← Back to the guideThe full worked example: Turn a month of plant updates into a one-page CHRO update
campo.
Field guide / L&D / UC-001 · SC-01
The result

7 updates in. 1 decision page out.

An L&D director at a nine-plant manufacturer turns a month of plant emails, exports and slides into three decisions, with any AI chat her company allows.

The resultWhat Dana sends her CHRO. Fictional company.
HALDEN INDUSTRIES
L&D monthly update · November 2026
For: Chief HR Officer
People review, Nov 24 · From: Dana Okafor

NEEDS A DECISION

URGENT
Monterrey: 14 forklift operators are working on expired certifications
Certifications lapsed Oct 31; recertification is booked for Dec 9. Decide: plant GM and EHS limit lift-truck work to certified operators until then. · Update 2
THIS MONTH
Cedar Falls: lockout/tagout refresher is 48 people short
212 of 260 done. The night shift has no date, only "after the holiday shutdown." Decide: fund night-shift coverage so the refresher finishes first. · Update 1
Q1 BUDGET
Supervisor onboarding is a pilot, not a launch
Reported as launched; it is a 12-person pilot starting Dec 2 at Joliet. Decide: include the company-wide rollout in the Q1 budget review. · Update 5

SAFETY-CRITICAL TRAINING BY SITELMS counts events, not people: asked for headcounts · Updates 1–3

MonterreyAt risk
Certs lapsed
Cedar FallsBehind
212 of 260
DaytonUnknown
Counts retakes
SaltilloUnknown
No data yet
QuerétaroUnknown
No data yet
GranbyUnverified
100% of events
HamiltonUnverified
98% of events
GreenvilleUnverified
95% of events
JolietUnverified
95% of events

WATCH

Emerging Leaders, 4.7 / 5: only 19 of 46 replied. Not yet a case for doubling. · Update 4
Greenville order desk: complaints began with the Sept 8 system change. Checking with IT first. · Update 7

HANDLED

Finance's spend per head (due Nov 20) mixes CAD, MXN and USD; we convert at one stated rate. · Update 6
Caught on review
1
"LOTO refresher: complete" is 212 of 260 done
2
Dayton's "140%" counts retakes, not people
3
"Onboarding launched" is a 12-person pilot in December
UC-001 · SC-01 · Rev C
FOR
Director / VP, L&D
SECTOR
Industrial manufacturing
COMPANY
11,500 people · 9 plants
TASK
Monthly CHRO update
WRITTEN FOR
ChatGPT · Claude · Gemini · Copilot
AI TEST
ChatGPT, Nov 12, 2026 · 8 of 8 found
Step 0 · The mess

What landed in Dana's inbox

Seven updates in one week, in seven formats. Each sounds fine on its own. Read closely, and most of them say more, or less, than the details support.

EMAIL · CEDAR FALLS[1]
LOTO refresher — complete
“Day and swing crews are done, 212 people through. Night shift we'll pick up… probably after the holiday shutdown.”
1
Says complete. It's 212 of 260.
TEAMS · MONTERREY[2]
quick one
“14 of our operators' certs expired Oct 31. Ops says they're still running the shift as normal.”
2
Casual tone, serious safety exposure.
LMS EXPORT · 9 SITES[3]
Safety-critical training, November
Completions = completion events in period. Dayton 434 of 310 assigned: 140%. Saltillo and Querétaro blank.
3
Counts events, not people. Two sites blank.
EMAIL · LEADERSHIP PROGRAM[4]
Cohort 3 wrap-up
“Average rating 4.7 out of 5! … I'd like to propose doubling the cohort size.”
4
Only 19 of 46 people answered.
SLIDE 7 · HR OPS DECK[5]
✔ New supervisor onboarding launched
Speaker notes: “pilot cohort starts Dec 2 at Joliet… rollout pending Q1 budget.”
5
“Launched” is a 12-person pilot.
EMAIL · FINANCE[6]
Training spend per head — need by Fri
“Hamilton 412 · Monterrey 6,900 · Dayton 340 …”
6
Three currencies in one list.
EMAIL · GREENVILLE[7]
Need customer service training ASAP
“Ever since the new order-entry system went live on Sept 8 the desk has been re-keying orders.”
7
Sounds like training. Started with a system change.
You're not alone
79%
of HR and L&D leaders feel pressure to show learning's business impact. A third say manual reporting gets in the way.
Go1, The Workforce Enablement Era, 2026, as reported by HRD. Sample size not published. Source
Steps 1–3 · How it works

Three steps, one chat

No special software and no setup. The prompt below does the heavy lifting, and each part of it has a job.

1
Open a new chat

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

2
Paste the prompt and the updates

Copy the prompt and all seven fictional updates from the appendix into one message.

3
Read it like an editor

Check each point against the update it cites. Then decide what your CHRO sees.

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

I lead Learning & Development at a manufacturing company. Below are seven updates I received this month from plants, Finance and my own team. I need to send my CHRO a one-page update before our monthly people review.

1
Who you are and who it's for. The AI writes differently for a CHRO than for your team.

Please read all seven carefully and give me:

  1. Needs a decision — up to three items where the CHRO or another senior leader has to decide or act. For each: what's happening, why it matters, who should decide, and which update it came from.
  2. Handled — for information — items my team is dealing with that the CHRO should know about.
  3. Watch — things that look important but aren't decision-ready yet.
  4. Site status — one line per plant on safety-critical training: on track, behind, or unknown.
2
Name the sections you want. A clear shape stops the AI writing an essay.

Be skeptical. Where a subject line, headline or summary says more than the details underneath support, go with the details and point out the difference. Don't add numbers together unless they measure the same thing in the same unit. If something is missing, say it's missing rather than guessing. Tell me which update number each point comes from.

3
The sentence that does the most work: check headlines against details. Then guard rails: no mixing units, no guessing, and a source for every point.

Keep it to one page, in plain language a busy executive can read in two minutes.

4
Set the length. ChatGPT still ran long.
Where to paste
Every tool has one message box. Paste the prompt, add a blank line, paste the updates, send. Attaching the updates as a file also works if your tool allows it.
Safe to try
Practise with the fictional updates first. With real reports, use only your company's approved tool and follow its rules on employee data.
The AI's first answer

What ChatGPT actually wrote

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

UNEDITED
ChatGPT · new chat · thinking: very high · Nov 12, 2026 · one run
memory may have been on

Halden Industries | November L&D Update

To: CHRO | People Review: November 24, 2026

1. Needs a decision

1. Monterrey — Expired forklift certifications [2] found

  • Situation: 14 operators' certifications expired October 31, but they are reportedly still operating forklifts. Recertification is scheduled for December 9.
  • Why it matters: Immediate safety and compliance exposure. Experience does not replace current certification.
  • Decision: Monterrey Plant Manager and EHS leadership must immediately address operator assignments, provide qualified coverage and pursue earlier recertification.

2. Cedar Falls — Incomplete safety training [1, 3] found

  • Situation: Only 212 of 260 employees have completed lockout/tagout refreshers. Night-shift training is deferred, possibly until after the holiday shutdown.
  • Why it matters: 48 employees remain outstanding. The email claims completion, but the details contradict this.
  • Decision: Plant Manager/Ops must commit to shift coverage and a completion date. Escalate if coverage cannot be secured.

3. Board reporting — Inconsistent training costs [6] A: moved

  • Situation: Finance needs plant-level training spend per head by November 20. Figures mix USD, CAD and MXN. Reporting periods, headcount definitions and conversion rules are unspecified.
  • Why it matters: A cross-plant comparison could mislead the board.
  • Decision: CFO/Finance leadership should confirm a common reporting basis and how to present figures that cannot be verified.

2. Handled — for information

  • Emerging Leaders: Cohort 3 completed. Average satisfaction was 4.7/5, but only 19 of 46 participants responded. Positive feedback focused on networking, not demonstrated business results. [4] D: merged
  • Supervisor onboarding: A 12-supervisor pilot is scheduled for December 2 at Joliet. The HR deck says "launched," but the pilot has not started. [5] B: promoted
  • Other HR operations: Benefits enrollment closed on time; HRIS upgrade phase 1 is reported complete. [5]

3. Watch

  • Greenville: Customer complaints increased following the September 8 order-entry system launch. Re-keying orders suggests a possible process or system problem. Ops and IT should investigate before L&D commits to customer service training. Complaint volumes and root cause are missing. [7] found
  • Emerging Leaders expansion: Proposal to double the 2027 cohort is premature. Evidence of learning impact, business outcomes and cost is missing. [4]
  • Supervisor onboarding rollout: Enterprise rollout depends on Q1 budget approval. Budget requirement and pilot results are not yet available. [5]

4. Site status — Safety-critical training C: reframed

The LMS counts completion events, not unique employees. It also lacks due dates. Event totals alone cannot confirm whether a site is on track.

Plant Status Evidence
Dayton Unknown 434 events/310 assigned; 140% is misleading. [3]
Greenville Unknown 265/280 events; individual completion unverified. [3]
Joliet Unknown 181/190 events; individual completion unverified. [3]
Cedar Falls Behind 48 refreshers outstanding. [1, 3]
Hamilton Unknown 236/240 events; individual completion unverified. [3]
Granby Unknown 150/150 events; unique completion unverified. [3]
Monterrey Behind 14 expired forklift certifications. [2, 3]
Saltillo Unknown No LMS data; migration pending. [3]
Querétaro Unknown No LMS data; migration pending. [3]

Bottom line: Two confirmed safety-training issues require immediate operational action. No plant can currently be verified as fully on track from the available LMS data.

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

What Dana changed, and why

ChatGPT found every problem planted in the updates. Dana's edits are not corrections. They are judgment calls about priority, audience and length.

What ChatGPT found: 8 of 8 planted problems
One run
✓“Complete” was 212 of 260
✓Forklift certs expired
✓140% counts retakes
✓Two sites had no data
✓4.7 came from 19 of 46
✓“Launched” was a pilot
✓Three currencies mixed
✓Greenville was a system issue
Four edits, all judgment calls
A
CHATGPT WROTE
Decision #3 for the CFO: “confirm a common reporting basis” for spend figures in three currencies.
DANA'S VERSION
Moved to Handled: “Finance's spend per head mixes CAD, MXN and USD; we convert at one stated rate.”
WHY · PRIORITY
Her team can fix this today. A decision slot is for something only a senior leader can approve.
B
CHATGPT WROTE
Supervisor onboarding under Handled, and the Q1 budget under Watch.
DANA'S VERSION
Needs a decision: “include the company-wide rollout in the Q1 budget review.”
WHY · PRIORITY
A budget call belongs to the CHRO. Left in Watch, it doesn't get made.
C
CHATGPT WROTE
Nine plants in one table: two “Behind”, seven “Unknown”, because the LMS counts events, not people.
DANA'S VERSION
The two confirmed problems first. The other seven marked “Unverified”, with event counts shown and headcounts requested.
WHY · AUDIENCE
The AI's rule is right, and stricter than our own first draft. Dana keeps it, puts the real problems first, and asks for the data that would settle the rest.
D
CHATGPT WROTE
About a page and a half. Emerging Leaders and onboarding each appear twice.
DANA'S VERSION
One page. Each item once.
WHY · LENGTH
A CHRO reads it in two minutes or not at all.
What only you can do
Decide what deserves your CHRO's attention, phrase it for that reader, and stand behind every number. The AI can find the problems. You sign the page.
Your turn

Make it yours

Swap in your own role, sources and audience. The method stays the same.

Your prompt: fill in the brackets

I lead [your function] at a [industry] company with about [number] people. Below are [number] updates I received this [week / month] from [who sends them]. I need to send [your audience] a one-page update before [the meeting].

Please read them all carefully and give me: 1. Needs a decision (up to three items, each with what's happening, why it matters, who should decide, and which update it came from); 2. Handled; 3. Watch; 4. [a section only your world needs, e.g. site status, program status, budget].

Be skeptical. Where a subject line, headline or summary says more than the details underneath support, go with the details and point out the difference. Don't add numbers together unless they measure the same thing in the same unit. If something is missing, say it's missing. Tell me which update each point comes from. Keep it to one page.

Five checks before you send
1
Every number traces to an update you can point to.
2
Words like complete, launched and approved match the details.
3
Nothing adds up different units: events and people, CAD and USD.
4
Safety and compliance items name who has to act.
5
It fits on one page, and you'd put your name on it.
Make it repeatable (optional)
Save the prompt once and next month just paste the updates: 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 reports
Use only the AI tool your company approves, and follow its rules on employee data. Keep personal details out unless your policy allows them.
What this example shows
One run of this prompt in ChatGPT found all eight problems planted in fictional updates, without inventing anything.
What it doesn't show
That every tool, setting or run will do the same, or how much time it saves you. Halden Industries and its people are fictional.
Appendix

Copy-ready: the prompt and the seven updates

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

The prompt

I lead Learning & Development at a manufacturing company. Below are seven updates I received this month from plants, Finance and my own team. I need to send my CHRO a one-page update before our monthly people review.

Please read all seven carefully and give me:

  1. Needs a decision — up to three items where the CHRO or another senior leader has to decide or act. For each: what's happening, why it matters, who should decide, and which update it came from.
  2. Handled — for information — items my team is dealing with that the CHRO should know about.
  3. Watch — things that look important but aren't decision-ready yet.
  4. Site status — one line per plant on safety-critical training: on track, behind, or unknown.

Be skeptical. Where a subject line, headline or summary says more than the details underneath support, go with the details and point out the difference. Don't add numbers together unless they measure the same thing in the same unit. If something is missing, say it's missing rather than guessing. Tell me which update number each point comes from.

Keep it to one page, in plain language a busy executive can read in two minutes.

[1] Email — Cedar Falls, IA

From: Rick Haldeman, Training Coordinator, Cedar Falls Subject: LOTO refresher — complete ✅

Hi Dana,

Good news, we wrapped the annual lockout/tagout refresher this week. Day and swing crews are done, 212 people through. Night shift we'll pick up when we can get coverage, probably after the holiday shutdown. Total headcount in scope was 260.

Thanks, Rick

[2] Teams message — Monterrey, NL

From: Lucía Garza, HR Generalist, Monterrey

Dana, quick one. The forklift recert vendor moved our session to Dec 9. 14 of our operators' certs expired Oct 31. Ops says they're still running the shift as normal, these guys have been driving for years. Let me know if you need anything from me.

[3] LMS export — Safety-critical training, November snapshot
Plant Assigned Completions %
Dayton, OH 310 434 140%
Greenville, SC 280 265 95%
Joliet, IL 190 181 95%
Cedar Falls, IA 260 212 82%
Hamilton, ON 240 236 98%
Granby, QC 150 150 100%
Monterrey, NL 420 398 95%
Saltillo, COAH — — —
Querétaro, QRO — — —

Export settings: Completions = completion events in period. Sites without data are not yet migrated to the LMS.

[4] Email — Leadership development

From: Priya Raman, Program Manager, Emerging Leaders Subject: Cohort 3 wrap-up 🎉

Hi Dana,

Cohort 3 wrapped last week and people loved it: average rating 4.7 out of 5! Strong signal the program is working. I'd like to propose doubling the cohort size for 2027 and would love 10 minutes at the next review to make the case.

Survey summary pasted below.

Emerging Leaders Cohort 3 — post-program survey Overall rating: 4.7 / 5 Responses: 19 (cohort: 46) Top comment theme: "great networking"

[5] Slide from the HR operations update deck

Slide 7 — Q4 highlights - ✔ Benefits enrollment closed on time - ✔ New supervisor onboarding launched - ✔ HRIS upgrade phase 1 complete

Speaker notes, slide 7: Supervisor onboarding: pilot cohort starts Dec 2 at Joliet (12 supervisors). Enterprise rollout pending Q1 budget approval.

[6] Email — Finance

From: Mark Ellison, FP&A Subject: Training spend per head — need by Fri

Dana, the CFO wants training spend per head by plant for the December board pack. Can you send by Friday Nov 20? Just use what the plants reported. Figures I have so far:

Hamilton 412 · Granby 388 · Monterrey 6,900 · Dayton 340 · Greenville 310 · Joliet 355 · Cedar Falls 360

(Hamilton and Granby report in CAD, Monterrey in MXN, US plants in USD.)

[7] Email — Greenville, SC

From: Tom Becker, Plant Manager, Greenville Subject: Need customer service training ASAP

Dana, we need customer service training for the order desk, and fast. Customer complaints are up since September. Ever since the new order-entry system went live on Sept 8 the desk has been re-keying orders and things are slipping through. Can you get something scheduled before year-end?

Tom