10 Prompt Templates for Every Presentation Type
These are starting points, not art. Fill the brackets, paste the result into any chat AI (or the advanced-prompt field of a slide generator), and you're most of the way to a working deck. Each template carries the structure that type of deck actually needs — the sequence of moves the room expects — which is where the quality comes from. The AI fills in text; you fill in specifics. Under each template: when to use it, how to adapt it, and what to check before you ship. At the bottom: the constraint block worth appending to every prompt you ever write.
1. Sales deck
Create a 10-slide sales deck offering [product] to [customer type].
Audience: [role], who currently solve this with [status quo].
Structure: their problem → cost of the problem → how we solve it →
proof (2 short cases) → pricing → next step.
Tone: concrete, no superlatives. End with a specific ask.
When to use: cold pitches, renewals, and the fifteen-minute slot at the end of a discovery call. Adapt it: swap the proof cases for ones in the prospect's industry — "a mid-market logistics company" means nothing to a healthcare buyer, and the status quo line is where the deck wins or dies, because the prospect's current workaround is your real competitor. Check: the pricing slide is a table, not a mystery; the ask names a date, not a feeling.
2. Quarterly business review
Create a 10-slide QBR for [team] covering [period].
Sections: results vs targets (as a table), what worked, what didn't
and why, what changes next quarter, asks.
Lead with the headline number. Be honest about misses. Bullets short.
When to use: any periodic review where the room mixes people who want the number and people who want the story. Adapt it: the "results vs targets as a table" line is the spine — keep it even if you change everything else; a QBR without a targets table is a highlights reel. Check: every miss on the table reappears in "what changes" — an acknowledged miss with no consequence reads as spin, and rooms track that better than you think.
3. Project kickoff
Create an 8-slide kickoff for project [name]. Audience: a mixed
technical and business team. Cover: the problem, the goal and the
non-goals, scope, milestones, roles, risks, communication rhythm.
End with the concrete actions for the first two weeks.
When to use: the first meeting of any cross-functional project. Adapt it: the "non-goals" line is the one that prevents the scope arguments that kill kickoff meetings — spend a real sentence on it, and get it agreed in the room, not just on the slide. Check: every milestone has an owner's name, and the two-week actions are small enough that someone could do one by Friday.
4. Training session
Create a 12-slide training deck teaching [skill] to [audience level].
Structure: why it matters → core concepts (one per slide, each with
an example) → common mistakes → a practice exercise → a summary
cheat sheet.
Tone: plain language. Explain any jargon on first use.
When to use: onboarding, tool training, process rollouts — anything where the audience has to do something afterwards. Adapt it: one concept per slide with an example beats a dense curriculum slide every time; if the skill has a checklist nature, make the cheat sheet slide the one people photograph and design it accordingly. Check: the practice exercise is actually doable in the session's time, and the mistakes section reflects the real ones your last trainees made, not the theoretical ones.
5. Conference talk
Create a 15-slide deck for a 20-minute talk on [topic] for [audience].
Structure: a hook (a story or a striking number) → the problem →
what I did → results, the honest ones → what the audience can apply
Monday morning.
One idea per slide, minimal text, speaker notes under each slide.
When to use: meetups, conferences, and the internal version — the tech-share slot. Adapt it: the Monday-morning slide is what people screenshot; design the deck backwards from it, and cut anything that doesn't serve that slide. The "honest results" line matters more than it reads — conference audiences are professionally skeptical of 100% success stories. Check: minimal text really means minimal; if a slide needs the notes to make sense, the talk structure is right and the slide is wrong.
6. Product demo
Create a 9-slide demo deck for [product] for [audience].
Structure: their workflow today, with the pain in it → the same
workflow with us → 3 feature slides framed as outcomes, not features →
proof → trial and next step.
Mark screenshot spots as [SCREENSHOT: what it shows].
When to use: structured demos — the deck frames the live product walkthrough, it doesn't replace it. Adapt it: the placeholders mark exactly where your real product images go; never let the AI invent UI, because a generated screenshot of a product you don't have is a credibility event waiting for its conference. Frame features as outcomes by asking, for each, "what does the user stop doing when this works?" Check: the "their workflow today" slide is accurate and a little bit generous — people defend their current workflow when it's mocked, and stop listening.
7. Investor update
Create a 7-slide monthly investor update for [company].
Sections: headline metrics (as a table), wins, misses and what we're
doing about them, asks (intros, hires, advice).
Tone: factual, no spin.
When to use: the monthly or quarterly update email-attached deck. Adapt it: investors read the miss list more carefully than the win list — a specific miss with a specific response builds more confidence than three vague wins, so keep both sections short and both specific. The asks section is the highest-leverage slide in the deck: name the intro you need, the hire you're stuck on. Check: metrics in the table match the numbers in the previous update — investors keep the last deck, and deltas that don't reconcile get questions you'd rather not answer cold.
8. Webinar
Create a 14-slide webinar deck on [topic] for [audience].
Structure: a title with the promise → agenda on one slide → 3 teaching
sections, each with one worked example → live demo placeholder →
Q&A slide with 5 seeded questions.
Notes: full conversational script, 60-90 seconds per slide.
When to use: lead-gen webinars, internal enablement broadcasts, anything to a camera. Adapt it: the seeded questions save you from a silent Q&A room — you pose the first one yourself, and the room follows. Webinars are the one case where a full script is the right default (see the notes article): the camera eats energy, and a script keeps pacing alive. Check: the title's promise is what the deck delivers — webinar audiences bail at the minute that the promise breaks.
9. Proposal follow-up
Create an 8-slide deck summarizing proposal [name] for [client].
Sections: their situation as we understand it → recommended approach →
timeline and team → pricing as one clear table → why us, in 3 points →
the decision and start date we propose.
When to use: after the proposal PDF goes out — the deck is what gets forwarded to the decision-maker who never read the PDF. Adapt it: "their situation as we understand it" is the trust-builder; use the client's own vocabulary from the sales calls, and get it right, because this slide proves whether you listened. Pricing appears once, as one clean table, because a hidden price reads as a negotiable price. Check: every number reconciles with the PDF — a deck that disagrees with its own proposal is a procurement question you don't want.
10. All-hands update
Create a 10-slide all-hands update for [company] covering [period].
Sections: one headline, metrics vs plan, 3 team wins with names,
challenges stated honestly, priorities for next period, one customer
story.
Tone: plain, warm, no corporate filler.
When to use: monthly or quarterly company updates, especially when the company is big enough that most people don't know what other teams shipped. Adapt it: names in the wins slide do more for morale than any slogan — specific people, specific work. The honest challenges section sets the tone for the whole culture slide-free; if leadership says the same thing in private, say it here. Check: the customer story is one story, told with a beginning and an end, not a list of logos.
Append this to every prompt
Constraints for every deck:
- one idea per slide; titles are takeaway sentences
- bullets under 12 words
- no agenda or thank-you slides
- no invented statistics; mark gaps as [NEED DATA]
- speaker notes on every slide
These five lines encode the defaults that separate a working deck from a topic summary. They come straight from the prompt formula, and they never need adapting — which is exactly why they should live at the end of every prompt, including the ones you write from scratch.
A template filled in, start to finish
Template 3, the kickoff, filled for a real-shaped project — this is what "replace the brackets" produces in about five minutes:
Create an 8-slide kickoff for project "Billing Migration".
Audience: a mixed team — 4 engineers, 2 support leads, the finance
partner; they know the current system's pain, they haven't seen the plan.
Cover: the problem (manual invoicing breaks past 500 accounts), the goal
(all invoicing on the new engine by March 1) and the non-goals (no
pricing changes, no customer migration in phase 1), scope (engine swap
plus a month of parallel run), milestones (parallel run starts Dec 15,
cutover Jan 20, old system read-only Feb 15), roles, risks,
communication rhythm.
End with the concrete actions for the first two weeks.
The brackets didn't just parameterize the template — they forced the project into sentences, and two of those sentences (the non-goals, the parallel-run scope) are the ones the team would have argued about in the meeting. Surfacing them before the room gathers is the template doing its real job: not saving typing time, but making the author decide the structure before the room does. The output deck will still need the specifics pass — real names on the roles slide, real risks from the last migration — but its skeleton will be right on the first try.
Your template library, maintained
A template that worked once is a prompt; a library of them is infrastructure. The mechanics are unglamorous: one file per deck type, named for the situation rather than the date ("sales-deck.txt", not "deck-final-v3.txt"), each containing the filled prompt from the last successful run plus three lines at the bottom — what got changed, what the room pushed back on, what to check next time. Prune quarterly: delete templates for deck types that stopped happening, merge two that grew into each other, and promote any constraint you've added three times into the standing constraint block, where it stops being your memory and becomes the system. Share the library — a shared notes document is enough — because a new hire starting from "the way we do QBRs here" beats a new hire starting from a blank prompt and their previous employer's assumptions about what a QBR is.
Mapping the calendar: which template when
Most deck types are predictable from the calendar, which means the library can be laid out against the year instead of discovered ad hoc. Planning season, usually Q4: the kickoff template drives initiative plans, and the investor-update structure doubles for the budget review — metrics table, changes, asks. Review season, at the end of each quarter: the QBR for teams, the all-hands structure for the company version, and the QBR's table discipline applied to any "how did we do" meeting. Launch weeks: the sales deck for the field, the demo deck for customers, the webinar script for the public session — all three from one launch master, as the repurposing article shows. Hiring seasons: the training template for onboarding modules, the demo structure for candidate "what we build" sessions. Fifteen minutes with the company calendar produces the list of which templates will be needed in the next ninety days — and that list is the difference between a library and a folder: a library gets maintained in advance of need, because the need is scheduled.
The calendar map also exposes gaps honestly: if your year contains a recurring deck type that no template fits — say, a monthly customer council — that's not a hole in the library, it's the next template to write. Take the best deck of that type anyone has given, strip its specifics into brackets, run it once, and append its three retro lines. A library built this way contains only templates your company actually needs, each earned by a real deck — which is why it stays small enough to maintain and never becomes a prompt-dump folder that nobody opens.
The five-slide version of every template
Every template above has a five-slide core, and knowing it saves you when the slot shrinks — the meeting moved from thirty minutes to ten, and the ten-slide deck won't fit. Sales deck → problem, cost, solution, one proof, ask. QBR → results table, what worked, what didn't, changes, asks. Kickoff → goal and non-goals, scope, milestones, roles, first two weeks. Training → why it matters, the two core concepts, common mistakes, cheat sheet. Talk → hook, problem, what I did, results, Monday morning. Demo → their workflow, yours, one feature as outcome, proof, next step. Investor update → metrics table, wins, misses, asks. Webinar → promise, two teaching sections, Q&A. Proposal → situation, approach, pricing, decision date. All-hands → headline, wins, challenges, priorities, story. Notice what survived in every single one: the ask, the evidence, and the misses. That's the irreducible core of all business decks — and if the five-slide version is honest, the ten-slide version is just it with support.
Templates for the small recurring meetings
The ten templates cover the big rooms; the small recurring ones — weekly status, retros, quarterly one-on-ones — deserve templates too, of a different kind. Status: three slides, ever — "done / doing / blocked," with blockers phrased as asks, because a blocker without an ask is a complaint. Retros: "keep / change / try," capped at five items per column, because a retro that lists everything changes nothing. One-on-one decks barely exist and mostly shouldn't — a shared doc beats slides for a two-person conversation — but the quarterly version ("what I need from you this quarter") works as a three-slide deck and forces clarity that conversation politely avoids. Standing templates for standing meetings are where compounding is fastest: the same structure every week means the room reads change, not format, and the meeting starts at the substance instead of the setup. Fill once, reuse forever, and let the constraint block ride along at the end as always.
What templates can't fix
Templates fix structure; they can't fix missing substance, and the boundary is worth stating plainly. A filled template with no real numbers in the input produces a well-organized deck of invented content — right skeleton, still a liability, because the structure makes the inventions look more credible. A template aimed at the wrong room — a sales-deck structure at a room that already decided not to buy — produces well-organized words that land on nothing. And no template supplies the position: the recommendation is yours, and if you don't write it, the AI will produce a plausible-sounding one, which is worse than producing none. So the standing order holds: brief first (the workflow article), template second. When a template's output feels hollow, the fix is almost always input, not prompt engineering.
How to adapt any template
Three moves cover ninety percent of adaptations. Reorder the structure: the templates lead with context because that's safe; executive rooms often want the recommendation first — move it and let the rest be support. Fuse two templates: a QBR that includes a demo is template 2's structure with template 6's demo section spliced in; keep one structure as the skeleton and import sections whole, because hybrid structures confuse the AI into mush. Change the audience line, not the structure: the same kickoff template serves a technical or a business kickoff — what changes is vocabulary, depth, and slide 1, and the audience line drives all three.
What not to change: the section count. Each template has the number of sections that fits its slot length, and adding a section means cutting a real one — if you can't name what gets cut, you don't have room for the addition.
From template to finished deck
The template is the opening move, not the game. After the first output, run the standing follow-up sequence: "Which two slides overlap? Merge them", then "Rewrite every title as the takeaway a listener should remember", then "Where would this audience push back, and which slide should preempt it?" Two minutes of pushes, and the deck goes from correct to yours. The workflow article has the full push list; the three above do most of the work.
Which template when
- Selling something → 1, and if the PDF already went out, 9 is its twin.
- Reporting a period → 2; reporting to the whole company → 10.
- Starting something → 3.
- Teaching something → 4; to a camera → 8.
- Convincing a crowd → 5.
- Showing a product → 6.
- Updating investors → 7, every month, without exception.
When two seem to fit — a pitch that's also a demo, an all-hands that's also a QBR — pick the one whose last slide matches your goal: the deck that ends on the decision you need is the one to start from.
Start from the closest template rather than a blank prompt — the structure carries half the quality, and the structure is the part you'd otherwise get wrong under deadline pressure. If none fits, write the brief from the workflow article and let the AI propose the structure from that; then save your filled version, because next quarter it's a template too.