Haute Lumière

Commerce · II.04 · MMXXVI · daylight

La Bourse  /  Volume II  /  Nº II.04  /  Workbook — the Gainshare employee

WORKBOOK — THE LUMINOUS GAINSHARE EMPLOYEE

Chapter II.04 · The Science of Wholeness

For the person working inside a gainshare arrangement — where a defined share of verified improvement returns to the people who created it. This chapter is about which claims hold up when somebody checks them, which is exactly the skill a gainshare runs on, because a gainshare is a claim that gets checked every quarter.

---

WHY THIS CHAPTER IS YOURS

A gainshare is a measurement with money attached. That means three things about your working life that nobody tells you on day one.

Your claim will be read by somebody who did not make it. Finance will read it. Possibly an auditor. Certainly somebody whose budget it comes out of. They are not hostile, and they will still find the loose number, because a loose number is the easiest thing in the world to find when you did not write it.

The verification is where the money actually is. Not the improvement — the verified improvement. An improvement nobody can check is an improvement nobody pays for, and the gap between those two sentences is where most gainshare disappointment lives.

And the rules of your scheme are a leverage point, in Meadows' exact sense. They sit at number five on her list, above information flows at six and far above the parameters at twelve. The percentage in your scheme is a parameter. The rule about what counts as a verified gain is a rule. The rule outranks the percentage, and this workbook will show you the arithmetic that says so.

---

PART ONE — DISCOVERY

Days 1–30: find the gain that is already provable

Exercise 1.1 — The provability sort (2 hours)

List every improvement your team has made in the last eighteen months. Now sort them into three columns, and be honest, because this sort is the whole exercise:

Most teams find the middle column is the largest, and that is good news rather than bad — a missing baseline is a thing you can start this month, and it converts column two into column one over one measurement period.

The third column is not wasted. It is the argument for changing the scheme's rules, which is Part Three.

Exercise 1.2 — The threshold check (90 minutes)

Before you propose anything, check the gradient. Ask, for each candidate:

• Is the effect at least three times the period-to-period noise in that measure?
• Is there a real difference in outcome between doing it well and doing it badly?

If the answer to either is no, the gradient is subcritical — the chapter's Rayleigh result, transposed. Below the critical gradient the same effort crosses the system and organises nothing. You will work hard and the number will come back inside the noise band, and the scheme will conclude that nothing happened.

The physics also hands you the fix, and it is the depth term, which enters cubed: a tenfold deeper layer needs 1,000× less gradient. In your scheme, depth means a longer measurement window or a pooled measure across more of the operation. Ask for a four-quarter window before you ask for a bigger percentage.

Exercise 1.3 — The maintenance you are already doing (one hour, with your team)

Ask your team one question out loud: what do we look after here that nobody counts? You will get five answers in ten minutes. Equipment somebody keeps running. A relationship somebody maintains. A dataset somebody cleans. Knowledge somebody rewrites.

Every one of those is throughput maintaining order — Prigogine's balance, on your floor. A body dissipating 100 W exports 0.3413 W/K to hold its shape, which is 29.49 kJ/K per day, and the same logic holds for a production line and a customer list. This is the most under-claimed category in every gainshare scheme, because the gain it produces is an absence: the failure that did not happen.

---

PART TWO — THE ARITHMETIC

Days 31–45: compute what you are owed

Exercise 2.1 — The counterfactual, written before the intervention

The single highest-value document you will produce is a signed baseline dated before you change anything, naming: the measure, the method, the period, the expected noise band, and who verifies.

An unagreed baseline is not a baseline. It is a future dispute in which you will be the one arguing from memory.

Exercise 2.2 — The avoided-failure calculation

For maintenance-type gains, the value is the failure that did not occur, and the chapter gives you the shape of it. On a condition model decaying faster the worse it already is, starting from a realistic 99 and failing at 55:

   maintenance at 100 % of the decay rate  →  failure in 31.75 years
   maintenance at  95 %                    →  failure in 21.75 years
   maintenance at  90 %                    →  failure in 17.68 years
   maintenance at  75 %                    →  failure in 12.09 years
   maintenance at  50 %                    →  failure in  8.26 years

A five percent funding shortfall costs 31.5 percent of the asset's life. Read that in the direction that pays you: the team that holds maintenance at full rate rather than 95 percent is buying ten years of asset life, and ten years of asset life on a £40,000,000 replacement-cost base is a number your finance function can price without being persuaded of anything.

Exercise 2.3 — Your scheme's own elasticity

Here is the arithmetic that decides what to ask for. A stock with goal-seeking inflow a·(G − S) and outflow k·S settles at S* = aG/(a+k). With a = 0.3, k = 0.1, G = 100 that is 75.00:

   +20 % on the parameter a   →   S* = 78.2609   (+4.35 %)
   +20 % on the goal G        →   S* = 90.0000   (+20.00 %)

   elasticity to the goal      = 1.000 exactly
   elasticity to the parameter = k/(a+k) = 0.250
   ratio                       = 4.00×

Now transpose it. The share percentage in your scheme is the parameter. The definition of a countable gain is closer to the goal. A 20 percent improvement in your share moves your take by 20 percent of a number that may be small; a change to what counts can move the number itself.

And the honest counter-case, which you should know before you walk into the meeting: with a = 0.1 and k = 0.9 the elasticities are 1.000 and 0.900, a ratio of only 1.111×. If your scheme is drain-dominated — if most of what you generate is already being captured and counted, and the constraint is the split rather than the definition — then the percentage really is the thing to ask for, and asking for a rule change instead will cost you a year.

Compute which world you are in before you choose. It takes an afternoon.

---

PART THREE — DESIGN

Days 46–70: make the uncounted countable

Exercise 3.1 — The rule proposal

Draft one page proposing a single change to what counts as a verified gain. Requirements, all five:

1. One change. Not a redesign of the scheme.
2. The method named. How it is measured, over what window, by whom.
3. A worked example from the last year, with the number it would have produced.
4. The cost of verification, honestly stated, including whose time.
5. What it would have paid the company, not only what it would have paid you. The fifth item is the one that gets it read.

Exercise 3.2 — The information flow

Meadows puts information flows at number six, immediately above rules. Ask a narrow, answerable question: what does the scheme's administrator see that we do not, and what do we see that they do not?

The answer is nearly always asymmetric in both directions, and closing it costs a shared report rather than a negotiation. A team that can see the running gain figure behaves differently from a team that finds out in April, and the difference is not motivation — it is the ability to correct in period.

Exercise 3.3 — The structural coupling argument, for your own defence

At some point your scheme will have a bad quarter because of something external — a price move, a shortage, a customer. Prepare the argument now, before you need it, because it is exact and it is fair to both sides.

The chapter's number: identical 10 percent shocks into two firms produce peak overshoots of 0.152 percent and 52.662 percent, a ratio of 345.6×, entirely from internal damping. The shock does not determine the response; the structure does. That cuts both ways and you should say so. It means an external excuse is usually weak — but it also means that where your team has genuinely improved the structure, the improved response in the next shock is yours, and it is attributable.

Write the two sentences now. Quarters where you need them are not good drafting conditions.

---

PART FOUR — DESTINY AND DELIGHT

Days 71–90: make it hold

Exercise 4.1 — Get the measure into the standing pack

Anything reviewed monthly persists. Anything reviewed by exception does not. The gain measure belongs in the pack that gets read whether or not anything is happening, and it should get there quietly, in a month when nobody has an opinion about it.

Exercise 4.2 — The second owner

One person tracking the gain is a hobby. Two is a practice. Recruit the second person by giving them credit for the first verified result — not by asking them to help.

Exercise 4.3 — Check somebody else's claim, and invite them to check yours

This is the practice that makes a gainshare trustworthy from the inside, and it is the chapter's own method turned on the scheme. The author of a claim is the one person who cannot see the gap between what the method says and what the write-up says, because they read the write-up through the method.

So pair up. Read each other's baselines and each other's arithmetic, on a route that shares no assumptions with the one that built it. A fault found now costs a line. The same fault found by finance in April costs the quarter and some of the trust.

Exercise 4.4 — The thing to enjoy

There is a specific pleasure in a verified number: not triumph, recognition. You knew the team was doing something good. Now it is on a page in a form another person cannot argue with, and it is going to be paid.

And a better one after it, which the chapter names: the meeting where somebody who was not persuaded uses your number in their own argument, for their own reasons, without attributing it. That is the moment a measurement becomes infrastructure. It has stopped being yours, which is exactly what you wanted.

---

THE ONE THING TO GET RIGHT IN YEAR ONE

If you do nothing else from this workbook, do this: sign a baseline before you change anything.

Not a spreadsheet. A dated page, naming the measure, the method, the period, the expected noise band and the verifier, with two signatures on it — yours and whoever will be asked to confirm the result. It takes forty minutes and it is the only document in the whole arrangement that cannot be reconstructed later.

Everything else in a gainshare can be argued about afterwards and usually is. The percentage can be renegotiated. The definition can be proposed. The window can be extended. A baseline that was never agreed cannot be recovered at all, because the world it described has already gone, and the only honest answer six months later is that nobody knows what the number was before.

That is not a small administrative point. It is the difference between a verified gain and a good story, and the people who do well inside gainshare schemes over several years are, almost without exception, the people who write the boring page first.

---

KNOW YOUR SCHEME — A CHECKLIST

Fill this in. Most people cannot fill it in from memory, and every blank is a place where money is decided without you.

The last row is the important one. If the answer is "nobody has ever proposed one", the answer is not that it is forbidden — it is that nobody has tried, and the first one-page proposal in a scheme's history is generally read with interest.

---

THE CONVERSATION, SCRIPTED

Opening. "I have a change to the definition, one page, with the number it would have produced last year and what it would have cost to verify."

When asked whether this is about your share. "No — the share is a parameter and it moves the outcome by about a quarter of what a definition change does. I ran the elasticity. I brought it."

When told the effect is inside the noise. "Agreed, over one quarter. Over four quarters the same effect clears the band, and a longer window costs the scheme nothing. Can we take the window before we take the percentage?"

When told maintenance is not a gain. "The avoided failure is the gain. At 95 percent of the maintenance rate the asset reaches failure in 21.75 years instead of 31.75 — a five percent shortfall costs 31.5 percent of the life. I can show the model."

Closing, always. "What would you need to see for this to be verifiable? I would rather build the thing you can check than argue for the thing I already have."

---

APPRECIATIVE QUESTIONS FOR YOUR TEAM

1. When has this scheme paid us for something we genuinely did, and what made that one provable when others were not?
2. What do we look after here that nobody counts, and who has been quietly doing it?
3. If we could change one definition rather than one percentage, which would we change, and what would it have paid last year?
4. What can the administrator see that we cannot, and what would it take to see it together in period rather than in April?
5. When the last shock came, which part of our response was ours rather than the shock's? What did we build that made it so?
6. What would make checking each other's numbers feel like a good hour here rather than an inspection?