What is a ghost deck?

Titles first, content later

A blog written by a former McKinsey consultant, Working With McKinsey, describes the ghost deck, also called a shell or skeleton at other firms, as a PowerPoint deck that is about twenty percent complete, with most of the effort going into the leads and headlines that push the storyline forward. Aside from those, most ghost pages are blank or carry rough sketches of exhibits the team intends to complete. Teams often sticker pages with notes on which data is already available and what still needs to be gathered. The deck is deliberately unfinished: its value lies in being quick to write, easy to review and painless to change.

Why consulting teams build them

The same source puts the purpose plainly: ghost decks are used to align on what the end product will look like and to minimise wasted work. A partner can review twenty headlines in minutes and redirect the whole project; reviewing twenty finished slides takes hours and tends to produce comments on design rather than argument. The ghost deck also doubles as a work plan, because every placeholder chart is a piece of analysis someone must do. The practice rests on the answer first thinking of Barbara Minto's pyramid principle, developed at McKinsey, in which the conclusion and its supporting points are decided before the evidence is assembled around them.

What goes on a ghost page

A useful ghost page has three things. The title, stating the slide's conclusion as a sentence, even if it is a hypothesis. A placeholder showing the kind of exhibit that will prove it, such as a bar chart of cost by region or a two column comparison, sketched or described in a box. And a note on the data: where it comes from, who is getting it and whether it exists yet. Pages that cannot be given a headline reveal that the team does not yet know what that part of the deck is for.

Common mistakes

The first is ghosting with topic titles, such as market overview, which creates the structure of a deck without its argument and defeats the purpose. The second is polishing ghost pages, spending time on layout before the storyline is agreed. The third is treating ghost headlines as findings and forgetting they are hypotheses; when the analysis disagrees, the headline must change. The fourth is skipping the review, building a ghost deck but never walking a senior reader through it. The fifth is showing a ghost deck to a client without explaining what it is, which can look like unfinished work rather than a deliberate stage.

Ghost decks outside consulting

The technique suits anyone building a high stakes deck with others: a fundraising deck reviewed by advisers, a bid response written by several departments, a board paper assembled from contributions. Sharing a ghost version early lets each contributor see where their part fits and what it must prove, which cuts the late merging of slides written in different styles to different lengths. It also gives a founder or project lead an early, honest view of the workload: a ghost deck with twelve placeholder charts that nobody yet has data for is a clear signal to shorten the deck or start collecting evidence now, weeks before the presentation date rather than the night before.

Where it shows up in the product

A slide added here from Library in the slide rail arrives as a shell: the chosen template with muted placeholder text in the right shapes, ready to fill. A deck of such shells, with the title of each written as its conclusion, is effectively a ghost deck built from real layouts, since the placeholders already show whether a slide will hold a chart, a table or three pillars. Speaker notes, stored per slide and never shown on the slide, are a natural place for the data notes a ghost page carries. Filling shells can be done by hand or by asking the AI assistant, whose changes arrive as a preview you apply.

Questions people ask