Two builders, both a decade in. On paper they are identical: same years, same titles, same scale of ventures. Watch what happens when each of them takes a month away.
The first one's work keeps moving. The triage rules she wrote still route. The pricing defaults she extracted from three deals still price. The playbook her last win produced is the onboarding doc for the next hire. Nothing about her presence is load-bearing, because the load was moved into structure.
The second one's work waits. Questions pile up in his inbox. Decisions stall at exactly the points only he has the instincts for. Ten years of pattern recognition, and it lives entirely inside one skull, available business hours, his timezone.
Both have ten years. One has ten years of experience. The other has one year, repeated ten times.
Experience is a conversion rate
The useful definition: experience is the conversion rate of wins into repeatable systems. Not exposure, not tenure, not the count of seasons survived.
The five-area frame used elsewhere in this archive treats experience as the last station of a chain: the track record of turning small wins into repeatable systems, as distinct from intellect, resilience, motivation, and cross-team behavior. That last clause is doing all the work. A win closes. The question is whether anything was extracted from it that survives the closing: a default, a threshold, an escalation rule, a playbook, a checklist that now catches what memory used to.
Why single experience out from the other four areas? Because the others are capacities you rent daily. Intellect, resilience, drive, and collaboration are all spent and renewed in the present tense. Experience is the only one of the five that deposits. It converts present tense work into assets that keep producing when the person is elsewhere. That is the compounding difference, and it is why Teams Ship at Their Weakest Area, Not Their Average treats the experience station as the one that decides whether this year's wins become next year's starting position.
The conversion is not automatic. Repetition does not produce extraction. This is the part the years-based resume cannot see: the builder who ships, reflects never, and ships again is running a loop with no learning step. From MVP to MVP Loops: Testing Ventures Continuously makes the same argument at the venture level. The loop only compounds if the test produces a retained conclusion. A win without extraction is a test whose result evaporated on contact with the next sprint.
What separates ten years from one year repeated?
One question, asked after each win: what now exists that did not exist before, and that makes the next one cheaper?
Three honest answers exist. An artifact: a rule, a default, a script, a document that carries the decision forward. A system: the artifact wired into how the work actually runs, so it executes without anyone choosing to remember it. Or nothing: the win closed, the lesson stayed in the room, and next quarter's version of the same problem starts from zero with better vibes.
The third answer is more common than anyone admits, because organizations reward the closing, not the extraction. Shipping is visible, celebrated, and measured. Systematizing is invisible until the systematizer leaves, at which point it is renamed, tragically, as a coverage gap. The person who spent Friday afternoon writing the triage rules produced no output that day by every dashboard in the company, and produced the only thing still running three years later.
There is a limit worth respecting, in the other direction. Extraction has diminishing returns, and the instinct to systematize everything produces procedure hoarders whose playbooks nobody reads. The standard is not documentation volume. It is whether the extracted thing runs. A single default that actually prices deals beats a wiki of lessons that died on publication.
Experience transfers only as systems
The hiring implication is uncomfortable. When you hire the second builder, the ten-years-of-instincts one, you did not acquire ten years of experience. You acquired access to it, business hours, his timezone. The experience remains where it was earned.
What transfers is what was converted. This is why Judgment Does Not Transfer deserves reading beside this piece: even exported systems carry the assumptions of the ground that grew them, so the transfer is partial at best. But partial transfer of real systems beats perfect retention of instincts in someone who might leave. The venture-building version of this is Venture Building Runs on Loops of Experts: the studio's asset is not the experts it can call. It is the loops the experts left behind.
So the years-of-experience filter, used flat, measures exposure. Two candidates with ten years can differ by a factor of ten in what the organization actually gets, and the resume cannot tell you which is which. The interview question that can: show me one system from your last venture that still runs without you. The answer is either an artifact or a silence, and the silence is data.
What this is not
Not an argument that raw time is worthless. Reps matter. The instincts of the second builder are real, and in a live room they are worth money. The claim is narrower: instincts are a capacity, not an asset. They rent. They do not compound, and they leave with the person.
Not an argument for documenting everything. The extraction standard is whether the artifact runs, not whether it exists. Documentation without abstraction is a museum.
And not a claim that everyone must systematize. A team needs some people converting wins into structure, and the chain-coverage logic applies: the conversion discipline needs to exist somewhere on the team, not everywhere on the team.
The move
Adopt one question as the closing step of every win: what artifact now exists that makes the next one cheaper?
- Ask it in the retro, out loud, before the room disperses. If the answer is nothing, say so plainly. That win deposited no experience.
- Cap the extraction at one artifact per win. The forcing function keeps the extraction honest and the museum small.
- For hiring, replace the years line with the systems line: name one system from your last venture that still runs without you.
Take-home
- Experience is a conversion rate, not a duration. Audit your last three wins for what they deposited. The count is your real tenure.
- If your best people's knowledge leaves the building every evening, that is not a retention risk. It is an extraction debt, and it compounds quietly.
- In interviews, ask for one system that outlived the person's involvement. The pause before the answer is the resume's true number.
Related reading:
- Teams Ship at Their Weakest Area, Not Their Average
- Judgment Does Not Transfer
- From MVP to MVP Loops: Testing Ventures Continuously
- Venture Building Runs on Loops of Experts
- Two of the Five Areas Leave Artifacts
If your team keeps reliving the same year with better instincts, the experience station is running without its conversion step. If you want help installing the extraction habit before more wins evaporate on contact with the next sprint, book a discovery call.
FAQ
What separates ten years of experience from one year repeated?
The conversion step. Experience is the conversion rate of wins into repeatable systems: defaults, thresholds, playbooks, escalation rules that keep working when the person is absent. Without conversion, each year deposits instincts only, which rent daily and leave with the person. With conversion, each year raises the starting position of the next one.
Why do years of experience mislead in hiring?
Because years measure exposure, and exposure without extraction does not transfer. Two candidates with identical tenures can differ enormously in what the organization actually receives. The discriminating question is to name one system from their last venture that still runs without them.
How does a team convert wins into systems?
By asking, at the close of every win, what now exists that makes the next one cheaper. The answer should be one running artifact, a rule, default, or playbook wired into how work executes. One artifact per win is the honest cap. More produces procedure hoarders; none produces a decade of relived years.
Is documentation the same as systematizing?
No. Documentation records; systems execute. The test for real extraction is whether the artifact runs without anyone choosing to remember it. A single pricing default that operates inside the workflow outperforms a wiki of lessons nobody opens, because the museum is not the asset.