Full name

Use the same name all week — it is how your submissions are matched. Each export names its own file from it, for example 1-muchson-day1-l1-diagnostic.

Route 3 · Level 3 · Management decision

Fund it anyway, and say what you left uncovered

Level 1 established what the file records. Level 2 showed that no offer wins on every axis. Level 3 gives you a budget that cannot fund everything it should, and asks you to fund it anyway — and to say, in writing, what you left uncovered.
A memo that funds everything has not used this exercise.

Materi C · Level 3 · Management decision · 60 min

Deciding under a budget that cannot fund everything

Six short cards (48 min of reading and exploring, the rest is discussion). They give you the tests you need for Task 3: how reversible a commitment is, why one success is not a forecast, how to score under a cap, what leaving something out costs, and who owns what.

Reversible and irreversible decisions

9 min

Most decisions can be undone cheaply. The few that cannot deserve the argument; the rest do not.

In plain words

Some decisions are like a door you can walk back through: if it goes wrong, you undo it cheaply (a ). Others are like a door that locks behind you: hard or expensive to undo (a ).

Why it matters

Most decisions are two-way doors, so make them quickly and save the long discussions for the few one-way doors. When something is a one-way door, look for a small, reversible first step before committing in full.

How to read the picture below

Click each of the three example decisions. It slides to its door, and a short note explains why it belongs there.

Door test

Exploratory · not graded Select a decision to send it to its door.

Three generic decisions. The case levers are not used here, so Task 3 stays yours.

Bezos (2015 Letter to Shareholders, Amazon.com), the Type 1 / Type 2 framing, popularly known as vs decisions: Type 1 decisions are consequential and irreversible (or nearly so) and deserve slow, deliberate, consultative process; the large majority of decisions are Type 2 — reversible, “like walking through a door: if you don't like the decision, you can go back through” — and should be made quickly, by small groups or individuals.
Micro-example (generic, not the case levers): staffing a role part-time for six months is a two-way door (reassign it back); signing a multi-year with a client is closer to a one-way door.

How to decide when this comes up in the task

Ask of each lever: if it goes badly, can you go back through the door? A part-time role you can reassign is a two-way door; a multi-year contract with and relationship cost is closer to a one-way door.
Slow, deliberate process belongs on the hard-to-reverse lever. Do not spend the same effort on the reversible ones.
The practical answer to a one-way-door lever is to stage it: a smaller, reversible first step before the full commitment (see C5).

The planning fallacy and reference-class forecasting

9 min

A single success does not forecast a program — an of similar programs does.

In plain words

When we plan, we tend to look only at our own case and its story (the ), and we end up too optimistic. The fix is to look at how similar projects actually turned out (the outside view) and see where our case sits among them. This is .

Why it matters

In Tasks 1 and 2 you studied one customer, Kessler. In Task 3 you plan for many customers. One good result with Kessler does not mean the whole programme will go that well.

How to read the picture below

Turn the outside view on and off. The narrow red curve is the optimistic story of one case; the wide teal curve is what similar programmes really achieved. Notice how far out on the optimistic side the Kessler marker sits.

Inside view vs outside view

What this showsKessler's one good outcome sits in the far optimistic tail of the wide outside-view curve — most similar programs land well to the left of it. A single success this far out is not the typical result; it's the lucky one.

The anecdote sits in the far tail of the narrow curve; the wider curve is what similar cases actually did.

Kahneman & Lovallo (1993), Timid choices and bold forecasts: A cognitive perspective on risk taking, Management Science, 39(1), 17–31: planners take an (focus on the specifics of the case at hand, its plan and obstacles) which systematically produces optimistic, narrow forecasts; the corrective is an / : look at the distribution of outcomes in a class of similar cases and locate the current case in that distribution, rather than reasoning from its own story.
Direct link to Routes 1 and 2: filing Kessler's Task 1 verdict and Task 2 recommendation is an inside view of one case (). A program in Task 3 needs its own reference class: the 70% rate from Materi A1, not Kessler's specific numbers.

How to decide when this comes up in the task

Your Task 1 and Task 2 work is an inside view of one case (). A program needs its own reference class: the 70% rate from Materi A1, not Kessler's specific numbers.
When you say what a program will produce, locate it in the distribution of similar programs first. Do not reason from the plan's own story alone.
Say what stays uncovered with a number of your own (Block 3.5). An optimistic plan that names no uncovered remainder is an inside view.

Scoring options under one constrained budget

10 min

When a budget cannot fund everything, score options on independent axes and let the arithmetic rank them — then override it in writing if you must.

In plain words

When the budget cannot pay for everything, give each option a score on a few separate questions (how much benefit, how doable, how different) and multiply them. and are well-known versions of this. The score gives you a ranking, but you still have to check whether each option fits the budget.

Why it matters

The best-scoring option is often also the most expensive. If it goes over the budget it cannot be funded, however good it looks. And if you choose against the ranking, write down why.

How to read the picture below

First look at the scores alone: Option 1 wins. Then switch on “Show cost”. Option 1 now rises above the cost cap line, so it cannot be funded. The numbers are made up just to show the idea.

Score vs cap

What this showsBy score alone, Option 1 ranks first. Generic, illustrative numbers.

Toggle the cost overlay.

The curriculum's own scoring method, Benefit × Feasibility × Differentiation, belongs to a family of multiplicative discrete-axis scoring frameworks used in industry to prioritise under scarcity:
(Reach, Impact, Confidence, Effort), a prioritisation framework popularised by Intercom's product team (2016): RICE = (Reach × Impact × Confidence) / Effort.
(Impact, Confidence, Ease), a lighter three-factor version associated with growth-hacking practice from Sean Ellis (see Ellis & Brown, 2017, Hacking Growth): ICE = Impact × Confidence × Ease.
Why discrete beats continuous here: a free slider on “impact” cannot be checked or compared across participants; a Low/Medium/High (or 1–3) scale can, which is exactly why Task 3's allocation board uses discrete switch positions instead of a budget slider.

The caution every one of these frameworks shares

Multiplying scores hides the fact that Effort (or Cost) is not optional: a high-scoring option that busts the budget is not fundable. That is the constraint check Task 3 enforces mechanically rather than asking you to remember it.

How to decide when this comes up in the task

Never stop at the score. Check every option against the cap: a high-scoring option that busts the budget is not fundable. The board does this check for you; you still have to choose what to leave out.
Use discrete positions (None / Partial / Full), not a free amount: they can be compared across people and re-checked, which is why the allocation board has switches and not a budget slider.
If the arithmetic ranks something first and you fund something else, say so in writing (Block 3.6), with the reason.

Portfolio economics: opportunity cost and capital rationing

9 min

Funding one option is always also the decision not to fund another — that trade-off has a name and a price.

In plain words

Every time you spend money on one thing, you also choose not to spend it on something else. What you give up is the . When the budget is fixed (), some good ideas will be left out, and that is fine as long as you say so openly.

Why it matters

Task 3 is about the pool of 42 customers, not only Kessler. Getting more of them to stay moves far more money than winning back one account.

How to read the picture below

On the left, 42 thin outlines: one revenue cliff for each one-off customer. On the right, the single Kessler bar. Compare the size of the whole pool with the one case you studied before.

Portfolio cliff vs Kessler cliff

The single-project cliff from Materi A1, redrawn 42 times, next to the one Kessler bar.

: the value of the next-best option forgone; : choosing among positive-value options because the budget, not the merit of any one option, is the binding constraint (standard corporate-finance framing, e.g. Brealey, Myers & Allen, Principles of Corporate Finance).
Reuse the formula from Materi A1 (Gupta & Lehmann 2003: CLV = m · r / (1 + i − r)) at scale: raising the r across a segment of customers, not just one account, is what actually moves the number A1 introduced.
for scale: TechSolutions serves roughly 60 active customers; at a 70% one-off rate that is 42 one-off customers, the pool Task 3's allocation board covers.

How to decide when this comes up in the task

Every lever you fund is a lever you do not fund at that level: state the opportunity cost of what you left at None or Partial, and put a number on it.
Under capital rationing the , not the merit of any one lever, is the binding constraint. Positive-value options are left unfunded, and that is not a mistake, provided you say so.
A problem is sized on the pool (42 one-off customers), not on one account: the r moves across the segment, not just for Kessler.
Levers can depend on each other, and a harsher scenario lowers what a lever delivers. A lever that needs another in place first delivers less without it; a price war takes away part of what it would win. Read each output for each scenario before you compare options, and say which scenario a number belongs to.
What stays uncovered is a subtraction you can state: the pool minus the accounts a lever reaches.

Governance: who decides, who executes, who is informed

9 min

A plan without an owner and a date is a hope, not a decision.

In plain words

A plan only works if it is clear who does what. is a simple table for that: (does the work), (the one person who answers for the result), (asked for input) and (kept up to date). Every task gets exactly one Accountable person.

Why it matters

If two people are Accountable, in practice nobody is. And a big commitment that is hard to undo is safer if you start with a small , then scale up once the first results are in ( thinking).

How to read the picture below

The example is a different company, Alpenwerk, replacing its IT ticketing tool. Click any cell, including the empty ones, to read why it holds that letter. Watch where the A sits: small operational choices stay with the team lead, while a big three-year contract or a vendor crisis goes up to the managing director. Below it, the four test questions and a table of what each Task 3 role typically decides, does, is asked about or is told.

Worked RACI · Alpenwerk replaces its IT ticketing tool

Worked RACI example, Alpenwerk replaces its IT ticketing tool
Activity	IT team lead	Service desk	Purchasing (Einkauf)	Managing director
Test three candidate tools
Sign the 3-year licence contract (€90,000)
Migrate tickets and train staff
Escalate when the vendor misses the go-live date

Sign the 3-year licence contract (€90,000) · Managing director · A

A three-year commitment of this size is beyond what one department may sign. The managing director decides and answers for it.

What this showsThe test behind this cell. Accountable: who answers for the result and has the authority to decide? Exactly one per row. Notice where the A sits: operational work stays with the team lead; a large, long commitment or a vendor crisis goes up to the managing director.

A different case from Task 3 (), so the task stays yours. Click any cell, including the empty ones.

Four test questions, one per letter

A: Accountable: who answers for the result and has the authority to decide? Exactly one per row.
R: Responsible: who does the work? There can be several.
C: Consulted: who must be asked before, because they know something important or could block it?
I: Informed: who only needs to be told the result afterwards?
Empty: Empty: this role has no part in this activity. That is allowed and common.

The four roles in Task 3
Role (Task 3)	Typically decides	Typically does	Typically asked before	Typically told after
Head of Sales / CCO (you)	Customer relationships, commercial deals, how the sales organisation is set up (who owns which account)	Raises risks on accounts, leads customer negotiations	On delivery changes that affect a customer	About delivery routines that run as planned
Delivery PM	How delivery and post- work is carried out	Runs projects, handovers and reviews with the customer	On what a customer actually needs day to day	About commercial decisions that change their work
Legal / counterpart	Rarely the commercial outcome itself	Checks contract terms, compliance and the purchasing process	Before anything contractual is agreed	About outcomes that touch a contract
(managing director)	Commitments beyond one department's authority, conflicts over resources, what to do when a key account is at risk	Rarely the day-to-day work	On large or hard-to-reverse commitments	About commercial commitments the sales side negotiates within its authority

Role profiles are a practitioner observation for a German firm, not a fixed law. A real company can distribute authority differently; say which profile you assume.

matrix (, , , ), a standard project-governance tool (documented in the Project Management Institute's PMBOK Guide): every activity gets exactly one Accountable owner; any number of Responsible, Consulted or Informed parties.
Applied to a retention program, typical activities are account-owner assignment, contract negotiation for framework agreements, running quarterly value-realization reviews, and escalation to the when a named account is at risk. Typical roles: Head of Sales / CCO (you, in Task 3), Delivery PM, Legal / counterpart, Geschäftsführer.
thinking (Amram & Kulatilaka, 1999, Real Options: Managing Strategic Investment in an Uncertain World): stage an uncertain, harder-to-reverse commitment (like the framework-contract push in Task 3) as a smaller first, buying the option to expand once early results are in, rather than committing the full amount up front. This is the practical answer to a lever identified in C1.

How to decide when this comes up in the task

Give every activity exactly one Accountable owner. Any number of Responsible, Consulted or Informed parties is fine; two Accountable owners means nobody is.
Use one test per letter. A: who answers for the result and has the authority to decide? R: who does the work? C: who must be asked before, because they know something important or could block it? I: who only needs to know afterwards? A role with no part stays empty.
A is not the busiest role. The role doing the work is R; the role answering for it is A. One letter per cell: if the same role both decides and does the work, give it A.
C is before and two-way; I is after and one-way. If leaving a role out could make the decision wrong or invalid (contract terms, a legal check), it is C. If it would only surprise them, it is I.
Put A at the level that has the authority, and no higher. Operational work stays with the team that runs it; decisions about customers and the sales organisation sit with sales leadership; commitments beyond one department's authority, resource conflicts between departments and an account at risk go to the managing director.
Read each role by what it decides, does, is asked or is told in this company (the role profiles above), not by its seniority alone.
Carry the grid into the governance table: every lever you fund needs a named owner and a date, and the next review needs one too.
For a lever, stage it: buy the option to expand with a smaller pilot first (real options thinking), then commit the full amount once early results are in.

Field method for Task 3

2 min

Four steps, twenty minutes: Allocate → Read both scenarios → Govern → Price what you postponed.

In plain words

A short recipe for Task 3. Split the budget across the levers without going over the , read both scenarios, give every funded lever an owner and a date, and say what you left unfunded and what that costs.

Why it matters

You have 20 minutes. A memo that funds everything has skipped the real decision.

Time box: 20 minutes.

Task 3 at a glance
Block	What you do	Kind
3.1 Allocation board	Set three levers, read six figures, file within the	Objective
3.2 grid	One per row	Objective
3.3 Risk and reversibility	Name the hardest-to-reverse lever and a reversible first step	Judged
3.4 Governance table	Decision, owner, date per funded lever	Judged
3.5 Not funding	Name what stays uncovered, with your own G6	Judged
3.6 Executive summary	Four live chips: lever, reversibility, scenario, owner	Judged

How to decide when this comes up in the task

Fund within the cap, open both scenario tabs before you file, give every funded lever an owner and a date, and finish by saying what stays uncovered with your own number. A memo that funds everything has not used this exercise.

References · Materi C (8)

Frameworks from Day 1 Levels 1–2 are listed in the references of Routes 1 and 2.

Gupta, S., & Lehmann, D. R. (2003). Customers as assets. Journal of Interactive Marketing, 17(1), 9–24.
Bezos, J. (2015). 2015 Letter to Shareholders. Amazon.com. (Type 1 / Type 2, “one-way door / two-way door” decisions.)
Kahneman, D., & Lovallo, D. (1993). Timid choices and bold forecasts: A cognitive perspective on risk taking. Management Science, 39(1), 17–31.
Intercom (2016). The RICE scoring model for prioritization (Reach × Impact × Confidence ÷ Effort), as documented in Intercom's product-management practice.
Ellis, S., & Brown, M. (2017). Hacking Growth. Crown Business. (ICE: Impact × Confidence × Ease.)
Project Management Institute. A Guide to the Project Management Body of Knowledge (PMBOK Guide). (RACI matrix.)
Brealey, R. A., Myers, S. C., & Allen, F. Principles of Corporate Finance. McGraw-Hill. (Opportunity cost, capital rationing.)
Amram, M., & Kulatilaka, N. (1999). Real Options: Managing Strategic Investment in an Uncertain World. Harvard Business School Press.

Task 3 · Level 3 · 20 min

Task 3 · Decision Memo · Customer Retention Investment Program

Level 3 gives you a budget that cannot fund everything it should.

Premise · from Routes 1 and 2

Task 1 verdict

Not filed yet. (a suggestion, not a gate).

Task 2 recommendation

Not filed yet. (a suggestion, not a gate).

Continuity note

In Task 2, “Apply to all TechSolutions customers” stayed locked because Kessler is one customer (n = 1) and cannot license a company-wide rule. Task 3 does not lift that lock by generalising Kessler's numbers. It asks a different, legitimate question: TechSolutions' one-off rate is a portfolio fact (70%, from Materi A1), and a portfolio problem gets a portfolio program, funded and governed on its own evidence, not on Kessler's file.

The situation

Case assumption

TechSolutions GmbH serves roughly 60 active customers. At the 70% rate established in Materi A1, 42 are one-off: delivered once, with no standing account relationship. Competitive pressure is rising (NordByte IT's win at Kessler is one visible instance of a pattern, not proof of it). The Managing Director has approved a €200,000, six-month program to raise retention across this one-off segment: enough to fund a real pilot, not enough to fund everything the program could use.

You write this memo as Head of Sales / Chief Customer Officer. Addressee: the Geschäftsführer (managing director). The memo assembles beside the questions, in reading order, with the executive summary first even though you write it last.

Block 3.1 · Allocation board

OBJECTIVE

FIND IT· Route 3 → Task 3 → “Allocation board”. Answer in the fields directly below the board.

Analyse in the app. Write your result in the answer area below.

Draws on

L1 · Named account owners

Existing delivery PMs take on part-time account ownership for closed projects.

Rope it pulls (A3): Relationship

L2 · Value-realization reviews

A fixed-term contractor runs structured post-go-live reviews (the QBR practice in Materi B6).

Rope it pulls (A3): Relationship + Benefit

L3 · Framework-agreement push

Two sales engineers convert one-off accounts to 3-year managed-service contracts, the contract type Kessler tendered in Task 2.

Rope it pulls (A3): Economic + Contractual

Total 6-month commitment

€0

€200,000 of headroom under the €200,000 cap.

Switching tabs updates the meters. Nothing to press.

Named owners

0accounts

Reviewed accounts

0accounts

Framework conversions · Status Quo

0accounts

Uncovered pool

42of 42 accounts

How the board counts

Total commitment = the cost beside each lever's position, added up.
Named owners and reviewed accounts = the number beside L1's and L2's position.
Framework conversions start from the number beside L3's position. If L1 is None, that number is halved and rounded down, because no named account owner is assigned. The Price War tab halves the Status Quo figure again, rounded down.
Uncovered pool = the 42 one-off accounts minus the reviewed accounts.

G1 · Total 6-month commitment (€)

Read the total-cost readout on the board.

G2 · Accounts with a named owner

Read the L1 output at your chosen position.

G3 · Accounts receiving a review

Read the L2 output at your chosen position.

G4 · Framework conversions, Status Quo

Open the Status Quo tab and read the L3 meter.

G5 · Framework conversions, Price War

Open the Price War tab and read the L3 meter.

G6 · Uncovered pool (accounts with no review at all)

42 minus G3 — the board prints this directly, no arithmetic needed.

Checks requested: 0

Filing needs the total within €200,000, both scenarios opened and G1–G6 filled. It does not need any particular combination of levers.

Block 3.2 · RACI grid

OBJECTIVE

FIND IT· Route 3 → Task 3 → “RACI grid”. Answer in the grid itself.

Draws on

One rule is enforced: exactly one A (Accountable) in each row. Which role holds it is your judgement, and it carries into the governance table below. Materi C5 teaches the four letter tests and what each of these four roles typically decides, with a worked example on another company.

RACI grid: four activities by four roles
Activity	Head of Sales / CCO (you)	Delivery PM	Legal / Einkauf counterpart	Geschäftsführer
Named account-owner assignment
Framework-contract negotiation
Value-realization review execution
Escalation when a named account shows churn risk

R Responsible · A Accountable (exactly one per row) · C Consulted · I Informed. Press a pressed chip again to clear it.

Block 3.3 · Risk and reversibility

JUDGED

FIND IT· Route 3 → Task 3 → “Allocation board”, then the lever you would find hardest to undo. Answer below.

Draws on

Which lever is closest to a one-way door?0 characters

Name the lever in your allocation closest to a (Bezos, 2015), and how you would de-risk it before committing the full amount. Frame: “___ is the hardest to reverse because ___. Before committing in full, we will ___ (a smaller, reversible first step).” At least 30 characters. Your framework push is unfunded, so you may instead name the hardest-to-reverse lever you did fund, or state that no one-way-door lever was funded this cycle.

Block 3.4 · Governance table

JUDGED

FIND IT· Route 3 → Task 3 → “RACI grid” and “Allocation board”. Answer in the table below.

Draws on

One row per lever you funded above None, plus one row for the next portfolio review. Date format: YYYY-MM or a named milestone (for example ‘Month +3’). Rows 1 to 1 are needed for your allocation.

Governance table
#	Decision	Owner	Date
1
2·
3·
4·

Block 3.5 · What we are not funding this cycle

JUDGED

FIND IT· Route 3 → Task 3 → “Allocation board” (the Uncovered pool meter) and your G6. Answer below.

Draws on

What stays uncovered, and its price0 characters

Name what stays uncovered, using your own G6 figure, and its price: not a vague risk, a number. Frame: “We are not funding ___ this cycle. That leaves ___ accounts with no review (G6), which is not zero. If this proves costly, the cost shows up as ___.” At least 30 characters.

Citesyour G6 figure or a lever you left unfunded — not found yet

Block 3.6 · Executive summary

JUDGED

FIND IT· Route 3 → Task 3 → the “Live memo”, which shows this summary first. Write it last.

Draws on

The summary a Geschäftsführer reads first0 characters

Say what you fund, how reversible it is, what the two scenarios produce, and who owns it. Four chips underneath light up as the text names each.

LeverReversibilityScenarioOwner

Your decision memo is still missing:

Preview of your memo

File name: 3-participant-day1-l3-memo.html

Your memo, live

0 of 6 blocks complete

Customer Retention & Buying Behaviour in B2B IT Sales

Decision Memo — Customer Retention Investment Program

Course: Customer Retention & Buying Behaviour in B2B IT Sales
Position: Day 1 · Route 3 · Level 3 · Task 3
Participant: —
Date: 3 October 2026

To: Geschäftsführer · From: Head of Sales / Chief Customer Officer

Executive summary

—

1. Situation

TechSolutions GmbH serves roughly 60 active customers. At the 70% one-off rate, 42 are one-off: delivered once, with no standing account relationship. The Managing Director has approved a €200,000, six-month pilot program to raise retention across this one-off segment. Case assumption.

2. Allocation decision

Lever	Position	6-month cost	Produces
L1 · Named account owners	None	€0	0 accounts get a named owner
L2 · Value-realization reviews	None	€0	0 accounts receive at least one review
L3 · Framework-agreement push	None	€0	0 accounts converted to a framework contract (before adjustments)
Total		€0 (cap €200,000)

3. What it produced

Reading	Status Quo	Price War
Total commitment (G1)	€—
Accounts with a named owner (G2)	—
Accounts receiving a review (G3)	—
Framework conversions (G4 · G5)	—	—
Uncovered pool (G6)	—

4. Risk and reversibility

—

5. Governance

Activity	Head of Sales / CCO	Delivery PM	Legal / Einkauf	Geschäftsführer
Named account-owner assignment
Framework-contract negotiation
Value-realization review execution
Escalation when a named account shows churn risk

#	Decision	Owner	Date
1	—	—	—
2	—	—	—
3	—	—	—
4	—	—	—

6. What we are not funding this cycle

—

Checks requested: 0