An AI model does not know what any human wants better than a human. It predicts. It has seen more examples than I will ever hold in my head, so it can guess the next click, the average preference, the likely headline. That is useful. It is not knowledge.

People do not fully know their own wants either. They change, contradict themselves, say one thing and buy another. The model is a sharp statistician of that mess. It is not the person in the situation. Treat it as a forecast, not a verdict.

Design beyond the object

That same gap is why an experience designer beats a graphic designer who only designs the object when the work has to reach different audiences.

Graphic design, practiced the old way, starts with the artifact: grid, type, brand, “make it feel right.” After enough years, you learn that “it looks right” is what clients say when they cannot tell whether it works.

Audiences are not palettes. They are people under constraints: time, language, device, distrust, bad light, fat fingers, a form they did not ask to fill.

Experience design starts there: the job to be done, the journey, the test, the edge cases. Different audiences are not a color swap. They are different first questions. The craft still finishes the work. Research and use decide whether the craft is aimed at anyone but us.

AI as a cheap rehearsal

Partnering with AI applications sits in the middle of those two facts. AI does not give me better taste, and it does not give me the audience’s inner life. It gives me volume and a cheap rehearsal.

Variations before I pick. Captions and proposals at the speed of the layout. Tags and culls for photographs. A rough stand-in for the user I have not met yet: “Explain this to someone who is not a designer.” Video and mockups a one-person shop used to need a team for.

That is how retainers stop being discounted hours. The application takes the repetitive middle. I keep the brief, the cut, the politics of the room, and the last look.

Three principles

Do not outsource wanting.

Do not design only for the file.

Do not confuse generation with judgment.

Use the model to see patterns and make options. Use experience design to find out whether those options survive contact with real people. Use thirty years of graphic craft so the thing that survives is not only usable but worth putting in front of them.

The product is not the picture, the prompt, or the persona slide. The product is whether the right people can do the right thing without me standing there explaining it.

Everything else—AI included—is staff. I am still the art director.

I want to argue the opposite. Accessibility is not a toll booth. Approached from the start, it is nearly free — and it makes the publication better for every single reader, not just the ones the requirement names.

What it actually means

Strip away the standards jargon and an accessible publication comes down to a handful of concrete things. The document has a real structure — headings that are actual headings, not just big bold text, so a screen reader (and a search engine, and a skimming human) can navigate it. It has a logical reading order, so content flows sensibly when read aloud or reflowed on a phone. Images and charts carry alt text — a written version of what the visual shows, for readers who cannot see it. Color never works alone — anything meaning is attached to also differs in lightness, label, or pattern, for the roughly one-in-twelve men with color-blindness and for everyone reading a grayscale printout. Text contrast is strong enough to read in bad light and small sizes. And in the final PDF, all of that is preserved as tags — the invisible layer assistive technology actually reads.

Notice what that list is not. It is not "make the document ugly," and it is not exotic. It is mostly what careful design does anyway, done deliberately and kept intact through to the final file.

The quiet secret: everyone benefits

Here is what the checkbox framing misses. Every one of those requirements improves the document for readers no requirement covers.

Real heading structure is what makes a long report navigable — the same layering that serves the four-minute reader. Alt text discipline forces the team to ask, of every figure, "what is this actually showing?" — a question that kills decorative charts and sharpens real ones. Color-independent design is what survives the projector in a bright room, the photocopier, the phone screen in sunlight. Strong contrast serves every reader over forty-five, which in the audiences of large organizations is a substantial share of the decision-makers. The delegate skimming your report on a phone in the back of a car benefits from the same reflow logic as the screen-reader user.

Accessibility, in other words, is a stress test. A document that passes it is simply better built.

Why retrofitting costs so much

The economics deserve to be understood, because they explain most of the frustration. Accessibility built in from the start is a set of habits: styles used properly, alt text written when the figure is made, colors chosen against a contrast check, tags exported with the file. Marginal cost: small.

Retrofitting inverts all of it. Someone must reverse-engineer the structure of a finished document, write alt text for a hundred figures they didn't make, re-solve color decisions that are now baked into every chart, and hand-tag a file that was never built to be tagged. For a long publication that can approach the cost of the original layout — the same reason a building gets ramps in the blueprints, not carved into the marble.

So the single most useful thing a publishing team can do costs nothing: put accessibility in the brief, not the launch checklist. Two lines — "final PDF must be accessible; figures need alt text" — written in week one instead of discovered in week twelve.

The standard is the floor, not the design

One honest caveat: a technically compliant document can still be a poorly designed one, and chasing the checklist alone produces exactly that. The point is not to pass a checker. The point is the readers the checker represents — and all the other readers who happen to benefit. Organizations whose work is meant to reach everyone have a particular reason to care that their publications literally can.

Build it in from the start and no one will ever notice — which is the highest compliment accessibility gets. The document will simply work: for the screen reader, the phone, the photocopy, the tired eyes at the end of a long meeting. That is not a checkbox. That is the job.

You can recognize these decks instantly. Forty slides, each individually reasonable. Background, context, methodology, findings, more findings, considerations, next steps. Everything the team knows is in there. And if you ask the person who built it, "what is this presentation arguing?" — there is a pause. That pause is the problem. Everything else is symptoms.

One idea, one sentence

Every presentation that works in the room can be compressed into a single sentence with a verb in it. Not a topic — "an update on the regional program" is a filing label, not an idea. A claim: The program is working in three of four countries, and the fourth needs a different model. We should fund the expansion now, because the window closes in March. The risk has changed, and our response hasn't.

That sentence is the presentation. The slides are its evidence, pacing, and delivery mechanism. Which is why the most valuable minutes in any deck project happen before slide one exists, when someone finally answers: if the audience remembers only one sentence next week, what must it be?

Senior audiences make this non-negotiable. As I wrote in the executive-ready piece, they are busy, between decisions, often reading the deck cold before the meeting. They are not going to assemble your argument from forty slides of ingredients. Either the deck hands them the idea, or they leave with whichever idea they assembled themselves — and it may not be the one you needed them to take.

The clear idea is a design tool

Here is what teams underestimate: that one sentence is not just rhetoric. It is the instrument that makes every design decision easier, and most of them automatic.

It decides what leads. The idea goes early and unmissable — a title slide that states it, an opening page that owns it. Not "buried at slide 32 as the reveal." Executives are not an audience you surprise; they are an audience you orient.

It decides what stays. Every slide faces the same test: does this advance the sentence? The methodology appendix everyone fears cutting, the six context slides, the third chart making the same point — the sentence settles those arguments impersonally. Forty slides become eighteen, and nobody had to win a turf war; the idea did the cutting.

It decides emphasis. When a slide has a job in the argument, you know what its headline should say, which number gets the color, what can sit quietly in the corner. When it doesn't, everything on it seems equally important — which is how you get the slide with four boxes, three arrows, and no point.

It even decides the room. A deck built on one clear idea produces a discussion about that idea — which is what you wanted. A deck built on coverage produces forty minutes of clarifying questions, and the decision deferred to a follow-up. Most "we need another meeting" outcomes were designed in, slide by slide, by a deck that never took a position.

Finding the sentence

If the sentence will not come, that is diagnostic, and it is worth being honest about which case you are in. Sometimes the thinking genuinely isn't finished — the deck is being asked to hide that, and no deck can. Sometimes the team knows the idea perfectly well but is hedging, hoping the audience will say it first so nobody has to own the recommendation. And sometimes the idea is there, buried under everything the team also wants credit for knowing.

Only the third case is a design problem, and it is the one I can help with. It is real work — reading the material, asking what the meeting is actually for, drafting the sentence and testing it against the owner's instinct. But it is the work that makes the rest almost mechanical. Structure, hierarchy, emphasis, charts: they all have something to serve now.

So before the template, before the brand colors, before anyone opens the software: one sentence, with a verb, that the room must remember. Get that, and the design has a job. Skip it, and the most beautiful deck in the world is forty well-dressed slides in search of a point.

It is not magic. Having produced hundreds of these transformations, I can tell you the honest secret: it is almost always the same seven moves, applied in roughly the same order. Once you know them, you can see them in every convincing redesign — and spot which ones your own documents are missing.

1. A decision about what matters most

The "before" page's real problem is rarely ugliness. It is equality: fourteen elements, all the same size, all shouting at the same volume, because nobody decided what the page is for. The first move happens before anything is touched — deciding the one thing a reader should take from this page. Every other move executes that decision. This is why a designer who asks "what should the reader do with this?" is doing the job, and one who asks "what colors do you like?" is decorating.

2. A real hierarchy

With the decision made, the page gets levels: one dominant element, a clear second tier, and everything else deliberately quiet. Type sizes stop being eleven slightly different values and become three or four committed ones. The reader's eye now has a path — first this, then this, then details if they want them — instead of a parking lot.

3. An honest grid

In the "before," elements sit where they happened to land. In the "after," everything aligns to a small number of invisible lines. Nothing about the content changed, but the page suddenly reads as intentional — and readers transfer that judgment to the organization behind it. Alignment is the cheapest credibility available anywhere in design.

4. Breathing room, taken from nothing

The most counterintuitive move: the "after" almost always has more empty space, yet fits the same content. The space comes from consolidation — three boxes saying one thing become one box, a paragraph restated in the chart's title disappears, decoration that carried no information goes. Crowding is usually redundancy wearing a layout problem's clothes.

5. One emphasis, spent well

The "before" tends to emphasize everything: bold here, red there, three boxes with borders, a highlight per paragraph. The "after" spends emphasis like money — one accent, placed on the one element from move #1. When only one thing on the page is loud, the reader hears it.

6. Charts rebuilt as arguments

If the page has data, this is where the transformation is most visible: the pasted spreadsheet chart with its gridlines, legend, and eight competing colors becomes a redrawn figure with a finding for a title, direct labels, and one highlighted series. Same numbers; the difference is that the "after" chart has an opinion about what the numbers mean.

7. Fewer, better choices everywhere

Last, the quiet cleanup that no one consciously notices and everyone feels: two typefaces instead of five, one style of icon instead of a collection, consistent spacing between things that repeat. Every choice removed is one less thing the reader has to process — the page gets calmer, and calm reads as competence.

Why this matters for your documents

Notice what is not on the list: talent-dependent flourishes, expensive imagery, trends. The seven moves are craft, not inspiration — which is exactly why they can be done reliably, on a deadline, to a deck due Friday. It is also why a before-and-after is the fairest way to evaluate a designer: not "can you make something beautiful from scratch," but "hand them a real, crowded, honest page — what do they do with it?"

And it suggests a useful self-test. Put your team's most important recent page at arm's length and ask: is there one clearly dominant element? Do things align? Is anything emphasized — or everything? You will know within ten seconds which side of the before-and-after you are looking at.

If it is the left side, that is a fixable condition — usually in days, not months. It is precisely what Design Rescue exists to do.

I want to explain why, using a project I am proud of: updating the visual identity of the Pan American Health Organization, whose emblem was created in 1948.

The wrong diagnosis

When an organization's materials start to feel dated, the symbol takes the blame, because the symbol is the thing everyone can point to. But look closer and the tiredness almost never lives in the mark itself. It lives in everything around it: a palette too narrow for the work the organization now does, typography that was never chosen so much as inherited, and decades of documents, decks, and web pages each drifting a little further from the last.

That misdiagnosis produces the classic failed rebrand: an expensive new mark, launched with fanfare, that solves none of the day-to-day problems — and destroys something valuable in the process. Because here is what a 75-year-old symbol has that no new mark can buy: recognition. Ministries, partners, and the public have seen it on documents, buildings, and vaccination campaigns for generations. That trust took decades to accumulate. A rebrand that throws it away to look contemporary trades the one asset that cannot be repurchased for the one quality that fades fastest.

What was actually wrong

At PAHO, the emblem did need care — a symbol drawn in 1948 needed an updated look and feel to work in places its designers never imagined: favicon-sized on a phone, embroidered on a field vest, animated in a video. But the deeper constraint was the color system. The organization was working in a single royal blue. One color, carrying everything: a scientific report, an emergency response, a public campaign, a youth program. Everything the organization made necessarily looked the same, which in practice meant everything competed for the same visual note and nothing could signal what kind of communication it was.

Renew the system, keep the soul

So the work ran the opposite of the classic rebrand. The symbol — the part with 75 years of trust in it — was respected and refined: redrawn to hold up at today's sizes and surfaces, its character intact. And the system around it, the part that was actually failing, was rebuilt. I introduced the orange and light blue the organization uses today alongside its blue — and that expanded palette did what a new logo never could. It gave the organization range: room to distinguish an emergency alert from an annual report from a community campaign, while every piece remains unmistakably the same house.

Someone who had known the organization for forty years saw continuity. Someone encountering it for the first time saw an organization that felt current. Both were seeing the same identity.

The pattern behind failed rebrands

With that story in mind, the common failure modes are easy to name:

Rebranding the symbol when the system is what's broken. The most expensive way to not solve the problem.

Chasing what looks current. Contemporary style has a shelf life of a few years; a symbol has to live for decades. Design toward the organization's character, not the year's fashion.

Launching a logo instead of a language. A mark without templates, palette roles, and typography guidance just starts a new cycle of drift — now with a logo nobody recognizes yet.

Skipping the rollout. An identity only exists in the materials people actually touch. If the decks, reports, and web pages don't change, the rebrand happened in a PDF and nowhere else.

Redesigning without a reason. New leadership wanting a new look is a motive, not a diagnosis. The identities worth changing are the ones failing at a job you can name.

The question to ask first

If your organization is circling a rebrand, ask one question before anyone sketches: what is the identity failing to do? If the answers are about range, consistency, and dated surroundings — as they usually are — you likely have a system problem wearing a logo problem's clothes. That is genuinely good news. It means the trust your symbol has earned can be kept, and the parts that are failing can be rebuilt around it.

The organizations that get this right end up where PAHO did: renewed without amnesia. Still themselves — just able, finally, to speak at full range.

The difference is not decoration. When every chart speaks in a different visual dialect, the reader spends energy re-learning how to read each one, and quietly starts to wonder how coordinated the work behind the report is. When the charts speak with one voice, the data itself feels more trustworthy. After three decades of redrawing charts for development banks and health agencies, I can tell you the standards that matter are few, and none of them require talent to follow. They require deciding once.

1. One chart, one message

If a chart needs three sentences to explain what it shows, it is two or three charts sharing one box. The discipline: before designing any figure, write the single sentence it exists to prove. If you cannot, the problem is upstream of design.

2. Put the message in the title

"Figure 12: Enrollment rates by region, 2015–2025" describes the axes. "Enrollment doubled everywhere except the regions that needed it most" states the finding — and it is what the skimming reader takes away whether or not they study the bars. Keep the descriptive part as a subtitle for the citation-minded. This single rule improves a report more than any palette ever will.

3. Label directly, kill the legend

Legends force the reader's eye to ping-pong between the data and a decoder key. Whenever geometry allows, put the label on or next to the line or bar itself. If a chart has so many series that direct labeling is impossible, that is usually rule 1 failing, not rule 3.

4. One palette, with jobs assigned

A chart palette is not "the brand colors, used freely." It is a small set with fixed roles: a sequence for categories, a ramp for intensity, one emphasis color reserved for the finding. The emphasis color is a spending decision — use it on the one bar, one line, one number the chart exists to show, and nowhere else. When everything is highlighted, the reader correctly concludes that nothing is.

5. Design for the color-blind reader — and the photocopier

Roughly one in twelve men cannot reliably distinguish red from green, and a surprising share of your report's life is lived in grayscale printouts and low-quality projections. Never encode meaning in hue alone: vary lightness, add direct labels, use patterns where needed. Print the chart in grayscale; if the story survives, it passes.

6. Redraw, never paste

A screenshot pasted from Excel or a partner's PDF announces itself instantly — wrong fonts, wrong colors, fuzzy at print resolution — and it cannot be corrected when a number changes. Every figure in a serious publication should be rebuilt in the document's own system from the underlying data. This is the most labor-intensive standard on the list and the one with the most visible payoff.

7. Let tables be tables

Not everything wants to be a chart. Precise values the reader will look up, many categories with no shape to reveal, audit-style completeness — these are jobs for a well-designed table: clear hierarchy, aligned numerals, generous row spacing, no prison-grid of lines. A clean table outperforms a heroic 14-series chart every time the reader's question is "what exactly is the number."

8. Start axes honestly

Bar charts start at zero — a truncated bar axis exaggerates differences and hands your critics a free shot at the whole report's credibility. Line charts may zoom when the meaningful variation demands it, but say so on the axis. For an organization whose product is trusted evidence, this is not a style rule; it is a reputational one.

Making standards stick

A standards document nobody opens changes nothing — I have seen beautiful 60-page chart manuals govern precisely zero charts. What works is making the right way the easy way: real templates in the tools authors actually use, a small library of pre-built chart styles, and one pass of professional redrawing at the end of each major publication to catch what slipped through. The manual can be four pages if the templates are good.

That end-of-cycle redrawing pass, by the way, is where most of my report work begins — a hundred author-made figures arriving in six dialects, leaving in one. If your next publication is heading that direction, it is a very fixable problem.

Once you accept that, everything about how the document should be designed changes. The goal stops being "make people read it" and becomes "make sure every kind of reader finds their part fast, and trusts what they find." Long technical reports fail when they are designed for a devoted cover-to-cover reader who does not exist, instead of the four readers who do.

The four readers

The four-minute reader — a minister, a director, a board member. They will read the executive summary, look at two charts, and form the judgment that matters most to your funding. The chapter reader — a specialist who goes straight to chapter five because that is their sector, and judges the whole report by it. The evidence hunter — a researcher or journalist who arrives from a search engine looking for one number, one table, one citation. The forwarder — the person who will not read much but decides who else receives it, based mostly on how credible it looks.

A readable technical report is one where all four succeed. None of them requires shortening the report. All of them require structuring it.

Layer the reading

The instrument for this is layering: the same content, accessible at three depths.

The first layer is the executive summary — and it has to be a genuine argument, not a compressed table of contents. "Chapter 2 discusses financing mechanisms" tells the four-minute reader nothing. "Current financing covers a third of the need, and two policy changes would close most of the gap" tells them everything, and invites them deeper. Give the summary real design attention: its own visual pacing, its key numbers made unmissable. It is read ten times more than anything else in the document; it should not look like page filler before the real report starts.

The second layer is the skimming layer, and it is where design earns its fee. Section headings that state findings instead of topics — "Coverage doubled after 2020," not "4.2 Coverage trends." Pull-quotes and callout boxes that carry the argument. Charts titled with their message. A reader flipping pages should be able to reconstruct the report's argument from headings, callouts, and figure titles alone. Test this literally: flip through and read only those. If the story does not hold, the layer is not built yet.

The third layer is the full text — which can now be as dense and rigorous as the subject demands, because the design is no longer asking it to also be the navigation.

Navigation is not the page number

In a 200-page document, the table of contents is a load-bearing element and usually the most neglected page in the file. It deserves findings-style entries and enough typographic hierarchy that the chapter reader can locate their chapter in seconds. Add running headers that say where you are, chapter openers that summarize what is coming, and — in the PDF, where most reading actually happens — working bookmarks and internal links. A 200-page PDF without bookmarks is a filing cabinet with the drawers glued shut.

Density is not the enemy — uniformity is

Teams often assume a readable report means fewer words per page. Not quite. What exhausts readers is unrelieved sameness: page after page of identical gray text with nothing for the eye to anchor to. A dense page next to a chart, a breathing spread at a chapter break, a callout that resets attention — rhythm, not dilution. The technical content stays intact; the experience of moving through it changes completely.

The same goes for the material everyone is tempted to cut: the tables, the annexes, the methodology. The evidence hunter came precisely for those. Design them properly — consistent table styles, readable at their real size, numbered so they can be cited — instead of exiling them to an unformatted afterthought. Annexes are where a report's credibility lives for the most demanding readers.

What this looks like in practice

When a long report comes to me, the work is roughly this order: fix the executive summary until it stands alone; rewrite the heading system into findings; build the figure and table styles once and apply them everywhere; design the chapter openers and navigation; then let the full text be itself. Not one page of content gets dumbed down. Every layer above the content gets built.

That is the real answer to the title. You do not make a 200-page report readable by making it shorter. You make it readable by making it navigable — so that each of its real readers, in their four minutes or their four hours, finds the report they came for.

A conference visual identity sounds like one thing. In practice it is thirty or forty things: the mark itself, then banners, stage backdrops, badges, lanyards, program books, signage and wayfinding, session slides, social cards, email headers, the registration page, and the report that comes out after. They share one look but not one deadline — print has lead times, venues have submission dates, and speakers send their slides whenever they feel like it.

The events that look effortless are the ones that treated this as a countdown, not a design project. Here is the version I run, working back from event day.

Weeks 8–7: Decide the system, not the pieces

This is the week for the identity — but the deciding question is not "what should the logo look like." It is "what will this look like on a badge, a backdrop, a slide, and a phone screen." A mark that only works at one size is a liability with six weeks to live.

Two other decisions belong here because everything downstream waits on them: languages (a bilingual event changes every single piece) and co-branding — which partner logos appear, in what order, at what size. That question looks trivial and can take two weeks of diplomacy. Start it now, not on the signage deadline.

What you should have by end of week 7: the identity, a one-page usage sheet, and templates for the big recurring formats — slide deck, social card, A4 page.

Weeks 6–5: Print and the venue

Now the pieces with physical lead times: banners, backdrops, badges, printed programs, signage. Printers need files days or weeks ahead, venues have their own art-submission specs and cutoffs, and both punish late changes with rush fees or worse.

The discipline that matters here is a piece list with owners and dates — every physical item, who supplies its content, when the printer needs the file. It is the least glamorous document of the whole project and the one that saves the event. Content that is still "being finalized" in week 5 needs a decision: cut it, or design around a placeholder that can absorb the late text.

Weeks 4–3: The digital layer

Registration page, email headers, social announcement cards, the event's slide template pushed out to speakers and moderators. Send that template earlier than feels necessary — speakers prepare late, and every week of delay converts future session slides from "on brand" to "whatever they had."

This is also when the wayfinding gets a reality check: someone walks the venue plan and asks where a confused delegate actually stands and looks. Signage designed at a desk always needs this pass.

Week 2: The mop-up

Whatever program content was late arrives now, and the last files go to print. Protect this week for corrections — do not let anyone spend it on a new idea. A change to the badge design in week 2 is not a refinement; it is a risk transferred to the printer.

Week 1 and event days: Live support

Something will change on site — a speaker cancels, a room moves, a minister confirms at the last minute and needs a name card and a program insert. The difference between a scramble and a non-event is having someone who can produce corrected, on-brand pieces in hours. Build that into the plan: the designer's job does not end at file handoff.

After: The pieces everyone forgets

The report of proceedings, the thank-you graphics, the photo assets cropped for each channel — decided before the event, they cost little; discovered after, they arrive a month late and off-brand, when attention has already moved on.

The honest version of "we're behind"

If you are reading this with five weeks left and no identity, the plan compresses — it does not collapse. The order of operations holds; you simply cut the number of pieces rather than the quality of the system. Fewer formats, done properly, beat a full set done badly. That triage call — which pieces the event actually stands on — is worth an experienced eye, and it is the conversation I usually have on the first call.

The visual system is the part of a conference every delegate touches, from the email that invited them to the sign that gets them to the right room. Eight weeks is enough. Eight weeks minus the co-branding argument is not — so have that argument first.

Here is the honest comparison — including where each one fails.

The freelancer: buy a pair of hands

A freelancer on an individual contract is the right purchase when the work is defined, bounded, and singular. One conference poster. One deck cleanup. One infographic from data you already have. You know what done looks like, and you need someone skilled to get there.

Where it wins: price, speed of contracting (individual consultant agreements are often the fastest vehicle in the system), and direct access — you talk to the person doing the work.

Where it fails: continuity and surge. A freelancer is one calendar. When your report and your event land in the same month, or when the person is sick the week of your deadline, there is no bench behind them. And each new freelancer starts from zero on your brand, your templates, and your approval quirks — you pay that learning curve every time, in money or in corrections.

The studio: buy an outcome

A studio contract is the right purchase when the job is a project — something with phases, formats, and enough moving parts that you want one party responsible for the whole result. A flagship report. A campaign. A conference identity with everything that hangs off it.

Where it wins: accountability and range. You are not assembling a layout person, a data-visualization person, and a web person yourself; the studio is on the hook for the outcome, not the hours.

Where it fails: small tasks. Routing a two-hour fix through a project-sized contract is slow and overpriced, which is exactly why those small tasks end up done in-house at 11 pm instead. Studios are also only as good as their scope — a vague scope produces a beautiful thing that solves the wrong problem, at studio prices.

The retainer: buy a working relationship

A retainer — a fixed monthly arrangement, or a long-term agreement drawn down by task order — is the right purchase when design need is recurring. If your unit produced a report, three decks, an event, and a dozen small graphics last year, it will do roughly that again this year. A retainer stops treating each of those as a surprise.

Where it wins: three places the other models cannot reach. First, speed — no new procurement per task; a request is an email, not a contract. Second, memory — the designer already knows your brand, your templates, and which director hates pie charts, so version one arrives closer to right. Third, priority — a retainer client's Friday emergency goes to the front of the line, because the relationship is the business.

Where it fails: genuinely irregular need. If you truly need design twice a year, a retainer is paying for readiness you will not use. And a retainer with the wrong partner locks in the wrong partner — which is why the sensible path is to start small.

The pattern that actually works

The organizations that get the best design for their budget rarely pick one model forever. They follow a sequence:

1. Start with a bounded, fixed-price task — small enough to approve quickly, real enough to test how the vendor handles feedback, deadlines, and your organization's particular weather.
2. Run one full project to see whether the quality holds at scale.
3. Then move the recurring work onto a retainer, once trust is earned and the numbers make sense — usually the point where the team notices they have contracted the same vendor four times in a year and each contract cost them three weeks of paperwork.

That sequence is not a sales funnel; it is risk management. Every step is reversible, and by the time you commit to a retainer you are committing to a known quantity.

It is also, not coincidentally, how most of my own long partnerships began: with a single rescue, due Friday.

If your team is at step three — tired of re-contracting the same work — the retainer options on my services page show how I structure ongoing arrangements, and the procurement page has the paperwork your organization will ask for.

"We need our annual report designed, approximately 80 pages, deadline November." True, and almost useless. I do not yet know who has to be convinced by this report, what happened the last time it went out, or which chart the director already hates. Those are the things that shape the design.

A useful brief is not a form. It is five answers. You can write them in an email in twenty minutes, and they will save weeks.

1. Who is this for, and what should they do after reading it?

Not the mailing list — the reader you most need to reach. A donor deciding whether to renew. A board approving a budget. A minister's advisor who will spend four minutes with the executive summary. Design for "everyone" produces documents for no one; design for that one reader produces documents everyone else can follow too.

And say what the reader should do next. "Understand our impact" is decoration. "Renew the grant," "approve phase two," "invite us to bid" — those are directions a design can aim at.

2. What is the real deadline, and what is it attached to?

Not the date you would like the files. The event the files exist for — the board meeting, the launch, the conference. That tells me which corners can never be cut and which intermediate dates are actually flexible. A designer who knows the report is for the Annual Meetings plans differently than one who only knows "end of October."

3. What state is the content in, honestly?

"The text is basically final" has cost more teams more money than any other sentence in this business. It is fine if chapters are still moving — layout can be sequenced around that — but only if I know. Say what is final, what is in review, and what does not exist yet. No judgment; sequencing is my job. Surprises are the only expensive part.

4. What are the fixed constraints?

Brand rules that cannot bend. Logos that must appear and in what order — with co-published reports this alone can take a week of back-and-forth, so settle it early. Formats you need at the end: print, web, accessible PDF, social cards. Languages that are coming, even if "maybe French later." Constraints named in the brief are design inputs. Constraints discovered in week five are rework.

5. Show one example you like — and one you don't

This is the question most briefs skip, and it is the most efficient sentence you can write. "We like how this report handles data, and we never want to look like that one" replaces three rounds of exploratory drafts. You are not doing the designer's job by pointing at examples. You are calibrating taste in one exchange instead of discovering it by trial and revision.

What to leave out

Everything about how the design should be executed. "Make it modern," "make it pop," "maybe orange?" — these feel helpful and are not, because they describe a solution without describing the problem. If there is a reason behind the instinct ("the last report felt dated, and younger donors ignored it"), give me the reason. The reason is useful forever; the instruction is a guess.

The twenty-minute test

Before you send a project out, check whether your brief answers the five questions: who must it convince, what event is it attached to, what state is the content in, what constraints are fixed, and what does good look like to you. If it does, any competent designer can start well — and quote accurately, which protects you as much as it protects them.

If you would rather answer them out loud than in writing, that works too. It is the first thing I ask on every call anyway.

Two reports with the same page count can differ in cost by a factor of five. That is not a designer being mysterious about pricing. It is that page count is one of the weakest predictors of design effort. After thirty years of doing this work for development banks, health agencies, and NGOs, I can tell you what the strong predictors are. There are six.

1. The number of figures, not the number of pages

A page of running text takes minutes to lay out well. A page with a complex chart can take hours, because the chart usually arrives as a screenshot from Excel and has to be rebuilt: redrawn in the report's visual system, its labels edited for readability, its point made visible instead of buried.

When you count your report, do not count pages. Count charts, tables, maps, and diagrams — and note how many exist as editable data versus pasted images. That number drives the quote more than anything else on this list.

2. How final the content is

Designing a finished manuscript is one job. Designing a manuscript while chapters are still being written is a different, more expensive job, because every content change after layout begins means rework. A paragraph added to chapter two can ripple through forty pages of a laid-out document.

You do not need perfectly frozen text to start — covers, visual systems, and divider pages can begin early. But the closer your manuscript is to final when layout starts, the less you pay for the same result.

3. Revision rounds, and who gives the comments

Two consolidated revision rounds is a healthy norm. What multiplies cost is not the number of rounds but the number of voices: when comments arrive separately from five reviewers across three departments, someone has to reconcile the contradictions, and if it is the designer, you are paying design rates for editorial diplomacy.

The cheapest thing you can do for your budget is appoint one person to consolidate comments before they reach the designer. One document, one voice, per round.

4. Languages

A second language is not a copy-paste. French and Spanish run roughly 15–25% longer than English, which means text reflows, charts need relabeling, and every page needs review again. Arabic flips the entire layout right to left. A trilingual report is closer to three reports than to one report with extra text.

If translations are coming, say so in the first conversation — the visual system can be designed to absorb them, which is far cheaper than retrofitting.

5. Formats and accessibility

"The report" is rarely one file anymore. A print PDF, a web PDF, an accessible tagged PDF that passes your organization's compliance check, social media cards, a summary deck for the launch event — each is real work. Accessibility in particular is not a checkbox at the end; a document built for it from the start costs modestly more, while one retrofitted after the fact can cost as much as the original layout.

List every format you need before asking for a quote. The list changes the number, and finding out in week six changes it more.

6. The timeline

Rush work costs more everywhere, and design is no exception — not as a penalty, but because compressing six weeks into two means other work moves aside and more hands join the job. If your report has a fixed launch date, the cheapest schedule is the one that starts earliest.

So what does it cost?

With those six factors defined, an honest range: a modest report — say 40 pages, a dozen charts, one language, two rounds — sits at the lower end of a few thousand dollars. A true flagship — 150+ pages, heavy data, multiple languages and formats, accessibility compliance, launch materials — is a five-figure project at any studio doing it properly. Anyone quoting a flagship at pamphlet prices is planning to template it, and it will read that way.

My own rates are public — the rate card is on the procurement page, along with the vendor paperwork your organization will ask for. Fixed quotes come from the six answers above, which means you control most of the number before I ever see the files.

That is the real point of this article. The cost of a flagship report is not set by the designer. It is set by decisions the team makes early — usually without knowing they are pricing decisions at all.

I have spent years on the vendor side of procurement at development banks, UN agencies, and large NGOs. The pattern is almost always the same: a team has a report, a presentation, or an event coming up. They know they need design help. And then they discover that hiring anyone — even for a small job — means navigating vendor registration, budget codes, approvals, and contract paperwork, usually with a deadline getting closer every day.

The good news is that the process is predictable. If you know what procurement needs before you start, you can compress weeks of back-and-forth into a few days.

What procurement actually needs from a design vendor

Every organization has its own forms, but the underlying checklist is nearly universal. Before your procurement or operations team can issue anything, they will need the vendor to provide:

1. Legal registration. The vendor's legal business name, address, and registration documents. In the United States, that usually means a W-9. For US federal work, it can also mean an active SAM.gov registration.
2. Banking details. Where payments go, on official letterhead or through the organization's vendor portal.
3. A rate card. Daily or hourly rates, or fixed prices for defined work. Procurement teams strongly prefer vendors who can show their pricing in writing before the negotiation starts.
4. References or past work. Evidence that the vendor has done this kind of work before, ideally for organizations like yours.
5. Insurance or compliance documents, depending on the contract size and your organization's rules.

If you are the person hiring, ask for all of this in your first email. A vendor who can send it back the same day is a vendor who has done this before. A vendor who needs two weeks to produce a rate card will need two weeks for everything else too.

Choosing the right contract type

The contract vehicle matters more than most teams realize, because it determines how fast you can start.

A purchase order or small-value contract is the fastest route. Most organizations have a threshold — often somewhere between $10,000 and $25,000 — under which a manager can approve a contract directly, without a competitive process. If your design job fits under that threshold, define it clearly and keep it there. One report, one presentation, one event package.

An individual consultant contract works when you need a person embedded in the team for a stretch of time. It is flexible, but it usually caps how the person can work and bill, and it ties the work to one individual rather than a studio.

A long-term agreement or retainer is the right tool when the need is recurring. If your unit produces publications, decks, and event materials all year, one negotiated agreement with agreed rates means every new request becomes a task order instead of a new procurement. The setup takes longer once; everything after it is fast.

If you are not sure, start with the small-value contract. You can always graduate to a retainer after the first project proves the fit.

Writing a scope that protects your deadline

Vague scopes cause slow procurement and slow projects. The fix is to write the scope in terms of what you will receive, not what the vendor will do.

Weak: "Vendor will provide design support for the annual report."

Strong: "Design and layout of a 60-page annual report, including cover, typographic system, up to 15 charts redrawn from supplied data, two rounds of revisions, and final print-ready and web PDF files, delivered by October 15."

That second version does three jobs at once. Procurement can price it. The vendor can plan it. And you have something concrete to hold everyone to when the deadline approaches.

Two more things belong in every scope: who supplies the content and when (design cannot start without it), and how many revision rounds are included (unlimited revisions is how projects miss deadlines).

The timeline, honestly

For a first-time vendor at a large organization, plan on this:

• Vendor registration: 3–10 business days, depending on the portal
• Contract or purchase order: 3–15 business days, depending on the value and approvals
• The design work itself: often the fastest part

This is why the smartest teams register their design vendor before the urgent project appears. If the vendor is already in the system with a signed rate card, an urgent request can move in days instead of weeks.

If the deadline is already here

Sometimes there is no time for any of this — the board meeting is Friday and the deck is not ready. For those situations I built Design Rescue: a fixed-price, fixed-timeline service that takes materials from rough to ready to present in 3–5 business days, with the paperwork kept deliberately simple so procurement can say yes quickly.

Either way, the principle is the same. Procurement is not the enemy of good design. Unclear scopes, missing paperwork, and late starts are. Handle those early, and the rest is just the work.

A report, presentation, proposal, or website can have all the right information and still fail to communicate clearly. Not because the idea is weak, but because the idea is buried under too much text, too many charts, too many messages, or a structure that does not guide the reader.

Good design creates that guide. It gives information a clear path. It helps people know where to look first, what matters most, and how each piece connects to the next. It removes noise. It creates order. It turns a page from something people have to decode into something they can follow.

This is especially important for teams working with research, policy, programs, funding, technical content, or data. The work may be meaningful, but meaning can get lost when everything is presented with the same weight.

Design helps make the message visible. It can turn a dense set of findings into a clear visual explanation. It can turn a complicated process into a simple sequence. It can turn a presentation from a collection of slides into an argument people can actually follow.

That is why design is not decoration. Decoration is added at the end. Good design is part of how the idea is understood. It helps a team explain their work with more clarity, more confidence, and less friction. It helps the audience see what matters without getting lost in the details.

Sometimes the problem is not the idea. The problem is that the idea needs a better way to be seen. That is the part of the design I care about most. Not making things look fancy. Helping people understand.

The content exists. The team knows what they want to say. The deadline is real — a board meeting, a donor presentation, a conference, a launch. But the document itself is not ready to show. The deck has forty slides in six different styles. The report reads well but looks like a draft. The web page explains everything and convinces no one.

Design Rescue is the service I built for exactly that moment. Here is what actually happens when you start one, so you know what you are buying before you buy it.

What qualifies as a Design Rescue

Design Rescue works on materials that already have their content: presentations, proposals, reports, infographics, one-pagers, and website pages. The words and data exist; the design is what needs help.

It is not the right fit for projects that still need writing, research, or a brand built from scratch. Those are real projects with real timelines, and pretending they fit in five days helps no one. When someone sends me one of those, I say so on the first call and we scope it honestly.

The test is simple: if you could present the material tomorrow if only it looked right, it qualifies.

Day by day

Day 1 — Handoff and triage. You send me the files and tell me two things: when it is due, and who is going to see it. That second question matters more than people expect. A deck for a board reads differently than a deck for a technical working group, and the design decisions follow from the audience. By the end of day one, you have a confirmation of exactly what will be done and when you will see it.

Days 2–3 — The rebuild. This is where the work happens. For a presentation, that means one typographic system, one color logic, one chart style, and a structure that lets a busy reader scan the deck and understand the argument. For a report, it means a real layout: hierarchy, spacing, figures that make their point without a paragraph of explanation. I do not decorate what is there — I reorganize it so the message is easier to find and easier to trust.

Day 4 — Review. You get the redesigned material and mark it up. One focused round of revisions is included, and one is almost always enough, because the structural decisions were confirmed on day one. Revision rounds multiply when nobody agreed on the direction early; the triage call exists to prevent that.

Day 5 — Final files. You receive everything in the formats you actually need — an editable deck your team can keep working in, a print-ready and web PDF, source files where that makes sense. The goal is that you are never locked out of your own materials.

Smaller jobs finish in three days. Bigger ones use all five. Either way, the date I give you on day one is the date you get it.

What it costs

Design Rescue is a fixed price, published on the site, agreed before work starts. No hourly meter, no estimate that grows halfway through. Fixed pricing is not just friendlier — for teams inside large organizations it is often what makes the purchase possible at all, because a fixed price under the small-contract threshold can usually be approved by a manager directly, without a longer procurement process.

If your organization needs vendor paperwork — a W-9, banking forms, a rate card, registration in a vendor portal — I have all of it ready to send the same day. I have been through procurement at development banks, UN agencies, and NGOs many times, and the paperwork should never be the thing that costs you your deadline.

What you get back

The honest promise is this: the same content, made clear. Your argument, your data, your voice — organized so that the person reading it can find the point, follow the logic, and take you seriously before you say a word.

Teams usually come back for the second rescue before the first one is a month old. Not because the emergencies stop, but because it turns out that "ready to show" is a standard worth having all the time. When that happens, we talk about a retainer, and rescues become simply how the work gets done.

If something important is due soon and the design is not there yet, that is exactly what this is for.

Credibility is visual before it is verbal

Decision-makers skim. Donors skim. Board members skim. They form an impression of competence long before they engage with the substance. When a report looks dated, cramped, or inconsistent, readers unconsciously transfer that judgment to the work itself — even when the underlying programs are excellent.

Design is not decoration. For an institution, it is the first evidence of seriousness.

The three credibility leaks

1. Inconsistent visual systems. Mismatched charts, shifting fonts, and ad-hoc color choices read as disorganization.
2. Buried impact. The most important numbers are stranded in dense paragraphs instead of being made unmissable.
3. No reading pathway. Without hierarchy, every element competes for attention and nothing lands.

What strong report design actually does

It builds a pathway through complexity — visual hierarchies that signal importance, layouts that guide the eye, and an information architecture that lets a reader with three minutes extract the critical insight without missing it.

That is the difference between a document that is filed and one that is forwarded.

Before they read every word, before they understand every service, before they decide whether to contact you, they are already forming an impression. Does this feel clear? Does this feel professional? Do I understand where I am? Can I trust the people behind this?

That first feeling matters.

A website does not need to be complicated to feel credible. In fact, many websites lose trust because they try to say too much at once. Too many messages, too many buttons, too many sections, too many visual styles. The visitor arrives and has to work too hard to understand what matters.

Clear design makes that experience easier.

It helps people understand what the organization does, who it serves, and what action they can take next. It gives the page a sense of order. It makes the content feel intentional instead of assembled. It gives the visitor a path.

This is especially important for organizations with complex work. A nonprofit, consultancy, or public-sector team may have many programs, audiences, reports, services, and priorities. The website has to carry all of that without overwhelming the person using it.

Good web design does not hide complexity. It organizes it.

The best websites feel calm because someone has made decisions. The headline is clear. The navigation is simple. The sections have enough space to breathe. The images support the message. The buttons are easy to find. The page gives people confidence instead of asking them to figure everything out alone.

That is the human side of web design. You are not just arranging content on a screen. You are helping someone feel less lost.

Trust also comes from consistency. When the typography, colors, spacing, and visual rhythm feel aligned, the website feels more prepared. The visitor may not notice the grid or the spacing, but they feel the care behind it.

A clear website tells people: we know who we are, we know what we offer, and we have made this easy for you to understand.

That is what good design can do. It can make a website feel more useful, more human, and more trustworthy before a conversation even begins.

Need a website, landing page, or digital section that feels clearer and easier to understand?

IGRAPHI designs and builds websites and landing pages built around clarity, structure, and strong visual design.

In an important meeting, that extra effort can change the room.

People are often reading quickly. They may be joining between other calls, reviewing the deck right before a discussion, or scanning it while someone is speaking. A presentation that is clear, calm, and well-structured helps them understand faster. It gives them a path through the material instead of asking them to figure it out alone.

That is what makes a presentation executive-ready. It is not about making slides look expensive. It is about making the message easier to understand, easier to trust, and easier to act on.

One of the biggest improvements you can make to any presentation is to decide what the audience should notice first. Every slide should help the reader understand the main point, the supporting evidence, and what the information means. When everything on a slide has the same visual weight, the reader has to do the sorting themselves. A strong presentation does some of that work for them.

This is especially important when the content is complex. Many presentations in large organizations are built from research, policy notes, project updates, financial information, timelines, maps, and data tables. The challenge is not that there is too much information. The challenge is that the information needs to be shaped into something people can follow.

Good design does not hide complexity. It gives complexity a structure.

A well-designed presentation also respects people’s time. Senior audiences are often busy, tired, and moving from one decision to another. They do not need every detail at once. They need to understand where they are, why the information matters, and what they are being asked to consider.

Clear titles help. Consistent spacing helps. Strong section breaks help. Simple charts help. But these things are not just visual choices. They are ways of being considerate to the person reading the deck.

That is the human side of presentation design. You are not decorating slides. You are helping someone think.

Charts and tables are often where a presentation becomes hardest to follow. The problem is usually not the data. The problem is that the data has not yet been turned into communication. A useful chart should help answer a question. A good table should make comparison easier. A strong data slide should help the audience see the point without having to decode everything from scratch.

Sometimes the best design decision is to remove something. Sometimes it is to highlight one number. Sometimes it is to separate one dense slide into two clearer ones. Sometimes it is simply to give the page more breathing room.

The goal is not to make the presentation flashy. The goal is to make people feel less lost.

Consistency also matters more than people think. When fonts, colors, icons, chart styles, and spacing change from slide to slide, the presentation starts to feel assembled under pressure. Even if the content is strong, the experience can feel uneven.

A consistent presentation feels more prepared. It tells the audience that someone cared enough to make the material easier to read. Most people will not notice the grid, the margins, or the type hierarchy. But they will feel the difference.

The best presentations also lead somewhere. They help the audience understand what needs to happen next. Maybe the goal is approval. Maybe it is funding. Maybe it is alignment, discussion, or a big decision. Whatever the goal is, it should not feel buried.

A good presentation builds toward that moment. It gives the audience enough context to understand the issue, enough evidence to trust the recommendation, and enough clarity to know what comes next.

Executive-ready does not mean over-designed. In nonprofit, public-sector, and international organization settings, restraint is often more powerful than decoration. The work needs to feel credible, calm, and useful. It does not need unnecessary effects or visual noise.

Good design knows when to stay quiet. It gives emphasis only where emphasis is needed.

A simple way to test a presentation is to ask this: can someone open it, scan it quickly, and understand the main message without you explaining every slide?

If the answer is yes, the deck is doing its job.

If the answer is no, the problem may not be the content. It may be the structure around the content.

An executive-ready presentation gives people confidence before the meeting begins. It helps them understand what matters, why it matters, and what comes next.

Need a presentation, report, or landing page cleaned up before an important meeting?

Design Rescue is a fixed-price service that makes your materials presentation-ready in 3–5 business days.