A deck that ends in a yes.

Investor decks (pre-seed through Series A is our sweet spot), sales and partnership decks, internal board and strategy decks, and the leave-behind versions of all three. The mechanics differ — an investor deck argues upside, a sales deck argues fit, a board deck argues a decision — but the discipline is identical: one argument, sequenced deliberately, every slide doing one job, evidence a reader can check.

Both, inseparably — the narrative and the design are the same job done twice badly when split. We start with the argument: interviews, your numbers, your market, until the story holds up out loud. Then the slides get built to carry that argument, with the visual system enforcing its hierarchy. A beautifully designed weak argument loses politely; a strong argument in chaotic slides loses avoidably. You are paying us to prevent both.

The presenter deck supports a human talking: sparse slides, big statements, the detail lives in what you say. The leave-behind travels without you — forwarded to a partner, reread at midnight — so it carries fuller sentences and the supporting detail on the page. Sending the sparse deck to be read alone is how great pitches die in forwarding; presenting the dense one is how audiences read ahead and stop listening. Where the stakes justify it we build both from the same system, and it is usually the best money in the project.

Often, and we will say which honestly after reading it. Decks usually fail in one of three ways: the argument is buried (a re-sequencing job), the evidence is weak or unverifiable (a content job), or the design fights the message (a system job). A rebuild targets the actual failure — we keep what earns its place, and plenty of decks need surgery on eight slides rather than a new deck. What we will not do is reskin a broken argument and call it fixed.

As long as the argument and no longer — which for investor decks usually lands between twelve and eighteen slides, not because of a rule but because arguments have natural anatomy: context, problem, solution, evidence, economics, team, ask. The length people worry about is the wrong variable; the right one is per-slide load. Ten dense slides read slower than sixteen clean ones. When a deck must shrink (a five-minute demo day), you cut whole limbs of argument, not adjectives — a shorter deck is a different deck, not a compressed one.

Consistency and restraint, mostly. One type scale actually adhered to; charts drawn in one style with honest axes; diagrams that explain rather than decorate; photography or none, but not clip art; and white space treated as a feature rather than a failure to fill. Templates feel templated because every slide is a different mood. A designed system makes slide 14 feel inevitable after slide 2 — that quiet coherence is what reads as “they have done this before”, which is often the point.

Persuasion through selection and clarity, never distortion: choose the metrics that genuinely support the argument, chart them with honest baselines and scales, and label the source. If growth looks better as a rate than an absolute, show the rate — and say so. Charts that start at suspicious zeros-that-are-not-zero, cherry-picked windows, or unlabelled units get found out in diligence, and the cost of being caught once is the whole deck’s credibility. We will also tell you when a number is not ready to be shown, which founders rarely enjoy hearing and always benefit from.

No. Every figure in the deck must be one you can defend in the room and in diligence — yours, sourced, or clearly labelled as an assumption with its basis shown. We structure and present your numbers, pressure-test the logic, and flag anything that needs a source or an accountant before it goes near a slide. A deck that survives the meeting but dies in diligence is a worse outcome than a modest deck that holds.

A conversation, your numbers, and whatever exists already — an old deck, a business plan, a data room, the back of the envelope. We interview like journalists: what you are raising or selling, who decides, what has been said before, where past pitches stalled. The material never needs to be tidy; finding the argument in the pile is the first half of the job. What genuinely accelerates things: honest numbers early, and access to the person who will present.

A sharp revision of an existing deck is a fast job. A ground-up raise deck where the story is still unsettled is not — the argument is the slow part, not the slides. Once the structure is agreed, design moves quickly, and we tell you at the start which kind of job yours is. If there is a real deadline (demo day, a booked meeting), say so first thing: honest triage of what is achievable by then beats a heroic all-nighter that shows.

The editable master in the tool you present from — PowerPoint, Google Slides or Keynote — plus a final PDF and the source assets: charts, diagrams, imagery. Nothing is locked to us; you can edit slide twelve at midnight before the meeting, and founders do. Where both versions were scoped, the presenter deck and leave-behind arrive as siblings from one system, so updating a number updates its everywhere.

Yes, and treated so by default. Raise materials and financials stay between us, we are happy to sign your NDA before you send anything, and nothing you share is reused as an example for anyone else — the sample deck on this page is fictional for exactly that reason. Real client decks stay in real client vaults.

Founders who want a designer to make a finished argument prettier — cheaper help exists for that, honestly. Anyone who needs the deck to say things the business cannot support; we do not build fiction, however well it would convert. And ideas so early there is nothing to argue yet — sometimes the honest advice is that you need three customers before you need a deck, and we have given it. If the argument exists and matters, we are the right room.

Where wanted, yes. The deck gets built around how the presenting founder actually talks — we hear the pitch out loud during the build, which changes slides more often than founders expect. Rehearsal passes, speaker notes, and the brutal-but-private question round before someone else asks the same question with money on the table: all available. The deck is a prop for a performance; we are happy to work on the performance.

Fixed, after scoping — the drivers are whether the argument exists or must be found, slide count and data density, and whether one version or the presenter/leave-behind pair is needed. A revision engagement is priced very differently from a ground-up build, which is another reason we read what exists before quoting. As with everything here: the number agreed is the number invoiced.