LEAD21 and What Is Still Missing

Reference to Series 1: all eight papers, and by name the second part of WP 7 · The Superintelligent Organization, in which LEAD21 is tested against seven conditions from AGI research. The result of that test appears below unchanged, including where it is uncomfortable. The evidence, with references, is there.

ABSTRACT

Here the name falls, third from last, and here the call is made. The episode has three movements: LEAD21 fulfils the four functions from Gate 5 with named instruments; it suffices today as an operating system for human coordination, with the evidence status declared as Series 1 demands; and what is missing is the platform with the app — every open place from WP 7 lies on the AI side. Gate 6 opens when a metric over twelve months is fixed that can also fail.

Suggested citation: Huesch, G. (2026). LEAD21 — and What Is Still Missing. Series 2 · The Superintelligent Leader, Episode 6. Working Paper Series 2026, Global Leadership Consultancy.

THE GATE

PART 1 · THE FINDING

WHERE IT COMES FROM

This episode stands on all eight papers of Series 1, and by name on the second part of WP 7 · The Superintelligent Organization. There LEAD21 is tested against seven conditions from AGI research, and there too stands the result that carries the third movement of this episode: three covered, four partly, in every one an open place.

The benchmark of the first movement — the four functions — comes from WP 0 · The Second Leap and WP 6 · The Legible Organization, via Gate 5. The placement of the author's own evidence level in the second movement refers to WP 3 · No Side Door to Belief, where the evidence levels of the leadership literature are recorded.

All evidence stands with references in Series 1. This episode introduces none — it assigns, and discloses what is open.

AN OPERATING SYSTEM WITHOUT A PLATFORM

That is the exact situation, and it is not a figure of modesty.

An operating system is worth no less than a platform. It is what a platform can build on at all — whoever reverses the order builds a surface over a mess. But as long as no surface exists in which the operation actually runs, every user has to make the connection himself, every day, by hand.

That is exactly what you have done in the last five gates. You counted, sorted, laid side by side and kept a resubmission in mind. Every one of those operations is work a machine would do better. That is not a criticism of your discipline — it is the specification of a product that does not yet exist.

This episode therefore has three movements, in this order.

FIRST MOVEMENT · THE FOUR FUNCTIONS ARE FULFILLED

Gate 5 set up the benchmark, not LEAD21. The four functions — context, relevance, feedback, continuity — are derived from WP 0 and WP 6 and stand there before any product name falls in this series. That is the order that matters: first the benchmark, then the thing that can be measured against it.

And it goes beyond the four. The system comprises 21 instruments, not four — Series 3 will show them individually, with their effect and the typical implementation error. What counts here is not the number 21. It is that every function hangs on a named instrument and not on an intention.

SECOND MOVEMENT · IT SUFFICES TODAY — AND THIS IS THE EVIDENCE STATUS

LEAD21 works. In post-merger integrations, in hyper-growth situations, over years, in more than 30 countries. It suffices as an operating system for human coordination.

And here comes the part that would be missing from a sales slide. Series 1 accused the leadership literature of omitting the difference between measured results and reported assessments. The episode in which one's own name falls is the place where that accusation must be applied most strictly to the author himself. So:

One of eight surveys is measured: a one-year study with 567 leaders, Prof. Beate Hanson & Prof. Mohit Bhandari, McMaster University, with improvements in 8 of 9 climate dimensions.

All others are expectation surveys with self-report. The figures cited in LEAD21 materials — +87% clarity, +58% efficiency, +50% performance, +46% alignment, −50% negative stress — are participant-reported assessments. They are real, they were collected, and they are not a measurement.

The in-house effectiveness study is publicly declared at evidence level 3 (scale: level 1 is the highest). Series 1 showed in WP 3 that the leadership literature sits at levels 4 to 5. Level 3 is better than the field and far from the top. Both belong stated.

Anyone who takes this distinction for a disadvantage has not read the series. A system that names its evidence level is the more testable one — and testability is the only reason to believe an assignment.

THIRD MOVEMENT · WHAT IS MISSING, AND WHY THAT IS THE CALL

WP 7 did not test LEAD21 against the four functions, but against seven conditions from AGI research. That is a harder benchmark, and the result stands here unchanged:

Three conditions covered, four partly, and in every single one at least one open place.

The decisive sentence of this episode appears in WP 7 at condition 0, integration, and reads there: covered for human coordination · open for AI adoption.

That holds for all seven. Every open place lies on the AI side, not on the human side. And with that, the finding is not a deficiency to be argued away. It is the specification.

What is missing is the platform and the app. The surface in which the 21 instruments run. Data that stays, instead of disappearing into minutes. And an AI that takes over what a human cannot do on his own material: make the implicit explicit.

That is exactly what the prompts of this series did provisionally, by hand. Every prompt in Series 2 anticipates a function that belongs built into that platform — the dependency test, the checking budget, the extension, the codebook. Not one of them ought to be a prompt. All of them ought to be buttons.

WHY THE ORDER CANNOT BE REVERSED

And with that, the claim this episode carries: the operating system is the condition, not the consequence.

WP 0 showed that AGI amplifies organizational maturity — including immaturity. Whoever sets AGI on an organization without a nervous system gets a faster mess, at higher resolution, with greater confidence. Whoever builds the operating system now has what a future superintelligence can dock onto at all.

In that sense LEAD21 is the sine qua non — the condition without which the next steps have nothing to grip. Not because it would be the only possible system. But because some chosen system is the precondition, and a system not chosen is none.

WHAT THE LEADER RECOGNIZES HERE

1 · He found the gaps himself. The inventory from Gate 2, the checking budget from Gate 4, the function assessment from Gate 5 lie before him as his own documents, with his date on them. This episode assigns. It no longer persuades, because there is nothing left to persuade.

2 · A system that names its open places is the more testable one. Whoever reports everything covered has not tested. The four partly covered conditions are the reason to believe the assignment of the three covered ones.

3 · Waiting is the most expensive option. The platform is coming — from whomever. The operating system has to stand first, otherwise the platform digitizes a mess. That is the only sentence in this series that advises haste, and it advises haste on the system, not on the purchase.

PART 2 · THE WORK

THE OBSERVATION TASK

One week, three documents side by side.

Lay your inventory (Gate 2), your checking budget (Gate 4) and your function assessment (Gate 5) side by side — physically, three sheets on a table, or three windows on a screen.

Mark every gap that appears in more than one document.

That intersection is your finding. Everything else is a single observation, and single observations are the reason leaders start five initiatives at once. A gap that appears in two of your own documents has shown itself twice, independently.

In most cases two or three gaps remain. That is a good number. If seven remain, you marked too coarsely.

THE IMPLEMENTATION TASK

Monday, one thing. One metric that can fail.

Formulate one metric for the next twelve months. Not five. With three entries:

1. Baseline — the number today, with the date.

2. Target — the number that should stand there in twelve months.

3. Method — measured or reported. Both are permitted. Concealing is not.

And underneath it the sentence that distinguishes most metrics from intentions: what will you do if it is not reached?

Without that sentence it is not a metric. A metric whose failure has no consequence is an intention with a decimal place.

Examples that meet the condition: the number of class D decisions (Gate 1); the average age of the three oldest open items (Gate 4); the share of decision minutes carrying an extension line. All three can be missed, and all three come from your own documents.

THE PROMPTS

Prompt 6.1 · The gap map ``` Below are three results from my own work: A) my inventory of leadership initiatives of the last three years B) my checking budget of open decisions C) my assessment of the four functions (context, relevance, feedback, continuity)

Derive a gap map from them: 1. Which gap appears in more than one of the three documents? That one first. 2. Which gap is a format problem (something is not spoken) and which a structural problem (something has no place)? 3. Which gap would close by itself if another were closed? State the order.

Format: table gap | evidenced in A/B/C | format or structure | depends on.

Important: derive only what stands in the three documents. Do not enter a gap you consider typical. Do not recommend a method or a product — I will do the assignment myself afterwards.

The three documents: [insert here] ```

Prompt 6.2 · The counter-check

This prompt is aimed against LEAD21, not for it. That is deliberate. Whoever introduces a system that does not survive a counter-check would have been better off not introducing it — and whoever does not offer a counter-check is selling.

``` Below are my gap map and the description of a leadership system I want to introduce.

Your task is the counter-check, not the confirmation: 1. Which gap on my map does this system demonstrably not address? 2. For which promise of the system is there no evidence in the attached description, only a claim? 3. Which of my gaps would close even without this system — through a change to an existing format? 4. Which metric would have to be different in twelve months for the introduction to count as successful? Formulate it so that it can also fail.

In your answer distinguish explicitly between measured results and reported assessments. Where the description does not permit that, write "method of measurement not stated".

Gap map and system description: [insert here] ```

Prompt 6.3 · The requirements specification for the platform

This is the prompt in which the call becomes concrete. What comes out is the specification of a product that does not yet exist — written from your work and not from a market study.

``` Below are my gap map and the list of manual operations I performed in Gates 1 to 5 — counting, sorting, laying side by side, keeping a resubmission in view.

Derive from them a requirements specification for a leadership platform that takes over those operations: 1. Which of these operations would software have to do automatically so that I do not repeat them monthly? Per operation: trigger, input, result. 2. Which data would it have to hold permanently so that continuity does not hang on one person? 3. Which analysis would only become possible once this data exists over twelve months — and is impossible today? 4. Where would an AI analysis be helpful and where dangerous? For "dangerous": name the concrete harm, not a general reservation.

Format: four sections, each as a list. No product names, no vendor recommendation — a requirements specification, not a comparison.

Gap map and operations: [insert here] ```

And an open invitation, stated openly. If you send this specification back, it enters the specification of the platform this episode speaks of. This is not hidden market research — it stands here so that you know before you decide. You can also keep it. It serves you either way, because a requirements specification is at the same time the test sheet for every tool you will be offered in future.

PART 3 · ABOUT GERALD HUESCH

Whoever names a product in his own episode must disclose the evidence status for it. It stands above, in the second movement, in full, including the level.

Systems shape behaviour. Through the selection of the right instruments, mechanics and leadership codes — derived from values and leadership guidelines — the behaviour of everyone involved is synchronized and aligned. Not through motivation, but through a system that produces the right behaviour when applied. Sales has Salesforce. Finance has SAP. Leadership has LEAD21.

About the author. Gerald Huesch is the creator of the leadership operating system category – "one of the most professional and creative minds in the domain of leadership and difficult change projects" (Boston Business School). With LEAD21 he built the first operating system for leadership – 21 evidence-based methods, the comprehensive solution and the missing link for complete transformation, extreme growth and high-volume post-merger integration: a unique leadership system for human coordination at scale, in which alignment and culture emerge as a by-product. He has trained more than 30,000 executives in over 30 countries (NASA, Rolls-Royce, Airbus, Omnicom and others) and delivers executive leadership training, leadership development and executive mentoring for executives in Germany, the Americas, Asia Pacific and the UAE. Participants from 300+ companies report +87% clarity, +58% efficiency, +50% performance, +46% alignment and −50% negative stress after LEAD21 (participant-reported); a one-year study with 567 leaders (Prof. Beate Hanson & Prof. Mohit Bhandari, McMaster University) showed improvement in 8 of 9 climate dimensions. Dr. Charles Pellerin (NASA) on LEAD21: "Unique in the world, and a capability that nobody else has." gerald-huesch.com/about-en

PART 4 · THE GATE AND THE PATH

THE CHECKPOINT — GATE 6 OPENS WHEN

A metric is fixed that can also fail in twelve months — with baseline, target and method of measurement, and with the sentence underneath stating what happens if it is missed.

And: the requirements specification from 6.3 exists as a document, with at least one point marked "impossible today". That single point is the most important of the whole gate, because it is the boundary between what discipline can accomplish and what needs a platform.

And if you find no metric that can fail? Then you have reproduced the most common finding in leadership work, on yourself: what looks good is what gets measured. Note it. Gate 7 returns to it, because a metric without a band is not a control but a direction.

STATUS OF THE GATES

Gates 1 to 5 worked on the operation. Gate 6 ordered it. Gates 7 and 8 work on what outlives the operation.

WHAT COMES NEXT

Gate 7 · The Sentient Organization. You have a metric. Gate 7 asks whether the words in which it is reported mean the same thing to everyone — and in almost every organization the answer is no. There it is about the codebook that was never worked on, and about the difference between body temperature and revenue.

The lock there: the same word means different things to two speakers.

OPEN QUESTIONS FROM THIS EPISODE

The assignment of the four functions to instruments is an assignment by the author, not an independent test. It is testable because the instruments are named — tested it is not.

The platform does not exist. Everything stated about it in this episode is specification and intent, expressly marked as such. A requirements specification is not a product.

Whether an operating system really is the necessary condition for beneficial AI adoption is the central prediction of Series 1 and not evidenced. WP 0 formulates it as a falsifiable prediction. This episode adopts it as a claim and not as a finding.

The evidence level 3 of the in-house effectiveness study is a self-classification against a publicly stated scale. An independent classification does not exist.

REFERENCES

Series 1, Working Paper Series 2026 WP 7 · The Superintelligent Organization, Part 2. The test of LEAD21 against seven conditions; the result three covered / four partly; the formula "covered for human coordination · open for AI adoption" at condition 0. WP 0 · The Second Leap. AGI amplifies organizational maturity including immaturity; the falsifiable prediction on the architectural condition. WP 3 · No Side Door to Belief. The evidence levels of the leadership literature, against which the in-house level 3 is to be placed. WP 6 · The Legible Organization, WP 5, WP 4, WP 2, WP 1 — as sources of the four functions and the kinds of gap.

On the author's own evidence One-year study with 567 leaders, Prof. Beate Hanson & Prof. Mohit Bhandari, McMaster University: improvements in 8 of 9 climate dimensions. The only measured one of the eight surveys. All further figures cited are expectation surveys with self-report; evidence level of the in-house effectiveness study: level 3 of 5, self-classified. Details at gerald-huesch.com/evidence-en.

The primary sources used in the Series 1 papers from the 19th Conference on Artificial General Intelligence (AGI-26), San Francisco, 27–30 July 2026, are identified there with volume and page numbers. This episode introduces no new references.