← Back to the exerciseThe full worked example: Build your first 90 days as CRO from the data, not the story you inherit
campo.
Field guide / Sales / UC-004 · SC-03
The result

"The pipeline is fine at 3x"? About 2x once stale and duplicate deals come out. And territories have to be set by February 26, not March 15.

A new CRO at a 45,000-person building services company builds her 90-day plan for the CEO, checking her predecessor's story against the data, with any AI chat her company allows.

The resultAmara's plan for the CEO, due January 22. Fictional company.
CINDREL BUILDING SERVICES · REVENUE
First 90 days: plan
For: Daniel Hart, CEO
From: Amara Osei · Friday, January 22

THE PLAN IN ONE LINE

Protect Q4, get the facts behind the misses, and set FY28 territories and quotas by the February 26 payroll deadline. I won't name a single cause for the misses until the data shows one. · Inputs 1, 3, 4, 5

WHAT I INHERITED

PipelineAbout 2x
$1.57B after $830M stale or duplicated
Churn reasons252 of 412
No reason recorded; SightLine 21 of 160
Attrition58 of 310
In 9 months; FY26 was 12%
TerritoriesFeb 26
Quotas by territory final for payroll

WHAT YOU'LL SEE, AND WHEN

BY FEB 5
Cleaned FY28 coverage, from sales ops, for you and the CFO
Any of the $1.57B that closes this quarter counts toward FY27, so FY28 coverage may be lower than 2x. · Inputs 3, 5
BY FEB 26
FY28 territories and quotas final
If the territory data isn't clean by February 5, FY28 starts with the current territory structure. We'll prioritize coverage for vacant territories and consider a fuller redesign after FY27 closes. · Inputs 1, 5
NOT YET
A single cause for churn or for the misses
First, customer success collects reasons for the largest of the 252 unexplained non-renewals. · Input 4
Caught on review
1
"3x coverage" includes $520M stale and $310M duplicated; clean, it's about 2x
2
SightLine is 21 of the 160 recorded exit reasons; 252 have none
3
Quotas by territory must be final February 26, before a March 15 announcement
UC-004 · SC-03 · Rev B
FOR
Chief Revenue Officer (new)
SECTOR
Building services (lifts, HVAC)
COMPANY
45,000 people · 200 branches · 20 countries
TASK
Build a 90-day plan for the CEO from the data, not the inherited story
WRITTEN FOR
ChatGPT · Claude · Gemini · Copilot
AI TEST
Claude, Nov 14, 2026 · 8 of 8 found
Step 0 · The mess

What landed on Amara's desk

A welcome email from the CEO that already contains the explanation, three lines from her predecessor's deck, and four short notes from sales ops, customer success, HR and finance. The story is tidy. The data underneath it is not.

EMAIL · CEO[1]
Welcome
“The pipeline is fine at $2.4B, three times next year's $800M target, so this is execution.”
1
The story.
PREDECESSOR'S DECK[2]
Turnaround
“Churn is up because of SightLine product issues … attrition has stabilised at 12%.”
2
Three claims.
SALES OPS[3]
Pipeline export
$520M older than 18 months … $310M duplicated.
3
What's real?
CUSTOMER SUCCESS[4]
Churn
Exit reason recorded for 160 of the 412.
4
Who counts?
HR AND FINANCE[5]
Deadlines
Quotas by territory final by February 26 … FY27 closes March 31.
5
Check the dates.
You're not alone
35%
of sales professionals in Salesforce's State of Sales survey said they completely trust the accuracy of their organization's data.
Salesforce, sixth State of Sales report, 5,500 sales professionals in 27 countries, 2024. Source
Steps 1–3 · How it works

Three steps, one chat

Here the AI helps you plan when someone has already explained the problem. The prompt tells it to treat that explanation as claims to check.

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 five fictional inputs from the appendix into one message.

3
Check one date yourself

Put the CEO's announcement date next to the payroll deadline. Then decide which first moves the CEO needs to see.

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

I've just joined an industrial building services company as Chief Revenue Officer. The sales organisation has missed plan two years running, and the CEO wants my 90-day plan by Friday, January 22. I started on Monday, January 11, and today is Wednesday, January 13. Our fiscal year runs April to March. Below are the CEO's email, my predecessor's turnaround deck, and notes from sales operations, customer success, HR and finance.

1
Who you are, when you started, when the plan is due, today's date and the fiscal year.

Please give me:

  1. What I've inherited: each claim in the CEO's email and the deck, with what the data shows: holds, doesn't hold, or unknown.
  2. Days 1 to 30: what I need to find out first, with the evidence that makes each item a priority.
  3. Days 31 to 90: the decisions and their dates, checked against our fiscal year-end and any other deadlines in the inputs.
  4. What I won't commit to yet, and what would let me.
  5. A short note to the CEO covering the plan and anything in his request I'd change.
2
What holds in the story you inherited, the first 30 days with evidence, the decisions to day 90 with dates, what you won't commit to yet, and a note to the CEO.

Treat everything my predecessor wrote as a claim to check, not a fact, and give the evidence for every first move. Don't swap one unsupported explanation for another. Check the arithmetic and the dates. Don't add facts that aren't in the inputs; if you assume something, say so.

3
The heavy lifting: treat everything your predecessor wrote as a claim to check, and don't swap one unsupported explanation for another. 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. Exports can be attached as files if your tool allows it.
Safe to try
Practise with the fictional inputs first. Real pipeline, churn and people data is 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 Amara did: check each point against the source it cites.

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

90-day plan inputs: CRO, industrial building services

Assumption about the year: the inputs don't state one. Kevin's deck is dated September 2026 and FY28 starts April 1, so I've taken "now" to be January 2027. On that basis every date in the inputs falls on the weekday given: Mon Jan 11, Wed Jan 13, Fri Jan 22, Fri Feb 26, Mon Mar 15, Wed Mar 31, Thu Apr 1. Counting Jan 11 as day 1, day 30 is Tue Feb 9 and day 90 is Sat Apr 10.


1. What I've inherited: claims and what the data shows

Claim (source) What the data shows Verdict
"Pipeline is fine at $2.4B, three times the $800M target" (CEO [1], deck [2]) The total of $2.4B is correct, and $2.4B is 3.0 times $800M. But $520M is stale (older than 18 months, no activity in 6 months) and $310M is duplicated renewals. Sales ops says the two don't overlap [3]. Removing both takes out $830M, 34.6%, which leaves $1.57B, or 1.96 times $800M. Doesn't hold. Coverage is about 2x on the export's own numbers, not 3x. found
Same pipeline counted toward FY28 The export is "total open pipeline". It doesn't say how much is due to close in FY27 Q4 (January to March), which is "usually our biggest bookings quarter" [5]. Whatever closes before March 31 counts toward FY27, not FY28. It also doesn't say whether the $800M includes renewals, which make up part of the pipeline. Unknown. FY28 coverage could be below 1.96x.
"Pipeline is healthy, so this is execution" (CEO [1]) This conclusion rests on the 3x figure, which doesn't hold. Doesn't hold as reasoned. Execution may still be part of the problem, but the data doesn't show it. found
"Service contract churn is up" (deck [2]) FY26 was 6.1% over 12 months. FY27 to date is 412 of 5,150, which is 8.0% (checked), over 9 months. FY27 has already passed the whole of FY26 with a quarter still to go. Caveat: I don't know whether 5,150 means contracts due for renewal or the whole contract base, or whether FY26 used the same base. Holds, provided both years use the same denominator (customer success to confirm). found
"…because of SightLine product issues" (deck [2], CEO [1]) SightLine is the reason in 21 of 160 recorded exits (13.1%) and 21 of 412 non-renewals overall (5.1%). The recorded reasons are price 61 (38%), response time 48 (30%), competitor bundle 30 (19%) and SightLine 21 (13%). Those add up to 160. 252 of 412 (61%) have no reason recorded. Doesn't hold on recorded data. Price isn't proven to be the cause either: 61% of exits have no reason, so the cause is unknown. found
"Sales attrition has stabilised at 12%" (deck [2]) 12% was the FY26 figure. In FY27 to date, 58 of 310 sellers have left, which is 18.7% in 9 months. The base isn't defined (start, average or current headcount), but 58 leavers would only be 12% of a base of 483 or more, far above 310. Doesn't hold. Attrition is higher than FY26, and the year isn't over. found

Other things I noticed:

  • The $310M duplication comes from 214 service renewals entered by both the service team and sales. That points to a process and ownership gap: nobody clearly owns renewals in the CRM. It isn't a seller performance issue.
  • 48 exits were for response time. That's service delivery. It isn't the SightLine product and it isn't sales. I'm noting where it sits, not claiming it's the cause.

2. Days 1 to 30 (Jan 11 to Feb 9): what to find out first C: three for the CEO

Items 1 to 3 are ordered by deadline: they feed the February 26 comp deadline and protect Q4.

  1. What FY28 coverage really is.

    • Evidence: $830M of $2.4B (34.6%) is stale or duplicated [3]. Coverage is 1.96x, not 3x.
    • Find out: whether sales ops agrees to remove both amounts, and how much of the remaining $1.57B is expected to close by March 31 (FY27) versus in FY28. Also whether the $800M includes renewals, and on what value basis.
    • Needed by: about Feb 5, so it can feed quotas.
  2. The Feb 26 comp and quota deadline: what's already in motion.

    • Evidence: "The FY28 sales compensation plan, including quotas by territory, must be final by Friday, February 26 for payroll set-up" [5].
    • Find out: who owns the plan, what draft exists, what territory structure the draft assumes, and the last date finance can accept territory changes.
  3. What Q4 bookings need to be protected. found

    • Evidence: Q4 is usually the biggest bookings quarter and FY27 closes March 31 [5].
    • Find out: the committed Q4 forecast, the deals in flight and which sellers own them. Anything I change between now and March 31 has to be weighed against this.
  4. Why sellers are leaving.

    • Evidence: 58 of 310 left in 9 months (18.7%) against 12% in FY26 [5].
    • Find out: the denominator HR uses, and which territories, tenure and performance levels the leavers came from (regretted or not). Also any exit reasons HR holds, and open requisitions. This matters directly for a territory redesign, which changes seller books.
  5. Why customers aren't renewing.

    • Evidence: 252 of 412 non-renewals (61%) have no reason [4]. SightLine accounts for 21 of the 160 recorded.
    • Find out: reasons for the 252, starting with the largest contracts. Whether FY26's 6.1% used the same base as the 5,150. Whether the response-time exits cluster in particular regions.
    • Ask, don't assume: no fix to SightLine, pricing or service until this is done.
  6. Who owns renewals in the CRM.

    • Evidence: 214 renewals were entered twice, by the service team and by sales [3].
    • Find out: who should own each renewal, and how the duplicates affect forecasting and possibly comp credit. Comp credit is an open question, not a finding.

3. Days 31 to 90 (Feb 10 to Apr 10): decisions and dates

Date Decision Checked against
Fri Feb 12 (day 33) What scope of territory change for FY28: full redesign, targeted changes (for example, vacant territories), or current territories with clean data. Has to come two weeks before Feb 26 so finance can set quotas by territory. A full redesign may not be possible in this window. A: decide the default
By Fri Feb 19 (day 40) Whether $800M stays as the FY28 target, based on the cleaned FY28 coverage from item 1. This is a decision for me, the CEO and the CFO. Quotas are set from the target, so this has to come before Feb 26.
Fri Feb 26 (day 47) Final FY28 comp plan and quotas by territory [5]. A fixed payroll deadline. This is the real territory deadline, not March 15. found
Early to mid-March (date to be set with HR) When and how sellers hear their FY28 territory and quota. This needs care with Q4 deals in flight, for example deciding who gets credit for deals that change hands. Territories must be set by Feb 26 anyway. Announcing during the biggest quarter, with attrition already at 18.7%, carries a risk (my judgment, not a finding) of distraction or more departures. Seller notice periods are unknown: confirm with HR.
By Wed Mar 31 (day 80) Pipeline rules live for FY28: stale-deal rules and single ownership of renewals. Stops the FY28 pipeline starting with the same $830M problem.
By Wed Mar 31 First churn actions, only where the reasons now known support them. They may involve service response time, pricing or product, and each needs its owner's agreement. Depends on item 5. If reasons are still missing for most exits, the decision is to keep investigating, not to act.
Thu Apr 1 (day 81) FY28 starts.
Fri Apr 9 (day 89, the last working day before day 90) Review with the CEO: FY27 actual against forecast, FY28 coverage, churn reasons and attrition trend. Day 90 is Saturday Apr 10.

4. What I won't commit to yet, and what would let me

  • That churn is a SightLine problem. First I need reasons for most of the 252 unexplained exits, and confirmation that FY26 and FY27 use the same denominator.
  • Any other single cause of churn, price included. Same condition: price is only 61 of 412.
  • Delivering $800M in FY28. First I need cleaned coverage split between FY27 Q4 and FY28, and the basis of the target (whether it includes renewals).
  • "This is execution." First I need clean pipeline, conversion rates by stage and the attrition breakdown. None of these are in the inputs yet.
  • A full territory redesign announced March 15. First I need to confirm the Feb 26 sequence with finance, check that the territory data is good enough, and agree with HR how to handle Q4 deals in flight.
  • An attrition target. First I need HR's denominator and exit reasons.

5. Note to the CEO

Subject: 90-day plan: outline and one change to the timing

Daniel,

Thank you for the welcome. You'll have the full plan by Friday, January 22. Here is the shape of it, plus one change I'd like to make to your request.

What I've found so far, from the January exports: - Pipeline: $520M is stale and $310M is duplicated renewals. Excluding both leaves $1.57B, about 2x the $800M target, not 3x. Some of that will close this quarter and count toward FY27, so FY28 coverage may be lower still. Sales ops and I are confirming the cleaned figure by early February. - Churn: it is up, from 6.1% in FY26 to 8.0% in the first nine months of FY27 (confirming the bases match). SightLine is the reason in 21 of the 160 exits with a recorded reason, but 252 of 412 have no reason at all. I won't name a cause until we have one. - Seller attrition: 58 of 310 sellers have left since April, 18.7%, against 12% last year. It hasn't stabilised.

The change I'd suggest: the FY28 comp plan, including quotas by territory, has to be final by Friday, February 26 for payroll. That means territories have to be decided by then, not announced on March 15. I'd decide the scope of territory change by February 12, set quotas by February 26, and agree with HR when sellers hear, in a way that protects Q4 deals.

One question for you and finance: if cleaned coverage confirms about 2x, I'd like to review the $800M with you before quotas lock on February 26. B: show, don't ask

Days 1 to 30: clean the pipeline, protect Q4, understand seller departures and the unexplained churn, and fix renewal ownership in the CRM.

Amara


Arithmetic I checked:

  • 520 + 310 = 830, which is 34.6% of 2,400. 2,400 − 830 = 1,570, and 1,570 ÷ 800 = 1.96.
  • 412 ÷ 5,150 = 8.0%.
  • 61 + 48 + 30 + 21 = 160, and 160 + 252 = 412. 21 ÷ 160 = 13.1%; 21 ÷ 412 = 5.1%.
  • 58 ÷ 310 = 18.7%. 58 ÷ 0.12 = 483.

What I didn't do: I didn't annualise churn or attrition. Renewal and departure timing within a year isn't given, so a full-year projection would be an assumption.

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

What Amara changed, and why

Claude found every planted problem and refused to swap one explanation for another. Amara's edits are about what a new CRO commits to, what she asks for in week one, and what the CEO needs to see.

What Claude found: 8 of 8 planted problems
One run
✓SightLine blamed without evidence
✓Stale pipeline
✓Duplicated pipeline
✓“3x, so it's execution”
✓Attrition hasn't stabilised
✓Churn periods differ
✓Feb 26 before March 15
✓Protect Q4
Three edits, all judgment calls
A
CLAUDE WROTE
“Fri Feb 12: What scope of territory change for FY28: full redesign, targeted changes (for example, vacant territories), or current territories with clean data.”
AMARA'S VERSION
Amara sets the default now: if the territory data isn't clean by February 5, FY28 starts with the current territory structure, coverage for vacant territories comes first, and a fuller redesign is considered after FY27 closes.
WHY · DECISION
Sellers are leaving faster than last year and Q4 is the biggest quarter. A default decided now keeps the February 26 deadline safe if the data isn't ready.
B
CLAUDE WROTE
“One question for you and finance: if cleaned coverage confirms about 2x, I'd like to review the $800M with you before quotas lock on February 26.”
AMARA'S VERSION
Amara commits to showing the CEO and CFO the cleaned coverage by February 5, and leaves the target out of her week-one note.
WHY · RELATIONSHIP
A new CRO asking to revisit the board's number in her first week reads as lowering the bar. The cleaned figure lets the CEO and CFO see the gap for themselves.
C
CLAUDE WROTE
Six first-month items, from coverage and the comp deadline to churn reasons and renewal ownership.
AMARA'S VERSION
The CEO sees three, each with an owner and a date: cleaned coverage from sales ops, the territory default, and churn reasons from customer success. The rest stay on Amara's own list.
WHY · LENGTH
The CEO needs what decides the February dates. Her working list is hers to run.
What only you can do
Decide the default before the deadline, show the number before you ask about it, and give the CEO what he needs to decide. The AI checked the story. You choose what a new CRO commits to.
Your turn

Make it yours

Use it whenever you step into a role and an explanation of the problem is already waiting for you: a new region, a turnaround, an acquired business, a function that missed its goals.

Your prompt: fill in the brackets

I've just joined [the company or function] as [your role]. [What went wrong], and [who] wants my [length]-day plan by [date]. I started on [date], and today is [date]. Below are [their brief, my predecessor's material, and the data and deadlines].

Please give me: 1. What I've inherited: each claim, with what the data shows: holds, doesn't hold, or unknown. 2. Days 1 to 30: what I need to find out first, with the evidence that makes each item a priority. 3. Days 31 to [end]: the decisions and their dates, checked against [the deadlines]. 4. What I won't commit to yet, and what would let me. 5. A short note to [who] covering the plan and anything in their request I'd change.

Treat everything my predecessor wrote as a claim to check, not a fact, and give the evidence for every first move. Don't swap one unsupported explanation for another. Check the arithmetic and the dates. Don't add facts that aren't in the inputs; if you assume something, say so.

Five checks before you send the plan
1
Every inherited claim has a verdict and a source.
2
The pipeline is counted without stale or duplicate deals.
3
Rates you compare cover the same length of time.
4
Your dates sit before the deadlines they feed.
5
You haven't replaced the old explanation with a new one.
Make it repeatable (optional)
Save the prompt once and reuse it whenever you take on something new: 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
Pipeline, churn and people data 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 didn't swap the inherited explanation for another one.
What it doesn't show
That every tool, setting or run will do the same, how much time it saves you, or why a real sales organisation missed its plan. Cindrel Building Services and its people are fictional.
Appendix

Copy-ready: the prompt and the five inputs

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

The prompt

I've just joined an industrial building services company as Chief Revenue Officer. The sales organisation has missed plan two years running, and the CEO wants my 90-day plan by Friday, January 22. I started on Monday, January 11, and today is Wednesday, January 13. Our fiscal year runs April to March. Below are the CEO's email, my predecessor's turnaround deck, and notes from sales operations, customer success, HR and finance.

Please give me:

  1. What I've inherited: each claim in the CEO's email and the deck, with what the data shows: holds, doesn't hold, or unknown.
  2. Days 1 to 30: what I need to find out first, with the evidence that makes each item a priority.
  3. Days 31 to 90: the decisions and their dates, checked against our fiscal year-end and any other deadlines in the inputs.
  4. What I won't commit to yet, and what would let me.
  5. A short note to the CEO covering the plan and anything in his request I'd change.

Treat everything my predecessor wrote as a claim to check, not a fact, and give the evidence for every first move. Don't swap one unsupported explanation for another. Check the arithmetic and the dates. Don't add facts that aren't in the inputs; if you assume something, say so.

[1] Email from Daniel Hart, CEO

Amara, welcome. We've missed plan two years running. Kevin's view was that churn is a product problem with SightLine, our connected monitoring service, and the pipeline is fine at $2.4B, three times next year's $800M target, so this is execution. I'd like your 90-day plan by Friday, January 22, including a territory redesign announced by March 15 so FY28 starts clean on April 1.

[2] From the previous CRO's turnaround deck (Kevin Marsh, September 2026)
  • "Service contract churn is up because of SightLine product issues."
  • "Pipeline: $2.4B, 3x coverage of FY28 target. Healthy."
  • "Sales attrition has stabilised at 12%."
[3] Sales operations: pipeline export (January 8)
  • Total open pipeline: $2.4B.
  • $520M is in opportunities older than 18 months with no activity in the last 6 months.
  • $310M is duplicated: 214 service renewals were entered once by the service team and again by sales. The stale and duplicate amounts don't overlap.
[4] Customer success: service contract churn (January 7)
  • FY26 (12 months): 6.1% of service contracts not renewed.
  • FY27 to date (April to December): 412 of 5,150 contracts not renewed, 8.0%.
  • Exit reason recorded for 160 of the 412: price 61, response time 48, moved to a competitor's bundle 30, SightLine issues 21. No reason recorded for the other 252.
[5] HR and finance notes
  • Sales attrition: 12% in FY26. In FY27 to date (April to December), 58 of 310 sellers have left.
  • The FY28 sales compensation plan, including quotas by territory, must be final by Friday, February 26 for payroll set-up.
  • FY27 closes Wednesday, March 31. Q4 (January to March) is usually our biggest bookings quarter.