← Back to the guideThe full worked example: Split next year's number across regions, on a like-for-like basis
campo.
Field guide / Sales / UC-003 · SC-04
The result

"Three options"? None works as proposed. History counts 9 months of India, potential mixes two market measures, and capacity adds up to $910M.

A CRO at a 26,000-person industrial equipment maker recommends how to split next year's $900M across three regions, with any AI chat her company allows.

The resultNadia's page for the CFO. Fictional company.
SELVANE MACHINERY · REVENUE
FY27 regional split: recommendation
For: Martin Cole, CFO
From: Nadia Brennan · Tuesday, December 1

MY RECOMMENDATION

Don't adopt any of the three as proposed. Put on one basis, two methods land close together. I'll bring that range on Friday and pick a single split once sales ops explains the $13.0M per rep. · Inputs 1, 2, 3

THE THREE OPTIONS

A: historyIndia short
9 months of India against 12 for the others
B: potentialNot like-for-like
Latin America measured differently; NA below FY26
C: capacity$910M
Not $900M; 6 hires counted as full-year
PipelineUnverified
Each VP's own figure

THE RANGE, IF THE ASSUMPTIONS HOLD

RANGE
North America $531M to $561M · Latin America $199M to $213M · India $140M to $156M
Each end adds up to $900M and keeps every region above FY26. History on a 12-month basis, if India's 9 months annualise evenly. Capacity with hires counted from their start dates (67.75 full-year reps), scaled to $900M: that implies $13.28M per rep, against sales ops' unverified $13.0M, before any ramp. · Inputs 1, 2, 3
BY THURSDAY
Confirm by noon, Thursday, December 3
Sales ops: how $13.0M was calculated. Anjali: hire start dates and ramp. Finance: whether FY26 is closed, and India's January to March bookings. · Inputs 3, 4
Caught on review
1
Option A counts 9 months of India against 12 months for the other regions
2
Option B compares total market with serviceable market
3
Option C adds up to $910M, not the board's $900M
UC-003 · SC-04 · Rev B
FOR
Chief Revenue Officer
SECTOR
Industrial equipment
COMPANY
26,000 people · 18 sites · US, Brazil, India
TASK
Recommend how to split next year's bookings number across three regions
WRITTEN FOR
ChatGPT · Claude · Gemini · Copilot
AI TEST
Claude, Nov 14, 2026 · 8 of 8 found
Step 0 · The mess

What landed on Nadia's desk

A short email from the CFO with two conditions, sales ops' three options in one table, the data behind them, and a line from each regional VP. Each option looks reasonable on its own. Side by side, they're built on different foundations.

EMAIL · CFO[1]
FY27 split
“Must add up to exactly $900M … no region below its FY26 actual.”
1
Two conditions.
SALES OPS[2]
Three options
History 584 / 207 / 109 · Potential 450 / 169 / 281 · Capacity 520 / 208 / 182.
2
Add them up.
SALES OPS[3]
The data
India's FY26 bookings cover April to December … Latin America $0.9B is serviceable market.
3
Same basis?
REGIONAL VPS[4]
Three notes
“We'll have 14 reps. Option C is the only fair one.”
4
Each one's own case.
You're not alone
1 in 4
sales reps made quota last year, according to Salesforce's survey of U.S. sales professionals.
Salesforce, Trends in Sales Compensation Report, 500 U.S. sales professionals surveyed August to September 2025. Source
Steps 1–3 · How it works

Three steps, one chat

Here the AI compares options that each claim to be fair. The prompt makes it rebuild them on one basis first.

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 four fictional inputs from the appendix into one message.

3
Add up one option yourself

Add the three regional numbers in each option and compare them with the board's number. Then decide what you'll show on Friday.

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

I'm Chief Revenue Officer at an industrial equipment manufacturer with three sales regions: North America, Latin America and India. The board has approved next year's bookings number, and I have to recommend how to split it across the regions to our CFO and the three regional VPs on Friday, December 4. Today is Tuesday, December 1. Below are the CFO's email, the three options sales operations prepared, the data behind them, and notes from the regional VPs.

1
Who you are, who you're recommending to, the meeting date and today's date.

Please give me:

  1. A like-for-like table of the three options: for each, the time period, what it measures and what it counts, and the number for each region.
  2. Each option tested against the CFO's two conditions and against the data.
  3. What's missing that I'd need to decide well.
  4. My recommendation, including whether any option works as proposed, and what must be confirmed, by whom, before Friday.
  5. What I tell each regional VP, in two sentences each.
2
A like-for-like table, each option tested against the CFO's conditions, what's missing, a recommendation with who confirms what, and a line for each VP.

Put every option on the same time basis and the same definitions before you compare them, and say plainly if none fits as proposed. Check the arithmetic. Don't add facts that aren't in the inputs; if you assume something, say so.

3
The heavy lifting: put every option on the same time basis and definitions before comparing, and say plainly if none fits. It found all eight planted problems.
Where to paste
Every tool has one message box. Paste the prompt, add a blank line, paste the inputs, send. Planning spreadsheets can be attached as files if your tool allows it.
Safe to try
Practise with the fictional inputs first. Real targets and regional plans 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 Nadia did: check each point against the source it cites.

UNEDITED
Claude · fresh session, no other context · Nov 14, 2026 · one run

Recommendation in brief: None of the three options works as proposed. Option A meets both of the CFO's conditions, but it rests on only 9 months of India data. Option B breaks the floor for North America and measures Latin America differently. Option C adds up to $910M, not $900M. A corrected version of Option C is the best starting point, provided the items in section 4 are confirmed before Friday.

Assumptions I've made (please confirm):

  • The fiscal year runs January to December. I took this from "India's FY26 bookings cover April to December."
  • All figures are in USD on the same bookings definition.
  • North America and Latin America reps are in seat for the full year, because the inputs mention no hires for those regions.
  • The dates are consistent: Tuesday December 1 and Friday December 4 are both working days in 2026.

1. Like-for-like table

Time period What it measures What it counts NA LatAm India Total
A: history (as proposed) FY26. NA and LatAm are 12 months; India is 9 months (Apr–Dec) Share of FY26 bookings NA 480 / 740 = 64.9%, LatAm 170 / 740 = 23.0%, India 90 / 740 = 12.2% 584 207 109 900 ✓ found
B: potential (as proposed) 2025 study, not a bookings period Share of market potential NA $2.4B total market, India $1.5B total market, LatAm $0.9B serviceable market only (sum $4.8B) 450 169 281 900 ✓ found
C: capacity (as proposed) FY27 headcount, every rep counted as a full year Reps × $13.0M (FY26 company average) 40 / 16 / 14 reps, total 70 520 208 182 910 ✗ found

Arithmetic checks:

  • A reproduces correctly from the inputs (583.8 / 206.8 / 109.5). It sums to 900.
  • B reproduces correctly (50% / 18.75% / 31.25% of 900 = 450 / 168.75 / 281.25). It sums to 900.
  • C is wrong: 70 reps × $13.0M = $910M. It fails the CFO's first condition.

The same three options on one basis:

NA LatAm India Total How it's built
A, annualised 561 199 140 900 India's $90M for 9 months becomes $120M for 12 months, which is an assumption. New total base is $770M.
B, like-for-like can't be done can't be done can't be done There is no total-market figure for LatAm.
C, part-year reps, scaled to $900M 531 213 156 900 India counts as 8 + 3 × 9/12 + 3 × 6/12 = 11.75 full-year reps. Total is 67.75 reps, so $900M / 67.75 = $13.28M per rep.

For reference, Option C counted as 70 full reps and simply scaled to $900M gives 514 / 206 / 180. That fixes the total but still counts India's six new hires as full-year reps.

2. Each option against the CFO's two conditions and the data

Adds to exactly $900M? Every region at or above FY26? Against the data
A Yes Yes: 584 > 480, 207 > 170, 109 > 90 The India number rests on 9 months, so it understates India. Annualised, India's share is $140M, not $109M. Pipeline coverage: NA 1.9×, LatAm 1.5×, India 2.4×. NA would need $14.6M per rep, above the $13.0M company average.
B Yes No. NA gets 450, below its FY26 480. Tom is right. LatAm 169 < 170 by about $1M after rounding (168.75 before). It mixes total market (NA, India) with serviceable market (LatAm). That understates LatAm against the others, and Lucas is right. India at 281 is above its $260M pipeline (0.93× coverage) and more than 3× its FY26 (9-month) bookings. The study dates from 2025. found
C No: $910M Yes as proposed (520 / 208 / 182) It counts India's 6 new hires as full-year reps, though 3 start in April and 3 in July. It allows no ramp-up time. The $13.0M average can't be checked because FY26 rep counts aren't given, and it may be distorted by India's part year. Anjali's "only fair one" is fair only if part-year hires count as full-year reps. found

Corrected C (531 / 213 / 156):

  • It sums to $900M.
  • Every region is above FY26.
  • Pipeline coverage is NA 2.1×, LatAm 1.5×, India 1.7×.

Note on the floor: On December 1, FY26 can't be closed if the fiscal year is the calendar year. The "FY26 actuals" ($740M) must include November and December figures that are estimates or incomplete. The floor is therefore being tested against numbers that may still move.

3. What's missing

  • India January–March FY26: what the distributor booked (or sold) in those months. Without it, India's history is 9 months and can't be compared with the other regions.
  • Whether FY26 is final: are the FY26 figures closed actuals, year to date, or a forecast for December?
  • The $13.0M per rep: how sales ops calculated it, how many reps there were in FY26 by region, and productivity per rep by region. Without these we don't know whether one company-wide figure suits all three regions.
  • Ramp time for new hires. Also any hiring, attrition or vacancies planned in NA and LatAm.
  • Pipeline definitions: stage, expected close date inside FY27, and weighting, plus each region's historical conversion rate. The three pipeline figures ($1.1B / $310M / $260M) are each VP's own number and may not be built the same way. found
  • LatAm total market: without it, Option B can't be made like-for-like.
  • Quota rules: whether the CFO's "exactly $900M" allows over-assignment at region or rep level, and the basis of the $900M (currency, FX for LatAm and India, what counts as a booking).

4. Recommendation

Don't approve any option as proposed: found

  • A quietly penalises India for a 9-month year.
  • B breaks the floor for North America and mixes definitions.
  • C is $10M over the board number and counts part-year hires as full-year.

Proposed working basis for Friday: corrected capacity (C), with NA $531M, LatAm $213M and India $156M. A: bring the range

  • It meets both CFO conditions.
  • It reflects FY27 resources.
  • No region's pipeline coverage falls below 1.4×.
  • It lands close to annualised history (561 / 199 / 140), which is a reasonable cross-check.

Present it as conditional, not final, until these items are confirmed: C: a deadline

To confirm Who Why it matters
The $910M error, how the $13.0M was calculated, FY26 reps by region Sales operations The whole capacity method depends on it
Start dates of India's 6 hires and the ramp assumption Anjali Rao, with HR/recruiting Each part-year rep moves India by several $M. A ramp period would lower India further.
Whether FY26 is closed, and India's January–March distributor bookings Martin Cole / Finance Sets the floor and the true size of India's history
Pipeline on one definition (FY27 close dates, stage, weighting) Each regional VP, collected by sales ops Shows whether the coverage figures can be compared
Whether "exactly $900M" allows over-assignment below region level Martin Cole Affects how much buffer the regions can hold

If the $13.0M can't be explained by Friday, tell the CFO that, and bring corrected A and corrected C side by side as the range (NA 531–561, LatAm 199–213, India 140–156). Don't pick one figure on the strength of a number you can't explain.

5. What to tell each regional VP

  • Tom Reyes (NA): You're right that Option B would put you below this year, so it's off the table. I'm proposing about $531M based on your 40 reps, against a $1.1B pipeline, and I'll confirm once sales ops validates the per-rep figure and we agree a common pipeline definition. B: the basis, not a number
  • Lucas Ferreira (LatAm): You're right that Option B compared your serviceable market with everyone else's total market, so I'm not using it. I'm proposing about $213M based on your 16 reps, which is above your FY26 $170M and below your $310M pipeline, subject to sales ops confirming the per-rep figure.
  • Anjali Rao (India): Capacity is the right starting point, but six of your 14 reps start in April and July, so they count as part-year: 11.75 full-year reps, or about $156M. Please confirm the start dates and expected ramp time by Thursday, and I'll adjust if they're earlier.

These figures come from the inputs alone. Everything marked as an assumption or "to confirm" should be checked before you present it.

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

What Nadia changed, and why

Claude found every planted problem, rebuilt the options on one basis and said none works as proposed. Nadia's edits keep her recommendation within what the evidence supports and decide who hears what, when.

What Claude found: 8 of 8 planted problems
One run
✓India: 9 months in history
✓Potential not like-for-like
✓Hires counted as full-year
✓Capacity adds to $910M
✓Potential breaks the CFO's floor
✓India's potential number
✓Pipeline unverified
✓None fits as proposed
Three edits, all judgment calls
A
CLAUDE WROTE
“Proposed working basis for Friday: corrected capacity (C), with NA $531M, LatAm $213M and India $156M.”
NADIA'S VERSION
Nadia brings the range, from history on a 12-month basis to capacity counted from start dates, and picks a single split only after sales ops explains the $13.0M per rep.
WHY · DECISION
The AI said itself that the $13.0M can't be checked and there's no ramp allowance. One precise split on that basis claims more than the data supports.
B
CLAUDE WROTE
“Tom Reyes (NA): … I'm proposing about $531M based on your 40 reps.”
NADIA'S VERSION
Before Friday, each VP hears which correction affects their region and what Nadia needs from them. The numbers come after the CFO agrees the basis.
WHY · AUDIENCE
A number given on Tuesday becomes the floor each VP negotiates from, before the CFO has seen it.
C
CLAUDE WROTE
A table of five items to confirm, with no deadline except India's start dates “by Thursday”.
NADIA'S VERSION
Every item is due by noon on Thursday, December 3, so Nadia has the afternoon to settle the split before Friday.
WHY · DATE
Without a deadline, Friday's meeting reviews open questions instead of deciding.
What only you can do
Keep the recommendation as firm as the evidence, decide who hears what first, and set the deadline. The AI rebuilt the options. You decide how much to commit to on Friday.
Your turn

Make it yours

Use it whenever a number has to be split and every option is presented as fair: regional targets, budgets by division, headcount by site, capital by plant.

Your prompt: fill in the brackets

I'm [your role] at a [industry] company with [the units you're splitting across]. [Who] has approved [the total], and I have to recommend how to split it to [who] on [date]. Today is [date]. Below are [their email, the options, the data behind them, and notes from the people affected].

Please give me: 1. A like-for-like table of the options: for each, the time period, what it measures and what it counts, and the number for each [unit]. 2. Each option tested against [the conditions] and against the data. 3. What's missing that I'd need to decide well. 4. My recommendation, including whether any option works as proposed, and what must be confirmed, by whom, before [date]. 5. What I tell each [person affected], in two sentences each.

Put every option on the same time basis and the same definitions before you compare them, and say plainly if none fits as proposed. Check the arithmetic. Don't add facts that aren't in the inputs; if you assume something, say so.

Five checks before you recommend
1
Every option adds up to the approved total.
2
Every option uses the same period for every unit.
3
Market figures use the same definition everywhere.
4
New people are counted from their start dates.
5
Your recommendation is no firmer than its least certain number.
Make it repeatable (optional)
Save the prompt once and reuse it every planning cycle: 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
Targets and regional plans are confidential. Use only the AI tool your company approves, and follow its rules.
What this example shows
One run of this prompt in Claude found all eight problems planted in fictional inputs and said plainly that none of the options works as proposed.
What it doesn't show
That every tool, setting or run will do the same, how much time it saves you, or what the right split for a real company is. Selvane Machinery and its people are fictional.
Appendix

Copy-ready: the prompt and the four inputs

Copy the prompt, then all four 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 industrial equipment manufacturer with three sales regions: North America, Latin America and India. The board has approved next year's bookings number, and I have to recommend how to split it across the regions to our CFO and the three regional VPs on Friday, December 4. Today is Tuesday, December 1. Below are the CFO's email, the three options sales operations prepared, the data behind them, and notes from the regional VPs.

Please give me:

  1. A like-for-like table of the three options: for each, the time period, what it measures and what it counts, and the number for each region.
  2. Each option tested against the CFO's two conditions and against the data.
  3. What's missing that I'd need to decide well.
  4. My recommendation, including whether any option works as proposed, and what must be confirmed, by whom, before Friday.
  5. What I tell each regional VP, in two sentences each.

Put every option on the same time basis and the same definitions before you compare them, and say plainly if none fits as proposed. Check the arithmetic. Don't add facts that aren't in the inputs; if you assume something, say so.

[1] Email from Martin Cole, CFO

Nadia, the board approved $900M in FY27 bookings. The three regional numbers must add up to exactly $900M, and I'd rather no region gets a number below its FY26 actual. Sales ops has put three options together. Please bring your recommendation Friday.

[2] Three ways to split $900M (sales operations)
$M North America Latin America India Basis
Option A: history 584 207 109 Share of FY26 bookings
Option B: potential 450 169 281 Share of market potential
Option C: capacity 520 208 182 FY27 reps x $13.0M per rep (FY26 company average)
[3] Sales operations data
  • FY26 bookings: North America $480M, Latin America $170M, India $90M. Total $740M.
  • India moved from a distributor to our own sales team on April 1, 2026. Its FY26 bookings cover April to December.
  • Market potential (strategy team study, 2025): North America $2.4B and India $1.5B are total addressable market. Latin America $0.9B is serviceable market, the segments we sell into today; no total market figure was done for Latin America.
  • FY27 quota-carrying reps: North America 40, Latin America 16, India 14. India's 14 includes 6 new hires: 3 start in April, 3 in July.
[4] Notes from the regional VPs
  • Tom Reyes, North America: "Option B would give us less than we booked this year. Our FY27 pipeline is $1.1B."
  • Lucas Ferreira, Latin America: "Option B measures our market differently from everyone else's. Our pipeline is $310M."
  • Anjali Rao, India: "We'll have 14 reps. Option C is the only fair one. Our pipeline is $260M."