You can follow PAS to the letter and still lose the reader by the second paragraph. The structure was never the problem. The words filling it were.
That gap is what the UPWORDS copywriting formula addresses. It stands for Universal Picture Words Or Relatable, Descriptive Sentences, and it comes from the direct response copywriter Michel Fortin, whose name almost never appears next to it.
It is also not a formula in the way AIDA and PAS are formulas. Those give you an order to write in. UPWORDS gives you no order at all. It governs which words fill a structure some other framework already chose, which is why so many explanations of it go wrong, and why several offer examples containing none of the technique.
What follows: Fortin’s three named techniques, a two-question test you can run on any line you have written, and one deliberately boring product rewritten five ways.
What UPWORDS Stands For, and Who Came Up With It
The acronym turns up in roundup after roundup of thirty, forty, two hundred copywriting formulas. Almost none of them name the person who invented it.
UPWORDS stands for Universal Picture Words Or Relatable, Descriptive Sentences. Fortin’s own definition: “Words that paint vivid pictures in the mind, or expressions that describe an idea to which the mind of your reader can quickly and easily relate to.”
He says he first taught it in 1998. The write-up everyone quotes secondhand went up in August 2009, eleven years later.
It is one of five frameworks he teaches:
- The 3 P’s. Pull, Prove, Push.
- OATH. Oblivious, Apathetic, Thinking, Hurting: how aware your market is of its problem.
- UPWORDS. Word choice.
- FORCEPS. Seven kinds of proof.
- The 3 Deadly C’s. Cogitation, Confusion, Complexity.
Four of those five govern structure or diagnosis. UPWORDS is the only one working at the level of individual words, and that distinction matters more than the acronym does.
Fortin’s original page is also no longer on the live web and survives only in the Internet Archive. Every current explanation is working from copies of copies, which explains a good deal about the state of them.
Why Picture Words Beat Abstract Ones
Fortin opens with a newscast. An anchor cuts to a reporter in a helicopter above a wildfire and asks how big it is.
“John, the forest fire so big, it’s covering well over 140 acres of land. That’s about 200 football fields back-to-back for you and me.”
140 acres is a measurement. 200 football fields is a picture. Without the second half, most of the audience either skips the number or gets it wrong.
The research supports the instinct, though not in the words copywriters use. The concreteness effect is one of the sturdiest findings in verbal learning: concrete nouns like “bicycle” are recalled reliably better than abstract ones like “justice”. Allan Paivio’s dual coding theory, 1971, explains it this way: concrete words get encoded twice, as language and as imagery, so there are two routes back to them.
The mechanism is less settled than the effect. A 2025 paper argues distinctiveness rather than dual coding explains it, and it holds up in people with aphantasia, who report no mental imagery at all.
So make the narrow claim: concrete language is remembered better. “The mind thinks in pictures” is a working metaphor, not a finding. The technique lives in the words, not in asking the reader to visualize.
Why UPWORDS Is Not a Formula Like AIDA or PAS
UPWORDS appears in every large formula roundup and belongs in none of them. Those lists are organized around a question it does not answer.
AIDA, PAS and QUEST are skeletons. They tell you what goes where: attention before interest, problem before solution. UPWORDS tells you nothing about order. There is no first step and no last step, which is why any explanation that gives it stages has invented them.
| Framework | What it decides | When you apply it |
|---|---|---|
| AIDA | The order of the argument | Before you write |
| PAS | The order of the argument | Before you write |
| QUEST | The order of the argument | Before you write |
| UPWORDS | Which words fill the structure | On the edit pass |
Here is one PAS block written twice. The product is Roundwise, a fictional route planning app for small delivery fleets, invented for this article and used for every example in it.
Plain:
Problem. Planning routes by hand takes too long.
Agitate. Mistakes cost you fuel and late deliveries.
Solution. Roundwise automates route planning.
Same structure, UPWORDS applied:
Problem. Every morning starts with a printed list, a highlighter, and forty minutes at the kitchen table.
Agitate. Then a driver calls at ten past ten, stuck on a one-way street, and the whole afternoon slides.
Solution. Roundwise hands every driver the day’s run on their phone, in the order that actually works.
Nothing structural moved. Problem, agitate, solution, three beats in the same order. Only the nouns changed.
The Two Questions That Tell You If a Line Is Working
Two questions, ten seconds, and you can run them on anything you have drafted.
Did an image arrive uninvited?
Read the line once at normal speed. Did you see something?
The tell is the instruction. If the sentence had to tell the reader to visualize, the words were not carrying the image. “Envision a productive office” is an instruction. “A desk at four on a Friday” is a picture. Any line opening with imagine, envision or picture this has handed the reader the job its own words were meant to do.
This is where most published UPWORDS examples fail. They ask you to visualize, then deliver abstractions: havens of creativity, actionable intelligence, opportunities that shape the future. The instruction does all the work.
Is it the same image for everyone in your market?
Everyone quotes the P in UPWORDS and skips the U. The U is the harder half.
An image arriving is not enough. It has to be roughly the same image across your whole market. A picture that lands with half your readers is worse than plain language, because half of them now hold the wrong picture rather than no picture.
Fortin’s rule is to borrow from the reader’s own world. If your market is artists, use art examples. If it is managers, use business analogies. If it is fishing aficionados, use fishing metaphors. The corollary catches B2B writers: jargon is a picture word for insiders and noise for everyone else. His example is the programmer selling to business owners in language only other programmers understand.
Both questions have to pass. One without the other is decoration.
The Three UPWORDS Techniques Fortin Actually Taught
Fortin named three concrete moves under this heading. None survive into current write-ups.
Repetitious Words
Repeat the idea. Never repeat the wording.
“Repetition is the parent of learning,” as Fortin puts it, and the objective “is not to repeat the same words over and over. It’s to use different examples to illustrate your point.” Each pass swaps synonyms and adds information, so the repetition teaches instead of nagging.
His worked set, four sentences carrying one idea:
- “Privacy policies promote purchases”
- “Privacy statements increase sales”
- “Confidentiality is a key to online success”
- “Posting a privacy policy is profitable”
Emotional Words
Near-synonyms point at the same thing and carry different weight. Fortin is not modest about the stakes: “your choice of words is the most important decision you will ever make.”
| Flat word | Weighted word | What changes |
|---|---|---|
| Cost | Investment | Reframes an outflow as an asset |
| Teeth | Smiles | Names the outcome, not the body part |
| Skinny | Slim | Takes the insult out of the compliment |
| Products | Solutions | Points at the problem being solved |
| House | Home | Adds the life lived inside it |
| Beat your competition | Murder your competition | Trades a scoreline for a body |
That last one is his: murder is “so much more powerful, visual and visceral” than beat.
One caution. The swap has to stay true. Calling a cost an investment is fair when there is a return and dishonest when there is not.
Positive Words
Say what a thing is, not what it isn’t. Naming what to avoid still puts it in the reader’s head.
Told not to think of a white carnation, you picture one. Fortin’s demonstration has a rigorous cousin in Daniel Wegner’s 1987 white bear experiment, where people asked not to think of one reported the thought more than once a minute.
In practice:
- “Painless” keeps the reader on pain. Write “there’s little discomfort”.
- “Error-free” software becomes “stable”.
- “Inexpensive” becomes “economical”.
Then there is “but”, the sharpest thing in the source. It reaches backwards and cancels whatever preceded it. A statement followed by “but”, Fortin writes, “can subtly communicate that what was said up to that point was a lie or unimportant.”
Use “and”, and stay on the positive. Not “it’s a great website but expensive”. Rather, “it’s a great website and worth every penny”.
UPWORDS Examples: One Product, Five Rewrites
Same claim, same unglamorous product, five places it has to work. Start with the control, roughly the sentence Roundwise’s site would launch with:
Roundwise optimizes multi-stop delivery routes to improve efficiency and reduce fuel costs.
Nothing in it is wrong. Nothing in it is visible either. Optimizes, efficiency and costs are all abstractions, and the reader reaches the full stop having seen nothing.
| Where it appears | Flat version | UPWORDS version |
|---|---|---|
| Headline | Smarter routing for growing fleets | Your drivers stop rewriting the route at every red light |
| Email opening line | Route planning is a major time sink for dispatchers | At six in the morning your dispatcher is holding a printed list and a highlighter. That is the route. |
| Paid ad | Optimize your delivery operations | Three vans. Forty drops. One highlighter. |
| Landing page subhead | Save time on daily route planning | Plan tomorrow’s runs in the time it takes to unlock the vans |
| Feature bullet | Automated route sequencing and driver dispatch | The day’s stops in the order that actually works, on the phone already in their pocket |
Now the version that looks like the technique and isn’t:
Envision a logistics operation where every route flows seamlessly toward peak efficiency.
Run the two questions on it. Did an image arrive? No. The sentence instructs rather than shows, and flows, seamlessly and peak efficiency are all abstractions. The second question never gets a turn, because there is no picture to test. Imagery-shaped text with no image in it.
One real example to close, because the rewrites above could suggest this means writing metaphors. It doesn’t.
At 60 miles an hour the loudest noise in this new Rolls-Royce comes from the electric clock.
David Ogilvy, 1957. Not one figure of speech in it. A speed, a car, a clock. The picture assembles because the details are specific enough to put you in the passenger seat. A good many of the best copywriting examples work the same way.
The part worth copying is how he got it. Ogilvy spent three weeks on the Silver Cloud and did not write that line. He found it in a sentence by the technical editor of The Motor. Picture words are usually discovered in the subject matter, not generated at the desk.
When UPWORDS Backfires
The failure mode of a technique like this is enthusiasm.
- The plain word is already the picture. “Saves forty minutes a day” needs no help. Reaching for a metaphor makes it worse.
- Mixed or over-extended metaphors. Once a reader starts tracking the image, they have stopped tracking the offer. If an image needs a second sentence to prop it up, cut it.
- Imagery that fights the claim. Vivid language about what could go wrong, aimed at a buyer who is already nervous, sells the fear rather than the fix.
- Regulated categories. Imagery smuggles in implied claims, and implied claims need substantiating exactly as express ones do. The FTC requires a reasonable basis for both before an ad runs.
That last one dates the source. Fortin’s flagship example takes a facial scrub sold on the word “microdermabrasion” and rewrites it as a “facelift in a jar” working “like a sandblaster applied with the gentleness of velvet glove!”
The sandblaster and the velvet glove are excellent, and they are not the problem. The same rewrite promises to “reverse the aging process” and deliver “beautiful skin in hours”, which in cosmetics today are performance claims needing clinical support. Keep the imagery and the promise separate in your own conversion copywriting and you can be as vivid as you like.
Why AI Writes Fake UPWORDS, and How to Use It Anyway
Ask a language model to write vividly and it reliably produces the exact failure this technique exists to prevent. Envision. A clear stream of insight. Havens of creativity. The sentence in the last section was not a straw man. It is what comes back.
The reason is structural. Models learn which words travel together, not what the words point at. A model knows “red” co-occurs with “warm” and “dynamic”. It has never seen anything red. Word-to-word relationships rather than word-to-world, which is precisely the gap UPWORDS was built to close.
Hollis Robbins arrives at the same test from the other direction, checking prose for signs of a machine: if nothing springs to mind, it’s probably AI. Fortin’s 1998 test, pointed at a problem he never saw coming.
A model also cannot pick the universal image for your market, because it does not know your market. Generic vividness is the only kind it has, which is worth being honest about.
Everything after the choice of image is where a model earns its keep:
- You choose the image. That call needs to know who is reading.
- The model drafts around it. Variations, lengths, channels, faster than doing it by hand.
- Hand it the picture in the prompt. “Write this using the printed list and the highlighter” gets you somewhere. “Make this more vivid” returns the fake version every time.
That split is roughly how we think about Instacopy as well. Deciding which image your market actually shares is your job. Drafting around it is what the tool is for.
FAQ
What does UPWORDS stand for?
Universal Picture Words Or Relatable, Descriptive Sentences. The expansion is Michel Fortin’s, from a technique he says he first taught in 1998 and wrote up properly in August 2009. It describes words that put an image in the reader’s head unprompted.
Who created the UPWORDS copywriting formula?
Michel Fortin, a direct response copywriter. UPWORDS is one of five frameworks he teaches, alongside the 3 P’s, OATH, FORCEPS and the 3 Deadly C’s. The technique now circulates widely without his name attached, largely because his original article is no longer on the live web.
Is UPWORDS the same as AIDA or PAS?
No, and the difference is the most useful thing to know about it. AIDA and PAS decide the order of your argument. UPWORDS decides which words fill that order. You can apply UPWORDS to a PAS block without moving a beat. You cannot write “in UPWORDS”, because there is no sequence to follow.
Does the UPWORDS formula work for B2B copy?
Yes, and the “universal” half does more work in B2B than anywhere else. Specialist buyers share a vocabulary that reads as jargon to everyone outside it, so an image that lands with a procurement lead can mean nothing to the engineer signing off. Fortin’s own examples are almost all professional services: cosmetic surgeons, web designers, consultants selling to florists.
Can AI write UPWORDS copy?
It writes the surface reliably and the substance badly. Models produce imagery-shaped sentences with no image in them, because they know which words co-occur rather than what those words refer to. The workable split is that you choose the image your market actually shares, and the model drafts around it. Pair it with a copywriting template if you want the structure decided too.
