity the Boston Consulting Group later generalized across industries. Plotting against cumulative output rather than time is what distinguishes a genuine learning effect from ordinary scale or price movement. withdrawn registry ↩