The definitive guide to sell sheets

Why this document gets read three different ways

Think about what actually happens to a sell sheet after it leaves your hands. Someone picks it up at a trade show table, or opens it as an email attachment between two other things they are doing. That is the scan: four seconds, maybe less, spent deciding whether this is worth anything further. If it survives, they come back to it later and actually read it, usually in a hurry, hunting for one specific fact. Does it do the thing they need, what does it cost, is there a number that proves it works. And then, whether they bought or not, it goes into a pile: a stack of papers on a desk, a folder of PDFs on a laptop, next to however many competitors' sheets they also collected. A week later they have to find yours again, and recognize it as the one they wanted, among all the others that look exactly like it.

Every design and writing decision in a sell sheet follows from those three moments, and a sheet built for only one of them fails at the other two. A sheet obsessed with looking striking in the scan but thin on substance fails the read. A sheet dense with real information but visually indistinguishable from every other sheet in the category fails the pile. The craft is building one document that survives being picked up, being read, and being found again, in that order, because a document that fails the first test never gets to the second or third.

The fixed page is the whole discipline

A website can scroll forever. A social post can be any length the platform allows. A sell sheet cannot: it has an exact page size and nothing beyond its edges exists. This sounds like a limitation. It is actually the most useful constraint in the entire format, because it forces a decision every other medium lets you avoid. What do you actually say when you cannot say everything?

The content has to fit the page; the page does not grow to fit the content. This single rule, honestly applied, solves most of what makes a sell sheet bad. When the content overflows, the honest fix is to cut content. Not to shrink the type, not to shave the margins, not to reduce the space between lines. A page set at a cramped type size with starved margins to force everything in is not a page that succeeded at fitting; it is a page that failed to be edited, and it signals exactly that to anyone who looks at it. Nobody reads a page that looks like it is apologizing for itself.

The same rule runs the other way. When the content underfills the page, the empty space needs a real job, a specification table or a proof point or a second use case, or the page should simply be shorter. A page and a half of real content stretched across two pages reads as padding, and readers notice padding faster than they notice almost anything else.

What this demands in practice: decide the page count before you write a sentence, and write to that length on purpose. One page, if the claim is sharp enough to earn attention alone. Two, only when the second page has a genuine, separate job, substantiating what page one already sold rather than repeating it.

Composing in zones

A dense page with no visible structure becomes a wall of text nobody enters. The fix is to think of the page as five zones, each doing exactly one job and nothing else.

HEADER   what this is, and the one-line claim          — this is where the scan lands
HERO     the single strongest claim or image           — the reason to keep reading
BODY     two to four blocks: outcomes, how, proof     — this is the read
DETAIL   specification table, requirements, use cases — this is the reference
FOOTER   mark, contact, one action, one URL           — this is the follow-up

The zones have to be visibly separated, through space and a genuine shift in visual weight, not through boxes drawn around every block. A page where every zone sits inside its own bordered rectangle reads as a form, not a document somebody chose to design.

And the rule that follows directly from the three-moments problem: only the top third is visible in a stack, in a file preview, and on a table someone is walking past. Everything that decides whether this gets picked up at all has to live there. What it is, who it is for, and the one claim worth remembering. A logo in the corner is not enough, because every competitor's sheet has a logo in exactly the same corner. If a second page is needed, page one has to stand completely on its own, because it is the page that gets read first and very often the only page that gets read at all. Never split one argument across the page break.

Density is a craft, not a failure to edit

A sell sheet is legitimately a dense document. That is part of the job, not a flaw in it. The actual skill is making density feel navigable instead of exhausting.

• Vary the size of blocks on purpose. Four equal-sized blocks read as a filled-in form. One large block and three smaller ones read as a page with a point of view about what matters most.
• Give every page one dominant element, even at high density. A hero number, a single strong claim, one real image. Without an obvious entry point the eye has nowhere to land, and the page reads as a flat gray field however good any individual line is.
• Put space between groups, not inside them. At high density, the gap between blocks is often the only signal telling a reader what belongs together and what does not.
• Keep one consistent internal padding across every block. Mismatched spacing inside blocks is the single most visible sign that a page was assembled from parts rather than designed as one thing.
• Alternate density down the page. Dense, then open, then dense again. A page held at one density the whole way down exhausts a reader no matter how good the content is.
• Set the type scale once and refuse to add to it. Size creep is how a dense page quietly grows from three sizes to nine, one local decision at a time, until the hierarchy has collapsed.
• Cut before you compress. The dense page that actually works is the one where a third of the planned content got removed, not the one where everything got squeezed smaller to survive.

Writing for the hunt, not the read

A reader scanning a sell sheet is hunting for one specific fact, not settling in to read prose. Design the copy for that.

• Labels beat prose. A block headed “Deployment” gets found by someone looking for deployment information. A block that opens “When it comes to deploying our product…” does not, because the word that matters is buried six words into a sentence a scanner has already skipped.
• Front-load every line. The first two or three words of a bullet are the only part a scanner reads before deciding whether to continue. “Single sign-on through any SAML provider” works. “Users can authenticate using single sign-on” buries the one word that matters.
• Bold the noun, not the sentence. Bolding half a paragraph defeats the entire purpose of bolding anything.
• Keep bullets parallel and genuinely short. Three bullets of different grammatical shapes is a paragraph wearing a disguise. Seven bullets in a row means nobody reads past the fifth; three to five per group is the real limit, not a style preference.
• Let real numbers pull the eye, and put them where you want attention to land. A stat, a count, a percentage draws the eye before any surrounding text does.

The specification table, where craft is most visible and most often missing

A comparison or specification table is the single place in a sell sheet where the gap between careful work and a rushed template is most obvious to the person reading it hardest. Remove every grid line you possibly can: a single hairline under the header row with generous spacing between rows beats a full box grid every time. Right-align numbers, left-align text, and put the unit in the column header rather than repeating it in every cell. Never center a column, because a centered column has no shared edge for the eye to follow down the page and is noticeably harder to scan than it looks at first glance. An empty cell needs a deliberate mark, an en dash or the words “not applicable,” never a blank space that reads as an oversight rather than a decision.

If the table compares you against a competitor, design the column that is yours to be visually dominant, with a subtle tonal ground or a slightly heavier weight, and do it once, cleanly, rather than stacking three separate visual tricks to make the same point. And never name a competitor in a comparison line without a sourced, dated claim behind it: the table is where a careful buyer checks hardest, which makes it the single place in the document where an invented figure does the most damage.

Surviving the format it is actually delivered in

Nearly every sell sheet is delivered as a PDF, read first on a screen, and occasionally printed on whatever printer happens to be nearest. It has to survive all three. Use a standard page size, US Letter or A4 outside the US, with real margins, keeping content clear of the edges even if a background treatment runs to the full bleed. Keep body type at a genuinely readable size; anything smaller is a decision, conscious or not, that this part is not meant to be read.

Check color choices for contrast in value, not just hue. Two brand colors that separate beautifully on a backlit screen can print as identical shades of gray on an office printer, and a sheet that only works on screen fails the moment somebody prints it to bring into a meeting. Keep text as real, selectable text rather than a picture of text, so it can be searched, copied, and read aloud by anything that needs to.

What a bad sell sheet costs, exactly

It is worth being specific about the failure, because “it did not look great” undersells it. A sheet that fails the scan was paid for in full and then never read, which means the entire cost of producing it, and the entire cost of whatever got the buyer's attention in the first place, bought nothing. A sheet that fails the pile gets read once and then lost, which is worse in a particular way: the buyer remembers the argument and attributes it to whoever they can find again.

And a sheet with an invented figure in the specification table does not merely fail. It teaches the most careful buyer in the deal, the one closest to actually signing, that your numbers are not checkable. An empty proof block is a known gap, and a buyer can decide for themselves how much that absence matters. A confident wrong number is a liability with your name on it.