A bad monthly business review is 40 slides of context, a status section nobody trusts, and three decisions buried in slide 35 that get pushed to next month. A good one is 8 slides, R/Y/G initiative status with named owners, the decisions called out by slide 3, and asks phrased as “Sarah, by Friday, because…”. These 11 prompts force exactly that shape, plus a 2-page board pre-read and a pre-cooked answer pack for the questions execs are most likely to throw. Pair them with the executive summary prompts for the cover page.
TL;DR
- Paste your raw metrics and initiative list into prompts 1–3 to get the spine of an 8-slide deck in one pass.
- The current best practice (2026) is a tight deck of roughly 8–12 slides, green metrics left in the pre-read, and only 2–3 in-room decisions; these prompts enforce that discipline.
- For slide prose (narratives, risk framing, pre-cooked Q&A) Claude Opus 4.7 or Sonnet 4.6 tends to read tightest; for fast iterating and image-heavy decks, GPT-5.5; if you live in Google Slides, Gemini 3.1 Pro inside Workspace.
- No chatbot exports a finished PowerPoint. These prompts give you decision-ready text you paste into Slides, Gamma, or PowerPoint.
Best for
- Executive monthly business reviews
- QBR prep
- Investor monthly updates
- Board pre-reads
- Cross-org rollups
Which model to run these in (June 2026)
You are generating decision text, not a designed file, so model choice is about writing quality and how much data you can paste in.
| Job | Pick | Why |
|---|---|---|
| Tight slide prose, risk framing, Q&A pack | Claude Opus 4.7 / Sonnet 4.6 | Cleanest exec prose; 1M-token context fits a quarter of raw metrics |
| Fast iteration + image-heavy decks | GPT-5.5 (ChatGPT) | Strong in-app image generation; Plus in-app context is ~320 pages |
| You build the deck in Google Slides | Gemini 3.1 Pro | Runs inside Workspace, so no copy-paste between windows; 1M context |
All three handle these prompts well. If you are pasting a big spreadsheet dump, a 1M-token model (Opus 4.7, Sonnet 4.6, or Gemini 3.1 Pro) swallows more raw data in one go; full 1M in ChatGPT requires the $200 Pro tier as of June 2026. For a deeper split, see ChatGPT vs Claude vs Gemini. To turn the output into actual slides, follow the work presentation outline guide.
A reliable framing for the whole deck is the Amazon-style three-act spine: What happened (metrics), what it means (narrative), what I need (decisions). The prompts below map onto it.
1. MBR deck outline (8 slides)
Outline an 8-slide MBR deck. Audience: [execs]. Output the 8 slide titles + 3-bullet body each. Structure it as What happened (metrics) / What it means (narrative) / What I need (decisions). Must include: headline summary, key metrics with month-over-month deltas, top risks, and the 2-3 decisions needed in the room.2. Red/Yellow/Green status report
Draft the R/Y/G status section of the MBR. Initiatives: [paste list]. For each, give status color, a 1-line reason, named owner, and expected status next month. Keep green items to one line each so meeting time goes to yellow and red. <=150 words total.3. Headline-first MBR summary
Write the headline summary slide of the MBR. 5 bullets max. Each bullet must contain a number or a decision. Ban "we continue to focus on...". Output for [company / org / function]: [paste data].4. Metric narrative (why did it move)
Metric: [name]. Last month: [value]. This month: [value]. Hypothesis on movement: [paste]. Write the 80-word narrative that interprets the move for an exec audience. Name 1 leading indicator to watch next month.5. Risks & mitigations slide
Draft the risks slide for an MBR. Inputs: [paste 5 candidate risks]. Output: top 3 risks selected, severity, owner, mitigation, and what specifically triggers escalation. Cut anything not actionable.6. Decisions-needed slide
Draft the "decisions needed" slide. Inputs: [paste decision candidates]. Output: top 3 decisions, 1-line context each, options considered, your recommendation, and the exec who must approve. If a decision has no real fork, drop it.7. Pre-meeting reading note
Write a 250-word pre-meeting reading note that lets execs skip the deck if they want. Include: top 3 facts, top 3 risks, top 3 decisions. Be ruthless about cuts; this should carry 80% of the status so the meeting is for decisions.8. Asks-from-execs slide
Draft a "what we need from leadership" slide. Inputs: [paste raw asks]. Output: 3 asks max, each <=20 words, specifying name + ask + deadline + why-now. No vague "more support".9. Q&A pre-cooked answers
Predict the top 8 questions execs will ask in this MBR. Inputs: [paste MBR draft]. For each, write a 60-word pre-cooked answer. Mark the 2 most likely "gotcha" questions and give a fallback line for each.10. Year-end / annual review
Draft a year-end version of an MBR. Inputs: [paste yearly metrics + initiatives]. Output: 6 slides - headline, what worked, what did not, key metrics with annual deltas, top 3 lessons, next-year priorities.11. Board pre-read
Convert the MBR into a 2-page board pre-read. Audience: board members who get an hour to read 5 docs. Output: page 1 = headline + metrics + risks; page 2 = decisions + asks. No marketing language.How to chain these in one sitting
- Run prompt 1 to lock the 8-slide spine.
- Feed your raw numbers to prompts 3 and 4 for the headline and each metric narrative.
- Run prompts 2, 5, 6, and 8 for status, risks, decisions, and asks.
- Paste the assembled draft back into prompt 9 so the Q&A pack is built from the actual deck.
- Run prompt 7 (pre-read) and, if it is a board month, prompt 11.
Keep one source-of-truth message thread per review so the model carries context across all 11 calls instead of re-explaining your org each time.
Common mistakes
- Dumping 40 slides of context that nobody reads before the meeting
- No R/Y/G initiative status, so the room can’t triage in 60 seconds
- Decisions buried in slide 35 and pushed to next month every time
- Asks without owner + deadline (“we need more support”) that execs can’t action
- Same deck month-to-month with no narrative on what actually changed
- Metric moves shown without a “why did it move” hypothesis or a leading indicator
- Pasting numbers and trusting the AI’s math; always re-check totals and deltas against the source
FAQ
Which AI model writes the best MBR slides? For pure writing quality, Claude Opus 4.7 or Sonnet 4.6 tend to produce the tightest exec prose and handle a quarter of raw metrics in one 1M-token context. If you want fast iteration or in-app charts and images, GPT-5.5 in ChatGPT is strong. If your deck lives in Google Slides, Gemini 3.1 Pro inside Workspace skips the copy-paste step. None of them exports a finished slide file; they give you text.
Can any of these tools build the actual PowerPoint or Google Slides file? Not directly from these prompts. As of June 2026, general chatbots output text you paste into a slide tool. For a designed deck, paste the output into Gamma or use Google Slides’ built-in Gemini. Treat the AI as your writer, not your designer.
How long should an MBR deck be? Current best practice is roughly 8–12 slides presented in about 20 minutes, with detail pushed to an appendix and pre-read. Green metrics stay in the pre-read; meeting time goes to yellow and red items and the 2–3 decisions that actually need the room.
Is it safe to paste real financials into these prompts? Check your company’s policy first. For sensitive numbers, use an enterprise or team plan with no-training guarantees (ChatGPT Team/Enterprise, Claude Team/Enterprise, Google Workspace), or replace exact figures with deltas and percentages before pasting.
How do I stop every month’s deck from looking identical? Lead with prompt 4 (metric narrative) on the two or three numbers that actually moved, and force prompt 3 to put a number or a decision in every bullet. The “why did it move” line is what makes a review read as new instead of recycled.
Related
- Weekly report prompts
- Executive summary prompts
- Meeting notes prompts
- Project planning prompts
- Action item extraction prompts
- Business Review Talking Points Prompts for QBR and MBR
- KPI Commentary Prompts: 12 Templates for Numbers That Tell a Story
- AI Business Driver Breakdown: The Two Numbers That Move Revenue
- AI Runs Your Monthly Finance Check-in