The Day You Become a Better Writer by Harry Dry | Growth Summit 24

Harry Dry | Six & Flow

Abstract

A copywriter opens a conference talk with a day he spent photographing his own commute — 107 advertisements collected, one of them remembered — and uses that ratio as the diagnosis. The argument is that there are two separate problems and only one of them is a writing problem: most companies have very little to say, and they are quite bad at saying it. Better words cannot supply a position, so the first half of the talk is about having something to say and the second is about saying it well.

• Mediocre is what gets signed off, because mediocre is what nobody can object to.
• An exact-match search in quotation marks counts how many companies already say your line.
• Positioning narrows to one workable input: the enemy — a direct competitor, a competing approach, or a belief derived from one of them.
• Every line gets three questions: can I visualise it, can I falsify it, and can nobody else say this.
• Concrete means a word you can drop on your foot, and an abstract phrase is converted by interrogation on paper rather than by better word choice.
• A very large advertising budget does not buy something to say — a competitor spending 185 times less has roughly the same customers.

Everything below follows the talk in order, from the commute through the three types of enemy, into the three litmus tests, and out to the closing comparison between a nine-figure advertising budget and a book. For anyone building a personal brand the useful part is that every test in it runs on one line, alone, in seconds, and none of them requires a budget or permission.

Chapter summaries

00:51 - Writing feels hard because the aim is to write well rather than to say something

Writing feels hard because the writer is trying to write well rather than trying to say something.

"If you think writing's hard, it's because you're trying to write well."

— 01:40

The difficulty is self-imposed by the target rather than by the task: aiming at a well-written sentence is a harder problem than aiming at a true one, and only one of the two has an answer. The frame comes from a children's book read at six, in which a character bangs his head against a wall in the hope that it produces a story. The advice offered to him is to keep banging.

01:47 - Mediocre is what gets signed off, because mediocre is what cannot be objected to

Most companies phone it in, most output is mediocre, and mediocre is what gets signed off because mediocre is safe.

"They phone it in. Most stuff's mediocre. Mediocre safe, safe gets signed off."

— 01:51

The approval step selects for what cannot be objected to rather than for what works, and those are different filters. A line nobody can argue with is also a line nobody can remember, so the safety that gets a piece signed off is the same property that makes it invisible once it runs. It is offered as the reason the rest of the day's collection looks the way it does.

01:55 - A category's copy is demonstrated by collecting it rather than argued about

The state of the category is demonstrated by collecting it rather than argued about; one ordinary commute is a sufficient sample.

"I want to show you, start off today by showing you the problem."

— 01:55

Side by side, each individual advertisement loses the benefit of the doubt it gets in isolation — which is why the demonstration is a collection rather than an argument. One dated day, an early start, a single underground line.

"Here's the first four ads I saw. Joins told me to live life to the max. XTB had more small print than print."

— 02:13

A supplier relationship shown as flat cardboard cutouts; a chocolate bar whose whole promise is the phrase and more. None of it is cherry-picked, and that is the claim: an ordinary commute is a sufficient sample.

03:12 - Large brand spend does not produce a distinguishable brand, so the message collapses to price

Large brand-strategy spend does not produce a distinguishable brand; the message collapses to price.

"I could not tell the difference between the two. It's all just save up to some money."

— 03:14

Where nothing separates two offers in language, the only remaining variable is the number — so the money spent on brand strategy ends up buying a discount war. Two mobile networks reduce to a small monthly saving; a payments billboard turns out to lead to a website that is more of the same, from a company he puts at eighteen billion pounds.

05:08 - One remembered advertisement is the measurable output of a whole day's category spend

The measurable output of a whole day of category-wide advertising spend is one remembered ad.

"I took 107 pictures in the day, and I remembered one."

— 05:09

Memorability rather than exposure is the scarce product, and volume does not convert into it. 107 photographs taken across one day, one of them remembered — a ratio he frames explicitly as reality rather than hyperbole. Everything upstream of that number, the budgets and the placements and the sign-offs, produced a single recalled line between them.

05:17 - Bad copy is not badly chosen words, it is words nobody chose

Bad writing is not badly chosen words, it is words not chosen at all — prefabricated strips already set in order by someone else.

"It consists in gumming together long strips of words which have already been set in order by somebody else."

— 05:36

Nobody is doing the selection step, so no line carries a decision. The passage is from 1946 and he says it summarises the problem better than he can, which is itself the point: the diagnosis is old, and the repetition of the failure is what is new. What the passage names is not bad taste but an absent process.

05:48 - An exact-match search counts how many companies already say the same line

You can check whether a line is yours in seconds: put it in a search engine inside quotation marks for an exact match.

"You add quotation mark to get a direct match. You'll see that hundreds, genuinely hundreds of other companies are promising to take your payments to the next level."

— 05:56

An exact-match count is an objective test of ownership that costs nothing to run and needs no judgement. The payments slogan comes back on what he says is genuinely hundreds of other companies promising the same thing, and his conclusion is that the words have lost all meaning. The test is a check on whether the line is yours at all, before any question of whether it is good.

06:14 - Ready-made phrases survive because nobody asks what they mean

The second test is to ask what the words actually mean, and the writer usually cannot answer.

"What do they mean by essential skills, and what do they mean by excel? Do they mean a £24,000 a year starting salary?"

— 06:24

A ready-made phrase is tacked together rather than meant, so interrogation finds nothing underneath it — which is what separates this test from the search one. He flags that he is asking genuinely rather than sarcastically, and the questions are specific enough to be answerable if anybody had an answer. The agency version of the same failure follows.

"It's abstract word bingo. You can play this with your own website."

— 07:05

Copy that reduces to telling stories, being award-winning and delivering results is what happens when there is no stated philosophy for the copy to be built from.

07:20 - There are two problems, not one: very little to say, and bad at saying it

There are two problems and they are separate: most companies have very little to say, and they are quite bad at saying it.

"There's two big problems. Most companies have very little to say, and they're quite bad at saying it."

— 07:22

Treating them as one problem sends the fix to the wrong layer, because better words cannot supply a position. It is the structure of the rest of the talk — having something to say, then saying it well — and the order matters: the second half is worthless applied to the first half's absence. He hedges it as what he will maybe perhaps manage in the time.

07:53 - The line an audience remembers is the residue of questions asked before any writing

The one line an audience remembers is the visible residue of positioning work the audience never sees.

"What are we competing against here? How are we different? Why us? What are we not?"

— 08:37

The output looks like a turn of phrase; the input was a set of questions answered before any writing began. A dating app's line on a bus is all anyone sees, and the questions behind it — what are we competing against, how are we different, what are we not, who is our enemy — are the part that never appears. The line is the residue, and the residue is what gets copied by people who skip the questions.

08:46 - Positioning narrows to one workable input, which is the enemy

Pick an enemy and position against them hard.

"I'm going to focus on one very specific thing: your enemy."

— 09:02

Positioning and storytelling are too large to cover, so a single input gets worked in the time available rather than the whole discipline gestured at. He interrupts himself to say the framing makes him sound like a guru and that he is definitely not one, and then gives the instruction the rest of the half unpacks: pick an enemy, position against them hard. An enemy supplies the contrast the message is written against, so the words have something to push on instead of standing alone.

09:20 - Three kinds of enemy get lettered, because the plain list will not hold

There are three types of enemy — competing approaches, beliefs you stand against, and direct competitors — lettered A, B and C.

"There's three types of enemy: direct competitors, competing approaches, and stuff you stand against."

— 09:22

He says the unordered list is hard to remember, so the letters are the retrieval device and the reordering is done live on stage: A for approaches, B for beliefs, C for competitors. It is a small thing, but it is the same move as everything else in the talk — the version that can be held is the version that gets used.

09:48 - One question produces an enemy: what the thing is actually competing against

The question that produces an enemy is: what are we actually competing against?.

"What are we actually competing against? You want to get into the habit of asking that question a lot."

— 09:56

It refuses the assumed category and makes the comparison set an explicit choice rather than an inherited one. He takes it from a book on positioning and calls it, hedged twice, probably the most important question in all of positioning and storytelling, and recommends asking it habitually rather than once.

10:03 - With a direct competitor named, the line comes from the contrast rather than the product

With a direct competitor named, the line writes itself from the contrast rather than from the product.

"Let's start with competitors. Imagine I've set up a privacy-focused email service."

— 10:03

The enemy supplies both the subject and the angle, so the copy becomes a comparison rather than a claim.

"I might write, Google read your email more than you."

— 10:17

The line is about the rival's behaviour, not about the product's features, and it is legible in one read because the reader already knows the comparison. The same exercise is run on the floor with a local bookshop, and the room supplies the enemy before he does.

11:06 - The enemy need not be a named company, because the status quo works as well

The enemy does not have to be a named company — the status quo works as well.

"Just the status quo. They don't have to be named."

— 11:13

A named rival is one instance of a general condition, and positioning against the condition avoids naming anyone while widening the claim. A new PR company positioned itself against traditional PR being dead and then wrote an essay on the subject — a post he says drew twenty million likes, and which its author told him was entirely organic. The status quo is available to anyone, including someone whose category has no dominant rival to push against.

12:53 - A competing approach is a different way of solving the same problem

A competing approach is a different way of solving the same problem, and it is often a better enemy than another product in your own category.

"Competed approaches. This is a different way of solving the same problem."

— 12:58

The competing approach is what customers actually use today, so it is the real alternative; rivals inside your own category are a much smaller pool. A biodegradable dishcloth positioned against paper towels rather than against other dishcloths, with a count of trees cut down daily to make them. The exercise handed to the room is transport advertising, where the competing approaches are taxis, walking and cycling rather than other trains.

14:18 - Beliefs are derived from a competitor or an approach, never invented

Beliefs are the third and hardest type of enemy, and they are derived rather than invented — they come from a direct competitor or a competing approach.

"You don't pull them out your backside. Traditionally they come from a direct competitor or a competing approach."

— 14:27

A belief pulled from nowhere has no constituency; one derived from an existing alternative already has people on both sides of it. A coding school whose competing approach is the computer-science degree produces the belief that you do not need a degree to be a good programmer — which is arguable, which is the point. He names it as the third and hardest type and zips through it.

14:56 - Every advertisement worth liking already has a conception of what it is up against

Every ad in the set he liked had some conception of what it was up against; the ones at the start of the talk did not.

"And they all had some conception of what are you up against. It's really important."

— 15:04

Knowing the alternative is what gives a line something to be true against; without it the copy is unanchored and can only assert. The two collections are set against each other — the commute's advertisements versus the examples worked in this half — and the difference is not craft. It is whether anybody decided what the thing was up against before the writing started.

15:10 - An admired line can be read backwards to the two inputs that produced it

A line you admire can be reverse-engineered back to the direct competitor and the competing approach that produced it.

"Look at the art I liked. No Swedish meatballs, no instructions, no divorce."

— 15:10

Both inputs leave fingerprints in the finished words, so the words can be read backwards to recover them. The vintage-furniture line about no meatballs, no instructions and no divorce hands over its direct competitor in the first clause.

"Ikea would be the competitor, for the Swedish meatballs. Now the competing approach."

— 15:38

The competing approach is in the second and third: building the furniture yourself, and what that costs. Nothing in the line is decoration — every element is one of the two inputs, which is why it can be reverse-engineered at all.

16:06 - Three questions get asked of every individual line

Every individual line gets three tests: can I visualise it, can I falsify it, can nobody else say this.

"I asked myself: can I visualise it, can I falsify it, and can nobody else say this?"

— 16:23

Three independent failure modes — unpicturable, unprovable, unowned — and a line can pass one while failing the others, which is why they are run separately.

"And if you get three yeses you probably got something good. You get three nos, you probably wrote a lot of rubbish."

— 16:31

He describes it as what he does to his own writing, line by line, ever so pretentiously. A line can also sound like ambition and fail all three at once — don't just get a job, change an entire industry. Nothing to picture; not binary, because whether a given figure changed an industry is subjective; and writable today by any competitor. Abstraction produces the three failures together, because there is no object, no claim and no specific.

17:11 - A passing line is one a competitor could not copy without lying

The passing line is specific enough that the competitor's version of it would be false.

"Worn by supermodels in London and dads in Ohio. Can I visualise this?"

— 17:13

Naming two opposite users of the same product is a checkable claim and one a rival cannot copy without lying. Both pictures form immediately: the dad at the barbecue and the model on the catwalk, in the same shoes. The equivalent line does not hold for the luxury brands, which is what makes it ownable rather than merely vivid — the claim is doing work no competitor can borrow.

17:59 - What cannot be pictured will not be remembered

What cannot be pictured will not be remembered.

"If I can't see it, it's not there yet. If you can't visualise it, you won't be able to remember it."

— 18:04

Memory attaches to images, so an unpicturable phrase has nothing to attach to. The demonstration is a live test.

"Six words are going to come up on the screen. Leg of lamb, a better way, muscly Irishman, available knowledge, seamless transition."

— 18:20

Three of the six are concrete and three are abstract, and the three he expects the room to recall are the three that can be seen. He hedges it hard — maybe he is wrong, maybe somebody has a photographic memory — which leaves the test as a test rather than a law.

19:05 - A concrete word is one that can be dropped on a foot

Concrete means real and tangible — a word you can drop on your foot.

"Concrete means real and tangible. It's a word that you can drop on your foot."

— 19:12

The physical test is binary and needs no judgement, which is what makes it usable during revision rather than only in theory. A leg of lamb gets dropped on stage; an attempt to drop a better way produces nothing, because there was nothing there to fall. Every other test in this half of the talk needs a decision from the writer, and this one does not.

19:34 - An abstract phrase is converted by interrogation on paper, not by better word choice

An abstract phrase is converted into a concrete one by repeated interrogation on paper, not by better word choice.

"I changed it to just zoom in. What I do is I get a sheet of paper."

— 19:47

Asking what you actually mean, repeatedly, forces each rewrite a rung lower until a physical object is left. The sheet has abstract at the top and concrete at the bottom, and the phrase is written at the top and rewritten down the page.

"What do I actually mean, over and over and over again. And I'll rewrite it over and over again until I'm left with a concrete object."

— 20:06

An app about regaining fitness becomes someone who has been on the couch for six months; fitness becomes running; running becomes five kilometres. The technique was renamed because the first name was not sticking, which is the same rule applied to itself.

20:49 - A concrete name travels without a budget, because the picture is already wanted

A concrete name travels without a budget, because it can be pictured and the picture is something people already want.

"You get couch to 5K, which is the world's largest fitness programme."

— 20:55

Visualisability substitutes for distribution spend, because the phrase carries its own promise and the picture is something people already want. It was not made by a sportswear giant with a budget; it came from a television assistant with no connections. The name spread because anyone hearing it can see both ends of it at once, which is the cheapest distribution there is.

21:19 - Only a falsifiable claim gets a reaction, because only it can be wrong

Falsifiable means able to be proved true or false, and only falsifiable claims provoke a reaction.

"Next up, can I falsify this. This just means able to be proved true or false."

— 21:19

A claim nobody could disprove costs the speaker nothing, so it is worth nothing to the listener. An astronomer was sentenced in 1633 to ten years' home imprisonment for a claim about what revolves around what — and the vague version, a harmonious connection with a radiant celestial body, would have cost him nothing at all. The sentence is the evidence that the claim was falsifiable.

22:03 - A falsifiable line puts the writer's head on the block, and risk is the signal

A falsifiable line makes an audience sit up, because the writer has put their head on the chopping block.

"Something falsifiable: our ears prick up, we sit up a little bit in our seat. It's like someone's putting their head on the block."

— 22:06

Risk is the signal: skin in the game is what distinguishes a claim from a flourish, and an audience can feel the difference without being able to name it.

"One's juicy, tasty and high quality, and the other one is organic, Irish and free range. Which one are you going to pick up?"

— 22:26

Which one gets picked up is called a rhetorical question, and it is — the same product described two ways is two different products. The easiest route to the same effect is mechanical.

"The easiest way for today is just to replace adjectives with numbers, with data."

— 22:41

Low unsubscribes becomes a stated percentage, loads of protein becomes a stated weight, save more becomes a stated figure. A number can be checked and an adjective cannot, so the number carries the risk the adjective avoids.

23:03 - A line a competitor could sign is doing no work for whoever paid for it

Never write an ad a competitor could sign.

"Never write an ad a competitor can sign. Delicious drinks, memorable moments."

— 23:17

A line any rival could put their name to is doing no work for the one that paid for it. He places an invented line about delicious drinks and memorable moments under a wine brand and then under a cola brand, and it fits both — which is the whole demonstration. He says he made them up, and that nobody knows or cares who said it first, because all of them are rubbish.

23:44 - Pointing requires an object, which forces the writing off adjectives

Imagine you cannot talk and can only point — cover your mouth and point at things.

"Imagine you can't talk, you can only point. So you cover your mouth like this and you just point at stuff."

— 24:29

Pointing requires an object, which forces the writing off adjectives and onto specifics — you cannot point at a better way. It came from a forum post by an account since deleted, buried in a long spiel, and it stuck with him. It is introduced as the contrast to a car brand's own guidance on writing its advertisements, of which he reads the first two: look at the car, and then look harder. Six points, of which he reads two, and both of them remove the writer's invention and put the attention on the object.

24:42 - Pointing on the page looks like counts rather than claims

Pointing on the page looks like counts, not claims.

"There are 3,389 men at our Wolfsburg factory with only one job: to inspect each stage of production."

— 24:50

The reader draws the conclusion from the figures, which is more persuasive than being handed it.

"More inspectors than cars. They're just pointing. They're not saying, oh, we're really safe."

— 25:01

A count of inspectors, a count of cars produced daily, and the arithmetic left to the reader. The claim that the cars are safe is never made, which is precisely why it lands — the reader arrives at it themselves and therefore owns it.

25:13 - The same company written both ways shows what talking leaves out

The same company written both ways shows the difference: talking has nothing substantial in it, pointing does.

"That's just talk. I'm just talking. There's nothing substantial there."

— 25:31

Talking asserts a future state; pointing cites what already happened, which can be checked. He invents a coding school on the train that morning — creating the coders of the future, one hundred per cent online, interactive technology, careers launched in up to nine months — and sets it against a real one: a count of graduates since a stated year, a stated job-placement rate, and one named path from stay-at-home parent to software engineer. Same category, same claim, two different amounts of evidence.

26:43 - A very large advertising budget does not buy something to say

A very large advertising budget does not buy something to say.

"Spent 345 million pounds on marketing in 2023, and what did they do with all that money?"

— 26:46

Money buys placement, not a position, and with nothing to say the extra spend buys more interchangeable placements. The figure, he notes, exceeds one small nation's GDP, and it bought billboards about moving work forward for any team, any project, any goal. Its named competitors, he says, think differentiation is using the words drive, impact and scale in different combinations. If everyone draws from the same small word pool, recombination cannot produce difference — which is the same failure as the commute's advertisements, arrived at with a far larger budget.

27:45 - A fraction of the spend coexists with profit and a comparable customer base

In the same category, a fraction of the advertising spend coexists with profit and a comparable customer base.

"They spend 185 times less than that. They don't lose hundreds of millions of pounds every year, they make tens of millions."

— 28:04

If the spend were producing the customers, the comparison would not hold. A project-management company in the same category spends by his figure 185 times less on advertising, makes tens of millions rather than losing hundreds of millions, and has roughly the same number of customers. The number that differs most is the one that matters least.

28:18 - An owned asset people seek out replaces the placement a budget buys

The company did something else instead of advertising: it wrote a book millions of people wanted to read, carrying its own ideas about how the work should be done.

"They wrote a book which literally millions of people wanted to read, which has their ideas about how to do project management software."

— 28:21

An owned asset people seek out replaces bought placement, and it carries the position rather than a slogan. A book people want to read does what three hundred and forty-five million pounds of billboards did not, because it contains something to say rather than a way of saying nothing. The closing ask is to work out what you are actually talking about first, and then write about it at length. The close is delivered against the clock and hedged throughout — maybe, and you might do a little bit better. Having something to say is upstream of every technique in the talk, which is why it is the last thing said and the first thing needed.

Personal Branding Lessons

The talk is aimed at companies with billboards, but every test in it is designed to be run by one person on one line, in seconds, with nothing but a search box and a sheet of paper.

Collect your own category's advertising for one day

Photograph every piece of copy your category puts in front of you on an ordinary day — no curation, no cherry-picking. Side by side, each line loses the benefit of the doubt it gets in isolation, and the pattern becomes impossible to argue with. One commute produced 107 images and a single remembered line. Do the same in your own field and you will have both the diagnosis and the evidence for it, which is the only argument that survives contact with somebody who signs things off. 01:55

Search your line in quotation marks for an exact match

Put the sentence inside quote marks in a search engine, which forces a direct match, and count how many companies wrote it before you. A payments company promising to take your payments to the next level comes back on hundreds of sites. The count is the entire verdict — no taste, no second opinion, no judgement. Anything above a handful means the words have already been spent, and whatever they once meant was spent by whoever got there first. 05:48

Ask what the words actually mean, and keep asking

Take the phrase and interrogate it, genuinely rather than sarcastically. What does essential skills mean — which skills? What does excel mean: a particular starting salary, promotion within six months? A ready-made phrase is tacked together rather than meant, so the questioning finds nothing underneath it, and that absence is the finding. The same sheet of paper is the conversion tool: write the phrase at the top, ask what you actually mean, rewrite it a rung lower, and repeat until a physical object is left. 06:14

Name the enemy before writing a single line

Positioning is too large to work all at once, so take the one input that pays immediately: what is this actually competing against? An enemy supplies the contrast the words are written against, so the copy becomes a comparison rather than a claim, and the line writes itself from the gap. There are three kinds, lettered because the plain list will not hold — A for competing approaches, B for beliefs you stand against, C for direct competitors. 08:46

Choose the competing approach over the direct competitor

A competing approach is a different way of solving the same problem, and it is usually the better enemy, because it is what people actually use today. Rivals inside your own category are a much smaller pool and a much smaller argument. A biodegradable dishcloth positioned against paper towels rather than against other dishcloths has the whole market to talk to instead of a slice of it — and the claim it can make is about trees rather than about absorbency. 12:53

Run the three litmus tests on every line you write

Can I visualise it. Can I falsify it. Can nobody else say this. Three independent failure modes — unpicturable, unprovable, unowned — and a line can clear one while failing the others, which is why they are asked separately. Three yeses is probably something good; three nos is probably a lot of rubbish. Change an entire industry fails all three at once. Worn by supermodels in London and dads in Ohio passes all three, and a competitor copying it would be lying. 16:06

Zoom in until a concrete object is left on the page

Draw a line down a sheet, write abstract at the top and concrete at the bottom, put the phrase at the top and ask what you actually mean — then rewrite it one rung lower, and again, until what remains is something you could drop on your foot. An app about regaining fitness becomes six months on the couch, then running, then five kilometres, then couch to 5K. That name was made by a television assistant with no budget, and it travelled because anybody hearing it can see both ends of it. 19:34

Replace the adjectives with numbers

Falsifiable means able to be proved true or false, and only a falsifiable claim gets a reaction, because only it costs the writer anything. The cheapest route there is mechanical: low unsubscribes becomes a stated percentage, loads of protein becomes a stated weight, save more becomes a stated figure. A number can be checked and an adjective cannot, so the number carries exactly the risk the adjective was avoiding — and the risk is what makes a reader sit up. 22:03

Cover your mouth and point at the thing

Imagine you cannot talk and can only point. Pointing requires an object, which forces the writing off adjectives and onto specifics, because there is nothing to point at in a better way. On the page it looks like counts rather than claims: a count of inspectors at one factory, a count of cars produced daily, and the conclusion left to the reader rather than asserted. The line never says the cars are safe, which is exactly why the reader believes it. 23:44

Questions

Each answer ends at the moment in the recording where it is given.

Why does my copy sound exactly like everybody else's?

Because the words were never chosen. The diagnosis borrowed here, from something written in 1946, is that bad writing does not consist of picking words badly — it consists of gumming together long strips of words already set in order by somebody else. A prefabricated phrase carries no decision inside it, so the sentence can be grammatical, professional and completely unowned. That is a different defect from clumsy wording, and improving the vocabulary of a ready-made line leaves it ready-made. 05:17

How do I check whether my tagline is actually original?

Put it into a search engine inside quotation marks so only exact matches return, and read the number. Run on one payments slogan, the test returned what he describes as genuinely hundreds of other companies promising the same thing. The count measures ownership rather than quality: a line can be clear and pleasant and still belong to several hundred organisations. It costs nothing, needs no agency and settles the question before any money is spent placing the words. A niche validation exercise starts in the same place. 05:48

What is a brand enemy, and how do I pick one?

An enemy is whatever your message is written against, and it supplies the contrast the words push on. Three types are given, lettered so the list will hold: A for competing approaches, B for beliefs you stand against, C for direct competitors. The instruction is to pick one and position against it hard, which is where a map of your landscape becomes a writing tool. A message written against nothing has to be true in the abstract, which is precisely the condition that produces interchangeable copy. 09:14

What am I competing against if I do not have an obvious direct competitor?

Look for the competing approach — a different way of solving the same problem, which is usually what people already do instead of buying anything. A biodegradable dishcloth competes with paper towels rather than with other dishcloths. The enemy also does not have to be a named company: the status quo works, and positioning against a condition rather than a rival widens the claim to everyone inside it. Mining what your audience already struggles with is how the real alternative surfaces. 09:48

Why do people remember some phrases and forget everything else I write?

Because memory attaches to images, and a phrase with no object in it has nothing to attach to. The rule offered is that if you cannot visualise it, you will not be able to remember it, demonstrated live with six phrases on a screen for ten seconds. The working definition of concrete is physical and needs no judgement to apply: a word you could drop on your foot. The prediction about which phrases survive is hedged, but the mechanism is not in dispute. 17:59

How do I make my writing concrete instead of abstract?

On paper, by interrogation rather than by better word choice. Draw a line down a sheet, write abstract at the top and concrete at the bottom, put the phrase at the top and ask what you actually mean — then rewrite it a rung lower, repeatedly, until an object appears. Regaining fitness becomes six months on the couch, then running, then five kilometres, and shortening the result gives a name someone can picture themselves inside. A content brief is the natural place to keep that ladder. 19:34

How do I make a claim that sounds believable instead of like marketing?

Make it falsifiable — able to be proved true or false. The illustration is an astronomer sentenced in 1633 to ten years of home imprisonment; a vaguer version of the same claim would have cost him nothing, because there would have been nothing in it to prosecute. A claim that could be disproved reads as risk taken, which is why it makes an audience sit up. Two chicken breasts, one labelled juicy and high quality and the other organic and free range, make the same point. 21:19

What should I use instead of adjectives?

Numbers. Offered as the easiest of several routes rather than the only one, and sourced, hedged, to a large retailer's writing guide: low unsubscribes becomes a stated percentage, loads of protein becomes a stated weight. A number can be checked and an adjective cannot, so the number carries exactly the risk the adjective was avoiding. The version that feels uncomfortable to publish is the one doing the work, because the discomfort is your own exposure becoming visible. 22:37

How do I tell whether my positioning line is actually differentiated?

Ask whether a competitor could sign it. The rule quoted is never to write an advertisement a competitor could put their name to, demonstrated with an invented line about delicious drinks and memorable moments placed under two different brands in turn. Where the exact-match search counts who already says your line, this test asks who could say it truthfully tomorrow. A sentence that would sit comfortably under a rival's logo is promoting a category and attributing it to nobody. Trust forms around claims only one person can make. 23:03

Do I need an advertising budget to get attention?

The closing comparison says no. One work-management company spent £345 million on marketing in 2023 and bought billboards any rival could have signed. A competitor selling the same software spends 185 times less, makes tens of millions rather than losing hundreds of millions, and holds roughly the same number of customers, give or take. What it did instead was write a book millions wanted to read. The ask, delivered under the clock, is hedged: maybe the time is better spent working out what you are actually talking about. 27:45

Sources

The Day You Become a Better Writer by Harry Dry | Growth Summit 24

The sections above follow the episode's own order. Each timestamp is the point in the recording where the idea is discussed, so the post reads as a map of the conversation rather than a rearrangement of it.

Quotations come from the episode transcript with spoken filler ("okay", "right", "you know", "um", "actually", "kind of", "sort of") and false starts removed. No words are added: square brackets mark anything inserted for sense, and an ellipsis marks any cut. British spellings are restored, and speech-to-text errors were repaired against context only where the intended word is unambiguous. Where a line appears both in a cold-open teaser and in its place in the conversation, the timestamp cites the second — the point at which it is actually said.

Quotes are stamped with the moment they are said rather than with a speaker's name.