← Back to the exerciseThe full worked example: Price freeze plus exclusivity: concede, counter or walk?
campo.
Field guide / Sales / UC-006 · SC-05
The result

"Concede both or walk"? Neither, on this evidence. The clause is written wider than the $38M our VP Sales describes.

A CRO at a 7,200-person automotive components maker decides how to answer his largest customer's demand for a price freeze plus exclusivity, with any AI chat his company allows.

The resultDaniel's decision page for the CEO. Fictional company.
NARROWGATE COMPONENTS · REVENUE
Corvane renewal: decision page
For: Marta Lindgren, CEO
From: Daniel Ash · Monday, January 18

MY RECOMMENDATION

Don't concede both and don't walk on the current evidence. On January 27, counter: hold 2026 net prices only with a steel and aluminium index, and agree no exclusivity until outside counsel has reviewed it. Corvane earns us $37.4M of gross margin after rebates; a freeze cuts that to about $30.7M in 2027, before the July wage rise (flat volumes, freeze at net price). · Inputs 2, 5, 6

WHAT'S REALLY ON THE TABLE

Exclusivity coversAt least $38M
Another $64M of tier-1 sales needs review; extra exposure unknown
Corvane margin today$37.4M (18.3%)
After rebates; $52.8M is at list
Under a freeze$30.7M (15.0%)
2027 cost of goods +4%; wages not yet costed
Fennick's 8%Unverified
One source: Corvane's buyer

WHAT WOULD CHANGE IT, AND WHO CONFIRMS

WOULD CHANGE IT
Fennick's offer confirmed like-for-like, or losing Corvane costing more than its margin
Then we'd give more on price. If Corvane won't drop exclusivity, walking comes back on the table. · Inputs 1, 2
BY MON JAN 25
Finance and legal: which of the $64M the clause covers, any overlap with the $38M, on the same revenue basis
Then finance sets a floor price and costs the July wage rise by product. · Inputs 3, 4, 5
LEGAL
Outside counsel asked now; no exclusivity agreed on January 27
· Input 6
Caught on review
1
The clause is written wider than the $38M: it also covers components and tier-1 suppliers
2
Margin is $37.4M after rebates, not $52.8M at list
3
Concede or walk is a false choice: a counter splits the two asks
UC-006 · SC-05 · Rev C
FOR
Chief Revenue Officer
SECTOR
Automotive components (largest customer renewal)
COMPANY
7,200 people · 4 plants · US and Mexico
TASK
Decide how to answer the largest customer's demand for a price freeze plus exclusivity
WRITTEN FOR
ChatGPT · Claude · Gemini · Copilot
AI TEST
Claude, Nov 15, 2026 · 7 of 7 found
Step 0 · The mess

What landed on Daniel's desk

A short letter from the customer, the account P&L, the clause itself, a note from the VP Sales, and notes from finance and legal. The VP Sales describes the clause. The clause says something else.

CUSTOMER LETTER[1]
Corvane Motors
“Prices frozen at 2026 levels for three years, and exclusivity.”
1
Two asks.
FINANCE[2]
Account P&L
“Gross margin at list: 24%.” Rebates 7% of list, on a separate line.
2
At list?
CONTRACT[3]
Exclusivity clause
“… or components for brake assemblies … or its tier-1 suppliers.”
3
Read it.
VP SALES[4]
The options
“Exclusivity only touches … $38M a year. Concede both or walk.”
4
Only?
FINANCE NOTE[5]
Costs
“2027 forecast: cost of goods for Corvane products up 4%.”
5
Frozen price.
LEGAL[6]
Exclusivity
“We need outside counsel's view … Not yet requested.”
6
Not yet.
You're not alone
4.7%
was the estimated average automotive supplier EBIT margin for 2024, according to Roland Berger and Lazard's study of 600 suppliers.
Roland Berger and Lazard, Global Automotive Supplier Study 2025, May 2025. Source
Steps 1–3 · How it works

Three steps, one chat

Here the AI works through a decision that arrives framed as two options. The prompt makes it test the framing before comparing anything.

1
Open a new chat

Use the AI tool your company approves: ChatGPT, Claude, Gemini or Copilot.

2
Paste the prompt and the inputs

Copy the prompt and the six fictional inputs from the appendix into one message.

3
Read the clause yourself

It says components and tier-1 suppliers. Ask which of the $64M of tier-1 sales it covers. Then decide what the CEO reads first.

The prompt, part by part
Copy-ready version in the appendix

I'm Chief Revenue Officer at an automotive components maker. Our largest customer wants a three-year price freeze plus exclusivity to renew its contract. Our CEO wants my decision page by Thursday, January 21, before I meet the customer on Wednesday, January 27. Today is Monday, January 18. Below are the customer's letter, the account P&L, their draft exclusivity clause, a note from our VP Sales, and notes from finance and legal.

1
Who you are, who decides, the meeting date and today's date.

Please help me work through the decision:

  1. Each claim in the customer's letter and our VP Sales's note: supported, contradicted or unverified, with the input that shows it.
  2. What the account really earns today and under a three-year freeze, with your calculation.
  3. What exclusivity would really cover, based on the clause as written.
  4. The options: concede, walk, and any option nobody has proposed, each with what it gains, what it risks and what's uncertain.
  5. My recommendation to the CEO, with what would change it and who must confirm what before January 27.
2
Each claim checked, what the account earns now and under a freeze, what the clause covers, the options including one nobody proposed, and a recommendation.

Check whether the choice we've been given is the real choice before you compare options. Read the clause as written, not as described. Check the arithmetic. Don't add facts that aren't in the inputs; if you assume something, say so.

3
The heavy lifting: check whether the choice you've been given is the real choice, and read the clause as written, not as described. It found all seven planted problems.
Where to paste
Every tool has one message box. Paste the prompt, add a blank line, paste the inputs, send. Contracts can be attached as files if your tool allows it.
Safe to try
Practise with the fictional inputs first. Real contracts, margins and customer letters are confidential: use only your company's approved tool.
The AI's first answer

What Claude actually wrote

Exactly as it came back, not edited. Read it the way Daniel did: check each point against the source it cites.

UNEDITED
Claude · fresh session, no other context · Nov 15, 2026 · one run

Corvane renewal: decision working notes

First, is "concede both or walk" the real choice?

No. The inputs show two separate asks, a price freeze and exclusivity. Nothing in them says Corvane will only accept both together; the letter states its position, and a buyer's opening position is not its walk-away. Carla's framing also leaves out the obvious third path: a counter-proposal that splits the two asks. And it rests on a misreading of the clause (see section 3). So the real decision for January 27 is what we offer, not whether we sign or leave.

There is also a timing gap. The award covers 2027–2029 and today is January 18, so 2027 has already started. The steel and aluminium contracts reset this month. The inputs don't say when the current contract ends or when new prices would take effect. Assumption: the 2027 cost increase applies to the whole award period.


1. Claims checked

Claim Source Verdict Evidence
Price frozen "at 2026 levels" Letter [1] Unverified (undefined) The letter doesn't say whether this means list price ($220.0M) or net price after the 7% rebates ($204.6M) [2]. That is a $15.4M a year difference in what we would be freezing.
"Fennick has offered us 8% below your price" Letter [1] Unverified No input supports it. "Your price" is also undefined. 8% below list = $202.4M, only about 1.1% below our current net of $204.6M. 8% below net = $188.2M. Very different threats. Nothing tells us whether Fennick's offer is like-for-like on scope, volume or terms. found
"Narrowgate keeps the award" Letter [1] Unverified / incomplete No volume, share or minimum commitment is stated. Exclusivity binds only us [3]; Corvane commits to nothing in return.
Corvane is "30% of our revenue" VP Sales [4] Roughly supported, basis unclear $220.0M list ÷ $700M = 31.4%. $204.6M net ÷ $700M = 29.2%. We don't know whether the $700M is on a list or net basis [2], so "about 30%" holds, but neither figure is exact. found
Exclusivity "only touches the two smaller automakers… $38M a year" VP Sales [4] Contradicted The clause [3] covers "any other vehicle manufacturer or its tier-1 suppliers" and "components for brake assemblies". Finance [5] shows $64M (net) of brake components sold to tier-1 suppliers in North America. That is outside Carla's figure. found
"The options are to concede both or walk" VP Sales [4] Contradicted / unsupported Nothing in the inputs rules out a counter-proposal or splitting the two asks. found
(Implied) we can agree exclusivity — Contradicted Legal [6]: we need outside counsel before agreeing to any exclusivity, and counsel has not yet been asked.

2. What the account really earns

FY26 today [2]

  • List revenue: $220.0M
  • Rebates and volume discounts: −$15.4M (7% of list, on a separate line)
  • Net revenue: $204.6M
  • Cost of goods: $220.0M − $52.8M = $167.2M
  • Reported gross margin: $52.8M (24%), but that is calculated on list and ignores the rebates
  • Gross margin after rebates: $52.8M − $15.4M = $37.4M, which is 18.3% of net revenue found

So the account earns about $37.4M of gross margin, not $52.8M. Gross margin is before overheads; the inputs don't show operating profit.

Under a three-year freeze at 2026 net price

Assumptions: flat volumes, rebates stay at 7%, and the freeze applies to net price.

  • 2027: materials cost +4% [5] gives COGS of $167.2M × 1.04 = $173.9M. Gross margin = $204.6M − $173.9M = $30.7M (15.0%), down $6.7M from FY26. found
  • On top of that, Silao wages rise 5% from July 1, 2027 [5]. This is not costed by product, so the real 2027 margin is below $30.7M by an unknown amount.
  • 2028 and 2029: materials contracts reset every January, but there is no forecast. Illustrative only: if materials rose another 4% each year, margin would be about $23.8M (11.6%) in 2028 and $16.5M (8.1%) in 2029, before labour. A: off the CEO page
    • Three-year total on that assumption: about $71.0M, against $112.2M if FY26 margin held. That is a gap of about $41M, plus labour.
  • If the freeze means list price with rebates continuing, the figures above hold. If Corvane expects rebates on top of a list freeze and new reductions, margin falls further.

If we matched Fennick (8% below)

  • 8% below list: net revenue $202.4M, 2027 margin about $28.5M before labour.
  • 8% below net: net revenue $188.2M, 2027 margin about $14.3M (7.6%) before labour.

3. What the exclusivity clause actually covers

As written [3]:

  • Products: brake assemblies and components for brake assemblies. This is not just finished assemblies.
  • Customers: any other vehicle manufacturer and its tier-1 suppliers. This is not just two automakers.
  • Geography: "in North America." It is ambiguous whether this means where we supply, where we produce, or where the customer is located.
  • Term: "during the term." The term is not defined in the clause.
  • No carve-outs: nothing for existing contracts or existing customers.
  • One-sided: it binds only Narrowgate. Corvane commits to no volume.

Revenue exposed, at least: $38M (direct automakers [4]) + $64M (tier-1 components [5]) = $102M a year. That is about half of Corvane's net revenue. D: total not yet known

Caveats:

  • Carla's $38M basis (list or net) is unknown.
  • We assume no overlap between the $38M and the $64M.
  • Any other brake business, now or future, with other manufacturers would also be caught. The inputs don't list it.
  • Margin on the $102M is unknown, so we can't yet say what it is worth in profit.
  • Legal flags a possible competition law issue because of the tier-1 coverage [6].

4. The options

A. Concede both (freeze plus clause as drafted)

  • Gains: keeps about $204.6M net revenue and about 30% of total revenue, if Corvane really does award.
  • Risks:
    • Margin falls from $37.4M toward about $30.7M in 2027, and lower after labour and later material resets.
    • $102M+ of other brake business would have to be given up or would put us in breach.
    • Possible competition law exposure, with counsel not consulted [6].
    • No volume commitment from Corvane in return.
  • Uncertain: freeze basis, the term, what "North America" means, margin on the $102M, Corvane's actual volumes.

B. Walk

  • Gains: keeps the $102M+ and freedom to sell. Avoids a shrinking-margin contract and the legal risk.
  • Risks:
    • Loses about $204.6M net revenue and $37.4M gross margin (FY26).
    • Fixed-cost absorption across the plant, and Silao, is unknown and could make the loss much larger than the gross margin figure.
  • Uncertain: whether Fennick's offer is real or comparable, and whether Corvane would actually switch.

C. Counter: split the asks (the option nobody proposed) B: no capacity promise

  • On price: offer to hold 2026 net price, with steel and aluminium indexation (the contracts reset every January [5]), or with a defined price path, in exchange for a volume or share commitment from Corvane.
  • On exclusivity: decline the clause as drafted. Offer at most a narrow, mutual commitment, for example dedicated capacity or priority supply for Corvane platforms. Make any exclusivity wording subject to outside counsel's review.
  • Ask Corvane what exclusivity is meant to protect (supply security? IP? capacity?) and what Fennick's 8% is measured against.
  • Gains: keeps the relationship and room to negotiate. Protects the $102M. Puts the cost increase back on the table. Avoids signing something legal hasn't cleared.
  • Risks: Corvane may refuse and go to Fennick, if that offer is real.
  • Uncertain: how firm Corvane's position is, and Fennick's basis.

D. Price concession only, no exclusivity This is a variant of C if Corvane pushes hard on price. Finance must first set a floor price below which the account no longer pays its way, including labour and the 2028–29 material resets.


5. Recommendation to the CEO

Don't concede both, and don't walk, on the current evidence. Go to the January 27 meeting with option C: found

  • Hold 2026 net price with materials indexation.
  • No exclusivity as drafted.
  • Ask Corvane for its volume commitment and the basis of Fennick's figure.
  • Set a finance-approved floor in advance.

The decision page should state plainly:

  • The clause covers at least $102M a year, not $38M.
  • The account earns $37.4M of gross margin after rebates, not $52.8M.
  • That margin falls to about $30.7M in 2027 under a freeze, before the uncosted wage increase.

What would change this:

  • Toward conceding more on price: Fennick's offer is confirmed as like-for-like and close to our net price; or losing Corvane would leave fixed costs that outweigh the margin given up.
  • Toward accepting some exclusivity: counsel confirms a narrowed clause is lawful, and the margin on the $102M is low.
  • Toward walking: finance shows the frozen price falls below the floor once labour and later material resets are costed; or Corvane refuses to drop the exclusivity clause.

Who must confirm what before January 27

  • Finance:
    • Whether "2026 levels" means list or net, and the basis of the $700M and the $38M.
    • Any overlap between the $38M and the $64M.
    • Margin on the $102M.
    • The Silao labour cost by product.
    • Fixed-cost absorption if Corvane leaves.
    • A floor price.
  • Legal and outside counsel: request the review now (not yet requested [6]). It is unlikely to be finished by January 27, so agree no exclusivity at that meeting; at most say the wording is "subject to counsel review." found
  • Carla Ruiz (VP Sales): the full list of customers, automakers and tier-1s, covered by the clause as written, and any other brake business.
  • CEO: whether we will risk losing about 30% of revenue rather than accept the clause as written. This needs to be decided by January 21. C: floor by Jan 25

Assumptions used: flat volumes; rebates stay at 7%; freeze applies to net price; no overlap between the $38M and the $64M. The 4% a year material increase for 2028–29 is illustrative only.

foundcaught a planted problemA–DDaniel changes it; see the next section
The human part

What Daniel changed, and why

Claude found every planted problem, read the clause as written and proposed the counter nobody had. Daniel's edits keep an assumption off the CEO's page, take out a promise no input supports, give finance a date, and stop short of a total the inputs don't establish.

What Claude found: 7 of 7 planted problems
One run
✓Margin at list
✓Clause wider than $38M
✓Fennick's 8%: one source
✓Costs rise under a freeze
✓Concede or walk is false
✓Counsel not yet asked
✓"30%" basis unclear
Four edits, one catch and three judgment calls
A
CLAUDE WROTE
“Illustrative only: if materials rose another 4% each year, margin would be about $23.8M (11.6%) in 2028 and $16.5M (8.1%) in 2029 ... a gap of about $41M, plus labour.”
DANIEL'S VERSION
Daniel keeps only 2027 on the CEO's page ($30.7M, before the wage rise) and says later years are unknown until finance forecasts them.
WHY · EVIDENCE
The AI labelled the 4% a year as its own assumption. On a CEO's page, a $41M figure reads as a forecast, however it's labelled.
B
CLAUDE WROTE
“Offer at most a narrow, mutual commitment, for example dedicated capacity or priority supply for Corvane platforms.”
DANIEL'S VERSION
Daniel's counter is price and exclusivity only. Capacity or priority supply isn't offered until the plants confirm they can give it.
WHY · PROMISE
No input says the plants have capacity to dedicate. A counter that promises it could be the next problem.
C
CLAUDE WROTE
“Finance: ... A floor price.” Listed among items to confirm “before January 27.”
DANIEL'S VERSION
By Monday, January 25, two days before the meeting, finance and legal confirm which tier-1 sales the clause covers, and finance sets the floor price and costs the wage rise.
WHY · DATE
A floor that arrives the morning of the meeting can't shape the counter.
D
CLAUDE WROTE
“Revenue exposed, at least: $38M (direct automakers) + $64M (tier-1 components) = $102M a year.”
DANIEL'S VERSION
Daniel writes that at least $38M is covered and that $64M of tier-1 sales needs review, and puts no total on it until finance and legal confirm which tier-1 sales it covers, any overlap, and the revenue basis.
WHY · EVIDENCE
The clause covers tier-1 suppliers of other vehicle manufacturers. Nothing says all $64M of tier-1 sales serves them, so $102M is an unverified sum of two revenue pools, not a confirmed exposure or an upper limit on all affected business.
What only you can do
Keep the AI's illustrations off the decision page, promise only what your company can deliver, and get the numbers in time to use them. The AI found the third option. You decide what you offer.
Your turn

Make it yours

Use it whenever a decision arrives framed as two choices: concede or walk, build or buy, hire or don't.

Your prompt: fill in the brackets

I'm [your role] at a [industry] company. [Who] wants [their terms]. [Who decides] wants my decision page by [date], before [the meeting], [date]. Today is [date]. Below are [the letter or request, the account numbers, the contract wording, our team's view, and finance and legal notes].

Please help me work through the decision: 1. Each claim: supported, contradicted or unverified, with the input that shows it. 2. What the account really earns today and under [their terms], with your calculation. 3. What [the clause] would really cover, based on the wording. 4. The options, including any nobody has proposed, each with what it gains, risks and leaves uncertain. 5. My recommendation, with what would change it and who must confirm what before [date].

Check whether the choice we've been given is the real choice before you compare options. Read the clause as written, not as described. Check the arithmetic. Don't add facts that aren't in the inputs; if you assume something, say so.

Five checks before you decide
1
Margins are after rebates and discounts.
2
The contract wording is read as written, not as someone described it.
3
Each threat has a source, and you know how many.
4
Rising costs are counted against any freeze.
5
At least one option is neither yes nor no.
Make it repeatable (optional)
Save the prompt once and reuse it whenever a decision comes framed as two choices: a custom GPT in ChatGPT, a project in Claude, a Gem in Gemini, or an agent in Copilot. Some companies switch these off; copy and paste always works.
With real data
Contracts, margins and customer letters are confidential. Use only the AI tool your company approves, and follow its rules. Legal reviews any exclusivity clause.
What this example shows
One run of this prompt in Claude found all seven problems planted in fictional inputs and proposed a counter nobody had raised.
What it doesn't show
That every tool, setting or run will do the same, how much time it saves you, or what a real company should offer a real customer. Narrowgate Components, Corvane Motors and their people are fictional.
Appendix

Copy-ready: the prompt and the six inputs

Copy the prompt, then all six inputs, into one message in a new chat. Everything here is fictional and safe to paste.

The prompt

I'm Chief Revenue Officer at an automotive components maker. Our largest customer wants a three-year price freeze plus exclusivity to renew its contract. Our CEO wants my decision page by Thursday, January 21, before I meet the customer on Wednesday, January 27. Today is Monday, January 18. Below are the customer's letter, the account P&L, their draft exclusivity clause, a note from our VP Sales, and notes from finance and legal.

Please help me work through the decision:

  1. Each claim in the customer's letter and our VP Sales's note: supported, contradicted or unverified, with the input that shows it.
  2. What the account really earns today and under a three-year freeze, with your calculation.
  3. What exclusivity would really cover, based on the clause as written.
  4. The options: concede, walk, and any option nobody has proposed, each with what it gains, what it risks and what's uncertain.
  5. My recommendation to the CEO, with what would change it and who must confirm what before January 27.

Check whether the choice we've been given is the real choice before you compare options. Read the clause as written, not as described. Check the arithmetic. Don't add facts that aren't in the inputs; if you assume something, say so.

[1] Letter from Greg Tanner, Head of Purchasing, Corvane Motors

For the 2027–2029 brake assembly award, we require two things: prices frozen at 2026 levels for three years, and exclusivity as set out in the attached clause. In return, Narrowgate keeps the award. Fennick has offered us 8% below your price.

[2] Corvane account P&L, FY26 (finance)
  • Revenue at list price: $220.0M.
  • Gross margin at list: 24% ($52.8M).
  • Rebates and volume discounts: 7% of list ($15.4M), booked on a separate line.
  • Narrowgate total revenue, FY26: $700M.
[3] Exclusivity clause (Corvane's draft)

"Supplier shall not supply brake assemblies, or components for brake assemblies, to any other vehicle manufacturer or its tier-1 suppliers in North America during the term."

[4] Note from Carla Ruiz, VP Sales, North America

"We can't lose Corvane: it's 30% of our revenue. Exclusivity only touches the two smaller automakers we sell brake assemblies to directly, $38M a year. The options are to concede both or walk."

[5] Finance note (January 15)
  • Revenue from brake components sold to tier-1 suppliers in North America, FY26: $64M (net).
  • Steel and aluminium contracts reset each January. 2027 forecast: cost of goods for Corvane products up 4%.
  • Silao labour agreement: wages up 5% from July 1, 2027. Not yet costed by product.
[6] Legal note (January 15)

An exclusivity clause that covers tier-1 suppliers may raise competition law questions. We need outside counsel's view before agreeing to any exclusivity. Not yet requested.