How to outline a presentation: build the middle first, not the opening. Settle on three to five core ideas, attach a real example to each one, then lift one of those stories to the front to open with and hold another back for the close. Almost nobody does it in that order. They open the slide software, start on slide one, and let the deck decide the shape of the talk. I have watched sharp executives lose a room that way long before they reached their best material.

Why this matters (and what most executives get wrong)

There is a default presentation shape in corporate life, and it is the data dump. Agenda slide, background slide, then forty slides of everything the presenter knows, arranged in the order the research happened to arrive. It feels thorough. It feels safe. It is the reason people check their phones by minute six.

The problem is not the volume of information, it is that nothing has been ranked. If you give an audience fifty two ideas, they do not walk out with fifty two. They walk out with whatever three happened to stick, and you had no say in which three. An outline is how you take that choice back. You decide what the three to five are, you build the talk to protect them, and everything else becomes support rather than competition.

This is also why starting with your opening slide is backwards. You cannot write a good opening until you know what you are opening toward. Decide the destination first, then the on-ramp practically writes itself. That sequencing sits underneath the whole Story-Driven Data™ approach I teach, and it is the single biggest change most executives can make to how they build a talk.

An audience does not remember fifty two ideas. They remember three, and if you did not choose which three, the room chose for you.

The 5-step way to outline a presentation

Draw a grid on a sheet of paper. Two columns, five rows. That is the whole tool, and it is faster than any outline feature in any slide program.

Step 1: Start in the middle

Skip the opening and skip the close. Go straight to the body of the talk, because the body is the reason the meeting exists. Everything else is packaging, and packaging built before you know what is inside it never fits. In the left column of your grid, you are about to write the ideas. In the right column, the examples. Nothing else goes on the page yet.

Step 2: Cut down to three to five core ideas

Write your main ideas in the left column, and stop at five. Not because you only have five, but because five is roughly the ceiling of what a room will carry out the door. If you have twelve, look for the three or four that the others are actually serving, promote those, and demote the rest into sub points underneath. This is uncomfortable and it is the most valuable twenty minutes in the whole build. The ideas you cut are not lost, they are subordinated.

Step 3: Attach a "for example" to every idea

Now fill the right column. Beside each core idea, write the story, the case, or the moment that proves it. The words "for example" are the hinge, because that phrase is what moves an idea from something the audience heard to something they can picture and repeat. An idea without an example is a claim. An idea with an example is evidence, and evidence is what gets passed along to the people who were not in the room.

Step 4: Open with fire

Only now do you write the opening, and you barely have to write it, because it is already on your page. Take the strongest story from the right column and move it to the front. Start there, cold, with no preamble. Then say hello, say who you are, and say what the next twenty minutes will cover. Almost everyone else in the building opens with a title slide and a housekeeping paragraph, so opening on a story is a low cost way to sound like the only person who prepared.

Step 5: Close on purpose, not on "any questions"

The standard ending is "so that is everything, any questions," which hands the last moment of your talk to whoever speaks first. Replace it with a signal and a landing. The signal is a line like "as I begin to close this out," which tells the room to pay attention again. The landing is one idea you want them to leave with, or one specific action you want taken, ideally carried by the story you deliberately held back in step three. Then stop and thank them.

The mistake most executives make

The mistake is building in the order the audience will hear it. It feels logical, and it produces a talk shaped like the research rather than like the point. You spend your freshest thinking on the introduction, which is the least important part, and you arrive at your core ideas tired, late, and out of time to find examples for them.

The second version of the mistake is treating the grid as an outline to be expanded rather than a filter to be enforced. I watch people fill in three to five core ideas, feel good about it, and then quietly add four more while they are building slides. By the time the deck is done the discipline is gone and the data dump has grown back. The grid only works if the number on the page is the number in the room.

The two column grid

Left column: three to five core ideas, no more. Right column: one story or example for each. Build that first, then pull one story forward for the opening and hold one back for the close. Slides come last, and only for the ideas that survived.

Case study: the forty slide update that became a nine minute talk

A division head at a professional services firm asked for help two days before an annual update to his firm's leadership. He sent me the deck first. Forty one slides, every one of them accurate, arranged in the order his team had produced the analysis through the year.

We never opened the deck again. We drew the grid on a legal pad and asked one question: if the room remembers three things next month, what should they be? He had them in about twelve minutes, and they were not the three the deck led with. Then we went hunting for examples, and that took longer, because two of his ideas had no story attached at all. That is usually the tell that an idea is a summary rather than a point.

He opened with a client moment from the second row of the grid, told cold, before he introduced himself. He closed with the one he had saved and a specific ask. The talk ran nine minutes instead of forty, and he took questions for almost half an hour, which is the version of that meeting everybody actually wanted. Same year, same numbers, same person. The only thing that changed was the order he built it in.

Two ways to build a talk

How you build itSlide one firstMiddle first
Where you startTitle and agendaThree to five core ideas
How many ideas shipHowever many you haveThe number the room can carry
Where the examples come fromWhatever is left at the endChosen alongside each idea
The openingHousekeepingA story you already had
The ending"Any questions"One idea, one ask
What the room repeats laterWhatever stuck by accidentWhat you chose

The last row is the whole argument. Every talk gets reduced to a few sentences the moment it ends. Outlining is just the decision to write those sentences yourself.

What to do next

Before you open your slide software for your next talk, spend twenty minutes with a sheet of paper and the two column grid. Three to five ideas on the left, one example each on the right. If you cannot find an example for an idea, that idea is not ready to be one of your five.

If you want help doing this with a talk that actually matters, that is the work I do with executives and leadership teams. You can get a quick quote for working together. For the two ends of the structure, here is how to open a presentation and how to end a presentation, and if the examples are the part you are missing, start with how to use storytelling in presentations.

Nobody ever complained that a presentation was too easy to remember.

Frequently Asked Questions

How do you outline a presentation?
Outline the middle first. Choose three to five core ideas, attach a concrete example to each one, and only then write the opening and the close. Lift your strongest example to the front so you open on a story rather than an agenda slide, and hold another one back to carry your closing point. Building in that order means the opening is written toward a destination you have already chosen, instead of guessing at it.
How many main points should a presentation have?
Three to five. That is not a limit on how much you know, it is a realistic estimate of how much an audience carries out of the room. If you present fifty two ideas the audience still leaves with about three, and you had no control over which ones. Picking the three to five yourself and demoting everything else into supporting detail is how you decide what gets remembered.
What is a data dump in a presentation?
A data dump is a presentation built by loading every fact and slide the presenter has into the order the research arrived in, with no ranking. It feels thorough and safe to the person delivering it, but nothing has been prioritized, so the audience does the ranking for you at random. The fix is not cutting content for its own sake, it is choosing which few ideas the rest of the material now exists to support.
Why should you write the opening of a presentation last?
Because an opening is an on-ramp, and you cannot build one until you know where it leads. Writing the introduction first spends your freshest thinking on the least important part of the talk and leaves you arriving at your core ideas late and out of time. Once the middle exists, the opening is usually already sitting in your examples column waiting to be moved to the front.
How do you end a presentation without saying any questions?
Use a signal and a landing. The signal is a short line such as as I begin to close this out, which tells the room the talk is arriving somewhere and pulls attention back. The landing is one idea you want them to leave with or one specific action you want taken, ideally delivered through a story you deliberately saved. Then stop and thank them, and take questions after the ending rather than as the ending.
Do you need slides to outline a presentation?
No, and starting with slides is usually what causes the problem. A two column grid on a sheet of paper, ideas on the left and examples on the right, outlines a talk faster than any outline feature in any slide program. Slides come afterward, and only for the ideas that survived the cut, which tends to reduce a deck substantially without removing anything the audience was going to remember.