Guide · 8 min read

Action Titles: The One Slide Habit That Separates Consulting Decks from Everyone Else's

Flip through a consulting deck reading only the top line of each slide and you get the entire argument — claim by claim, in order. Do the same with most other decks and you get a table of contents: "Background", "Q3 Revenue", "Next Steps". The difference is one habit, it costs nothing, and you can retrofit it onto any deck in an afternoon.

What is an action title?

An action title is a full-sentence assertion at the top of a slide that states the slide's conclusion — "Q3 revenue grew 14%, but the growth came from two accounts" — where a topic title merely names the subject: "Q3 Revenue". The title does the arguing; everything below it is evidence. That single change redefines what a slide is. A labelled slide is a folder of related material the audience must interpret for themselves. An action-titled slide is a claim plus its proof, and a claim can be tested: the reader can agree, push back, or point at the chart and say "that doesn't show that". A label can't be wrong, which is exactly what's wrong with it.

The habit comes out of strategy consulting. Barbara Minto, who taught writing at McKinsey in the 1970s, codified it in the Pyramid Principle: lead with the governing thought, then support it — on every page, at every level. The same idea reached technical audiences as the assertion–evidence slide format, long advocated in engineering-communication circles: a sentence headline stating the finding, supported by visual evidence rather than bullet decoration. Consulting firms add a discipline worth stealing whole: the storyboard. Partners draft the deck as titles alone — a "ghost deck" of empty slides, each carrying only its sentence — and debate the argument before anyone builds a chart. If the titles don't survive that argument, no amount of slide-craft will save them. It's the spine of the whole genre, and our guide to consulting-style decks builds the rest of the skeleton around it.

Two versions of the same slide side by side: the left slide has a short muted topic label at the top, the right slide has a full-width sentence bar stating the conclusion — the body content of bullets and a small chart is identical on both. Topic title: a label “Q3 Revenue” — the reader must work out the point rewrite Action title: a full-sentence conclusion “Growth came from two accounts” — the point rides on top
Same evidence on both slides; the only change is the top line doing the arguing. A deck of right-hand slides can be read title-by-title and still deliver its whole story.

Six titles, rewritten

The fastest way to internalise the habit is to watch it work. Six labels you have seen a hundred times, rewritten as assertions — each pair imagines the slide's body staying exactly the same.

Topic titleAction title
Q3 RevenueQ3 revenue grew 14%, but the growth came from two accounts
Competitive LandscapeThree of our five competitors now bundle the feature we charge for
Customer Survey ResultsCustomers stay for support quality, not price — 61% ranked it first
Project TimelineShipping in March requires the vendor decision by 15 December
Hiring UpdateEngineering hiring is six weeks behind plan; two offers close the gap
RisksThe biggest risk isn't the competitor — it's our own integration backlog

Illustrative rewrites, not templates — your slide's assertion has to come from your slide's evidence, which is rather the point.

Notice what the rewrites do mechanically. A verb appears. The number that matters climbs out of the chart and into the sentence. The so-what stops being homework for the audience. And notice the uncomfortable side effect: writing the assertion forces you to decide what the slide actually says — and occasionally reveals that it says nothing, which is the most useful thing a title can ever tell you.

The horizontal logic test

Consultants call it horizontal logic: read only the titles, first slide to last, and you should hear the complete argument as a coherent paragraph. Try a churn story — Churn doubled in Q2. The increase is concentrated in customers onboarded without training. Trained customers churn at a third the rate. A mandatory onboarding call would cost about £40k a year. We should pilot mandatory onboarding for enterprise accounts in Q3. Five titles, and a reader who never looks below the top line still knows what you found and what you want.

The companion check is vertical logic: on each individual slide, does the material below the title actually prove the title? Horizontal logic catches structural problems — arguments that don't chain, slides that repeat each other, an ending that doesn't follow. Vertical logic catches slide-level problems — a bold assertion sitting on top of a chart that shows something adjacent. Run both and you have reviewed a deck properly without reading a single bullet. This is also why the habit matters most in decks that travel without you: a pre-read, a forwarded file, the board deck a director skims at 11pm. Nobody narrates those. The titles are the talk.

How to retrofit an existing deck

You rarely get to start from a storyboard. The retrofit pass takes under an hour for a twenty-slide deck, and it improves the deck even if you keep half the old titles.

  1. Pull the titles out. Copy every slide title into a plain list, in order. PowerPoint's outline view does this in seconds. You are about to edit the list, not the slides.
  2. Say each slide's point aloud. For each slide, finish the sentence "This slide shows that…". Whatever finishes it is your action title. If you can't finish it, you've found a slide with no point — cut it or merge it before polishing anything.
  3. Replace the label with the sentence. One sentence, with a verb, ideally a dozen words or so. If a specific number is the point, the number goes in the title. Resist the caveats — they live in the body.
  4. Read the new list top to bottom. This is the horizontal logic test. Reorder until the list reads as an argument; delete any slide whose sentence just restates a neighbour's.
  5. Make each body earn its title. Everything on the slide is now either evidence for the assertion or clutter. Move clutter to an appendix — cut slides love it there, and it keeps them alive for Q&A.

When a topic title is fine

Not every slide argues, and forcing assertions onto slides that don't makes a deck faintly ridiculous. Section dividers ("Part 2: Churn") are signposts; label them. Agenda and closing slides are furniture. Reference decks — a rate card, a glossary, an appendix people consult rather than read — are lookup tools, and lookup tools want labels the way a filing cabinet does. And the photo-plus-one-word keynote genre lives at the opposite pole entirely: the slide is a backdrop and the argument lives in the speech. The rule of thumb: if the deck is meant to persuade, title with assertions; if a slide is furniture or a lookup, a label is the honest choice.

Where action titles go wrong

Three failure modes account for most bad attempts. The first is the observation that stops short of an implication: "Churn rose 4%" is a full sentence and a true one, but it just moves the label problem into grammar — the audience can still reply "so what?". The strong version answers before they ask: rose where, because of what, demanding which decision. The second is the two-line ramble — a title carrying three clauses and two caveats. If the sentence genuinely needs two lines, it is usually two slides pretending to be one; split it, and let the caveat live in the body where it belongs. The third is the overreaching title — an assertion the slide beneath doesn't prove. This one is worse than a label, because assertions get noticed: the sharpest person in the room will test your boldest sentence against your own chart. Vertical logic is not optional once you start making claims.

Why AI-generated decks default to topic titles

Ask most AI deck generators for a sales strategy deck and the titles come back as labels — "Market Opportunity", "Our Approach", "Next Steps" — because the tool is filling a template skeleton, not making an argument; it answers "what sections does a deck like this have?" rather than "what is this deck claiming?". It's the same slide-by-slide generation habit that makes AI presentations look generic in every other respect too. Narrative-first tools invert the order: argument first, slides after. That's the approach Kinsy takes — it plans the deck's narrative before laying out a single slide, so titles descend from an argument rather than a template, and because the export is a native, fully editable .pptx where every text element opens as a real text object, sharpening any title it gets wrong is ordinary PowerPoint editing rather than image surgery.

However the deck gets made, the habit is the same one Minto was teaching fifty years ago: decide the point, put the point on top, make everything underneath earn it. It is the cheapest upgrade in presentation craft — no design skill required, just the nerve to write down what you actually think.

Keep reading

Titles that carry the argument

Try a deck builder that plans the narrative first, backs claims with live research and numbered citations, and exports a native, fully editable .pptx — so every title is a real text box you can sharpen. 100 free credits — about one full deck. No card required.

Make a deck free