Pitch decks built on the argument first, then the slides

The problem is rarely that the slides are ugly

A founder books this the week before a meeting, usually with a deck that has grown by accumulation. Forty slides, seven lines on each, three years of edits by four people, and a section that was relevant to a strategy the company has since abandoned. The owner knows it is not working. What they cannot see, because they are inside it, is which parts are load-bearing and which are just there.

The sales version of the same problem is quieter and more costly. A deck built three years ago is still going out every week, with an old logo, superseded pricing and a product line that was discontinued. Nobody owns it, so nobody fixes it, and every prospect sees a company that looks slightly behind. And there is a third reason people buy this that is worth stating honestly — they do not want to spend two days sitting with it themselves, and two days of a founder's time is worth considerably more than this work costs.

What actually decides whether a deck works is the argument underneath it. Slide order, what gets cut, and whether each slide makes one point or three. The visual design matters — a deck that looks careless undermines a company that claims to be careful — but it is the second problem, not the first, and studios that start with the second one produce forty beautiful slides that still lose the room.

How a deck gets rebuilt

1. Read everything and find the argument. Your current deck, the memo, the website, whatever exists. We come back with the running order — what comes first, what moves, what merges, what is cut and why, and what is missing entirely. Investor decks have a conventional sequence — problem, solution, market, product, model, competition, team, numbers, ask — and departing from it is allowed, but it should be a decision rather than an accident. One day.
2. Write the headlines. Every slide gets one line at the top that states the point, not the topic. Market is a label; a sentence carrying the actual finding is a headline. This is the stage that does the most for the deck and the one most often skipped, because it is writing rather than design. It also exposes the slides that have no point to state. One to two days.
3. Reduce to one idea per slide. If a slide is making three arguments it is three slides or it is one slide and two cuts. Detail that matters but does not belong in the flow moves to an appendix behind the ask, which is where the questions get answered. Half a day to a day.
4. Design the system, then the slides. One layout grid, one type scale, two colours used consistently, and a small set of slide templates that everything else is built from. Charts are rebuilt from your data with labelled axes and stated units rather than pasted in as screenshots at whatever resolution they happened to be. Two to five days.
5. Deliver in both forms and check it on the machine that matters. A PDF for sending and a live editable file in your tool. Every deck is opened and paged through in the destination application before handover, because fonts that are not installed on your machine get substituted and a substituted font reflows a slide. Half a day.

That last check is not busywork. A deck built in a design tool and exported to PowerPoint will use whatever typeface the receiving machine can find unless the file is prepared for it. The failure shows up at the worst possible moment — on a borrowed laptop, in a boardroom, with the headline now wrapping onto three lines.

Where the cost sits

Whether the content exists. Tightening and redesigning material that is already written is a fundamentally different job from developing the narrative with you, pulling the argument out of conversations and drafting slides that have never existed. This is the largest single factor in the range.

How much is drawn rather than laid out. A deck of type, photographs and clean charts moves quickly. A deck that needs a product diagram, a market map, a process schematic or an architecture illustration slows down, because each of those is an original drawing that has to be made comprehensible at a glance from the back of a room.

Which tool it has to live in, and who edits it next. A deck that only ever exports to PDF can use anything. A deck your team will edit weekly has to be built inside their tool, with real master slides and styles, so that the next edit does not quietly dismantle the design. That constraint costs time and it is worth paying for if the deck has a life after the meeting.

The three failures we watch for

Forty slides become forty beautiful slides. You send what you have and it comes back with the same structure, the same seven lines a slide, now in a nicer typeface. The cutting is the work being paid for. A deck that has not lost weight has not been designed, whatever it looks like — and the founder still has no idea which three slides carry the pitch.

A number gets retyped wrong. Charts and figures are transcribed by hand at some point in every deck project, and a transposed digit, a mislabelled axis or a growth figure quoted for the wrong period is caught instantly by exactly the audience you cannot afford to lose credibility with. Worse, the embarrassment lands on you in the room, not on the designer. Every figure on every slide should be traceable back to the source you supplied, and where two of your own sources disagree, the right move is to stop and ask.

The deliverable cannot be edited when it matters. A gorgeous PDF and a design-tool link, handed to a founder who needs to change the ask figure ninety minutes before the meeting, from a hotel, on a laptop that does not have the fonts. The question of who edits this next has to be asked before the first slide is designed, because it decides the tool, the typeface choices and how the file is built.