How to Sequence a deck

The symptom

Sixty slides.

Nobody set out to build sixty. The number came up once in a meeting about whether the deck had gotten too long, everyone agreed it had, and three more slides went in that quarter. The core argument is still in there. It’s on slides four, five and nine, surrounded by fifty-five slides of everything else the firm has ever needed to say.

You’ve watched what happens in the room. Somewhere around slide twenty the prospect’s attention changes texture. They’re still nodding. They’ve stopped asking anything. By slide forty the meeting has quietly turned into a performance with an audience, and the questions at the end are polite ones about timing and next steps.

Then the follow-up email asks for something the deck covered twice.

Why it keeps happening

A deck accumulates because every addition to it was defensible.

Each slide got requested by somebody senior. Each one answers a question a real prospect asked, once, in a real meeting. Nothing in the review process ever asks whether the slide earns forty-five seconds of a first conversation, because the deck has no defined job. It’s the firm’s collected answer to every question it has ever been asked, sorted roughly by the order in which the questions arrived.

We’ve added the slide. It answered a real question from a real prospect, it was correct, and it made the deck worse for everyone who came after.

That’s the whole mechanism in one move. The slide was correct on the day it went in, and it became permanent on the same day.

Across three decades of sitting in on firms reviewing their own materials, the removal conversation almost never happens. Reviews add. Somebody has to own a deletion, and cutting the slide a managing partner asked for last spring costs more internally than keeping it costs in any way the firm can measure. So the deck grows toward completeness, which nobody chose and nobody wanted.

What the buyer is doing at the Proof gate

Deciding whether the evidence in front of them settles the specific thing they’re uncertain about.

That’s narrower than it sounds. A buyer at this gate is carrying one or two live doubts, and proof works when it lands on those. Volume doesn’t substitute. A case study can be well built, credible and true, and still arrive at the wrong time, in which case it reads as material rather than as an answer.

Evidence delivered before the corresponding question exists creates noise. The buyer has to take it in, work out whether it matters, and file it somewhere. None of that moves them closer to comfort. Do it thirty times in forty minutes and you get information fatigue, which persuades nobody. What it produces is a buyer who stops evaluating and settles for whatever feels good enough to end the work.

Then the firm reads that disengagement as insufficient interest, and responds by adding proof.

The correction

Split the deck in two, and make the split visible.

The core carries the argument and nothing else. Give it a hard ceiling and name the number out loud: the count of slides you can present properly in the time a first meeting really gives you. Twelve is a defensible answer for forty minutes. The specific number matters less than the fact that one exists, because a ceiling is what turns every new addition into a trade rather than an insertion.

Everything else moves to an appendix designed to be read alone rather than presented. That appendix used to be a safe place to overstuff, on the theory that nobody read it. Now the file gets fed to a model for a summary before a human opens it, and whatever you padded it with is legible in seconds to somebody with no context at all.

Then the sorting rule, which is the part that does the work.

Go slide by slide and name the buyer question each one answers. Three outcomes:

- A slide that answers no question comes out of the deck entirely, appendix included.

- A slide that answers a question this buyer hasn’t asked yet goes to the appendix, where it’s ready the moment the question shows up.

- Slides answering the same question get merged. There will be more of these than you expect.

Give the deck visual discipline while you’re in there. A narrative slide, a proof slide and a technical slide are doing different jobs, and they shouldn’t look or behave the same way. Someone flipping through should know what kind of thing they’re looking at before they’ve read a word. When every page is formatted identically, the reader has to decode each one to find out whether it matters.

What order the surviving core slides run in is a separate problem, and snippet 10 handles it. This one only decides what’s in the room at all.

Before and after

Before. Slide 14 of 60. “Our Proprietary Six-Phase Engagement Methodology.” Presented at minute nine of a forty-minute first meeting, to a buyer who hasn’t yet said what’s broken.

After. Core slide 6 of 12. “You said the last two providers missed go-live. Here’s where our schedule puts checkpoints at the two places that usually break.” Appendix A, page 3: the full six-phase methodology, checkpoints marked, sent after the second meeting, because by then somebody asked.

The hard call

The slides that have to come out belong to people. Somebody asked for that one, and they’re still at the firm, and they’ll notice.

There’s no version of this where the deck gets shorter without that conversation happening. What makes it survivable is the ceiling, because it moves the argument off the merits of any single slide, which are usually fine, and onto what it displaces. The partner’s slide is fine. It’s also the eleventh best use of twelve.

Run this on your own material

Open the deck you’re taking into the next first meeting. In the margin of every slide, write the buyer question it answers, phrased the way a buyer would say it out loud rather than the way your practice group would.

One sentence per slide. No credit for “our capabilities.”

Count the blanks. Count the duplicates. Those two numbers are the size of the problem, and they’re usually large enough that the ceiling stops feeling arbitrary.

Do this before the deck goes out, with the slides you already own. It’s an afternoon, and it doesn’t require design help.

Signals you’ve cleared it

Advance when the buyer interrupts the core deck with a question you were about to answer, and when they ask for something in the appendix by name. Both mean the question arrived before the evidence did, which is the sequence you’re after. Advance too when a first meeting ends early because the argument landed.

Hold when you presented every slide and got no questions. Silence at the end of a long deck reads as agreement and usually isn’t. Hold when the follow-up asks for material you already showed, because that means the buyer couldn’t find it under everything around it. And hold when the answer to “which slide comes out” is that they all have to stay.

Where this connects

Engagement Success in an AI World · Built to Be Read, on evidence arriving before the question exists.

Research Library · Entry 01, bounded rationality, on satisficing under information load.

Next move · Go to How to write a case study.