Tom Goodwin and Rory Sutherland Fix Marketing (Part 2)

creator | channel

Abstract

The second half of a three-way conversation, opening on the agency of the future and a prediction that the whole model runs backwards: when production becomes cheap and fast, the work gets made first and sold afterwards, the way pottery and paintings always were.

• Marketing is a philosophy of business rather than a department, and naming is its strongest instrument.
• Creativity has a mathematical definition — it is what gets you out of a local maximum.
• Moving the perception costs less than moving the physics, because the speed problem is usually an uncertainty problem.
• A badge changes the product: identical cars diverge on faults, satisfaction and resale value.
• There are no trends, only vectors, and every movement creates its own counter-movement.
• Everything starts as an option and ends as an imposition, which is worth having at the margin and self-defeating at saturation.

For one person building a brand, most of what the two describe as organisational obstruction has no equivalent. A company keeps a department that can veto an idea, a measurement system that discounts what it cannot count, and a client relationship too valuable to risk; a solo operator holds all of those roles at once, so the reframe, the price and the publishing decision belong to the same head. Corporate theory here is, for one person, simply the next thing done that afternoon.

Chapter summaries

00:10 - When production becomes cheap and fast, the work gets made first and sold afterwards

When production becomes cheap and fast, the work gets made first and sold afterwards.

"the whole thing works backwards, which is that you have production companies that are very, very good at high-speed"

— 00:10

Commissioning made sense when materials and time were expensive; once they are not, the maker can produce and then find a buyer. Jewelled eggs were commissioned one at a time and pottery was made in quantity and then sold — and nobody ever commissioned a song about a submarine. The supplier becomes the competitor, having no relationship to defend.

03:05 - Marketing is a philosophy of business rather than a department, and naming is its strongest instrument

Marketing is a philosophy of business rather than a department, and naming is its strongest instrument.

"I think the creative industry should reposition themselves across the whole of business, the whole mix"

— 03:05

Value creation happens in the mind rather than in the factory, and what a thing is called decides whether people will do it. A drinking occasion renamed into something respectable; a railway line renamed into part of a transit network. The most important marketing here was not done by marketers at all.

08:28 - People in real business want a better solution rather than a creative one

People in real business want a better solution rather than a creative one.

"we have the people to do this, and we have the process to do this, but the one thing that's letting us down is we're obsessed with the"

— 08:28

Almost every role in a company concerns what has already happened or how the company buys things, which leaves exactly one function looking outward and forward. Very few jobs anywhere are about the future, let alone the far future — and that vacancy is the opportunity rather than a complaint.

09:46 - Creativity has a mathematical definition: the escape from a local maximum

Creativity has a mathematical definition: it is what gets you out of a local maximum.

"there is a kind of mathematical explanation for what it is, which is it's the kind of thinking"

— 09:46

Everyone has defined the problem the same way and is trapped by the definition, so the lateral move is the only route to higher ground. It is the escape from a local maximum rather than a temperament — and it is the process rather than the result that earns the name.

11:31 - Moving the perception costs less than moving the physics

Moving the perception costs less than moving the physics.

"imagine going to a bank and saying, 'Let's come up with a way to get people to pay more for their bank account.'"

— 11:31

An apparent speed problem is often an uncertainty problem, and expectations can be reset for nothing where the process cannot be changed at all. An insurer about to spend a fortune accelerating claims solved it by promising four weeks and paying in two, and the complaints stopped.

14:52 - Government has no marketing function, so it imposes rules where it could trade them

Government has no marketing function, so it imposes rules where it could trade them.

"Government has no marketing function. Spends 45% of GDP. Doesn't have a creative function."

— 14:52

Without a way to speak to citizens directly, the only channel available is journalists trying to catch you out, which leaves compulsion as the only instrument. A speed restriction imposed is resented; the same restriction offered in exchange for removing something people dislike is accepted.

17:22 - Logical rigour only reaches the answers competitors have already found

Logical rigour only reaches the answers competitors have already found.

"we've fetishized Where I'm sympathetic with the word creative, although I define it slightly differently"

— 17:53

Requiring that everything make sense in advance limits the explorable solution space to a fraction of itself — and that fraction is the part every competitor has already searched with the same logic. None of the important discoveries happened forwards.

19:38 - A process with no creativity in it can still reach answers no human would

A process with no creativity in it can still reach answers no human would.

"in detective work, a an AI plus a human, uh, each of them will have specific strengths"

— 19:38

Trying an unparalleled number of combinations while dismissing assumptions about what the answer should look like is not a creative process and produces creative outcomes. An antenna designed by evolutionary search is asymmetrical where any person would have made it symmetrical.

21:42 - A business is a discovery mechanism rather than an efficient supplier of wants people already have

A business is a discovery mechanism rather than an efficient supplier of wants people already have.

"There is a thing something which needs to change before marketing marketers in general"

— 21:42

One tradition treats preferences as already known and competition as pressure on price; the other treats the firm as exploring adjacencies where new value can be created. Discovery rather than supply — and only one of the two is taught nearly everywhere.

23:18 - Advertising is now two industries that mistakenly believe they are one

Advertising is now two industries that mistakenly believe they are one.

"I think there's this fascinating divide in advertising now and there's almost two completely different industries that think they're the same industry"

— 23:18

One half of the industry admits it cannot fully explain why anyone buys anything; the other believes buying is rational and the job is interception. The second does produce immediate measurable return, which is a different thing from producing the sale.

25:18 - Advertising earns more by raising willingness to pay than by intercepting people already buying

Advertising earns more by raising willingness to pay than by intercepting people already buying.

"there's some really weird maths out there that shows if you use advertising to get people to double the number of people to buy your product, you probably"

— 25:18

Margins in most categories are thin enough that doubling the number of buyers returns less than persuading the same people to pay thirty per cent more. The job is to make more people aware, admiring and willing — some of whom buy now and some in years.

26:40 - A badge changes the product, so the trusted one resells higher and complains less

A badge changes the product: the trusted one resells higher, and the dearest watch keeps worst time.

"The same car. Because it was basically down to the perception of the brand."

— 27:04

Expectation colours the experience, so identical goods produce different fault reports, different satisfaction and different resale values. Two cars off one production line, badged differently, diverged on every measure that supposedly described the car. The badge is part of the product.

29:10 - A company's value depends on the altitude of the category it is judged from

A company's value depends on the altitude of the category it is judged from.

"it's absolutely ludicrous that it's worth more than every single other car company in the world put together"

— 29:24

Software makes companies horizontal rather than vertical, so what one is worth depends entirely on which category you place it in and how far regulation lets it expand. A valuation that is absurd at one altitude becomes arguable at another.

31:27 - There are no trends, only vectors, and every movement creates its own counter-movement

There are no trends, only vectors, and every movement creates its own counter-movement.

"My wife doesn't like it, not because of the technology or because she feels unsafe, but she prefers talking to the driver."

— 31:33

As one direction saturates, the opposite becomes noticeable and therefore valuable — which is the surprise mechanism again. Retail automation produced a return to markets; airports converged until the one without shops became the prize. There are no trends, only vectors.

33:01 - The people designing the default future generalise from lives almost nobody else lives

The people designing the default future generalise from lives almost nobody else lives.

"One of the strange findings is people, if you measure happiness at a kind of real-time level, people are actually really, really happy while"

— 33:01

A small and unrepresentative group decides what everyone else's defaults will be, generalising from lives almost nobody lives. Findings about how people feel while driving came from a place with unusually bad traffic — and many of those achievements have been miraculous, which is stated plainly.

35:45 - Self-service is a gain beside the human option and a loss replacing it

Self-service is a gain beside the human option and a loss replacing it.

"The classic example of that is is what I call the railway station ticket machine. Brilliant option."

— 35:45

Removing the alternative to the lowest-cost channel destroys exchanges that built trust and produced knowledge nobody counted. A machine cannot sell the cheaper split fare that a member of staff would have found. What efficiency kills is not on the ledger.

39:16 - New technology changes little until adoption tips and the world rearranges itself around it

New technology changes little until adoption tips and the world rearranges itself around it.

"there's something really interesting about both EVs and self-driving cars. Cuz in a way, they're two separate technologies"

— 39:16

Below the threshold the old infrastructure must still exist, so almost nothing changes; above it, roads, insurance, property and cities all reorganise at once. The technology is the same in both states and the world around it is not.

42:08 - Everything starts as an option and ends as an imposition

Everything starts as an option and ends as an imposition.

"what have you lost at this point because should you optimize for optionality or optimize for optimality?"

— 42:08

An option is welcomed, adopted widely, and then made compulsory by the quiet withdrawal of the alternative. The thing itself is often not undesirable — the question worth asking each time is what it kills.

44:13 - Such innovations are worth having at the margin and self-defeating at saturation

Such innovations are worth having at the margin and self-defeating at saturation.

"One of the things marketers are trying to do is basically maximize possible buying situations. And that means you go for variety."

— 44:23

A few instances add something and the same thing everywhere removes the conditions that made it valuable. Short-term letting helps a rural owner and empties a city of the residents who made it worth visiting, so the visitor no longer experiences the place they came for.

46:56 - Most celebrated new marketing is old mechanics rediscovered from a still-valid playbook

Most celebrated new marketing is old mechanics rediscovered, and the old playbook is still the starting point.

"if you were a CMO of a brand today, given everything we've talked about, what would be the one"

— 46:56

The principles and the mechanics have not changed; the tools and the context have, which means you get both rather than choosing. A celebrated recent idea had been running in restaurants for decades, and an eighteenth-century arrangement was already hardware as a service.

50:23 - Adoption is a sigmoid, and business keeps doing linear arithmetic on it

Adoption is a sigmoid, and business keeps doing linear arithmetic on it.

"I'd use your intuition, but I'd also marry it to complexity maths. Because my argument is that most of business is using the wrong maths."

— 50:23

Habit and social copying drive adoption, so early linear extrapolation kills products that would have worked. A wearable abandoned early returned a decade later in an acceptable form — and being first asks someone to break two instincts at once. Adoption curves, it does not climb.

52:22 - Social proximity beats reach, because people adopt what the people near them have adopted

Social proximity beats reach, because people adopt what the people near them have adopted.

"rather sell five solar panels to five people who live in the same street than 15 solar panels to people who live 10 miles apart"

— 52:22

Five neighbours adopting is worth more than fifteen strangers, because the mechanism is contagion rather than exposure. Asked why he bought the thing, he would have produced plausible reasons — and then realised he would never have bought it had a family member not bought one first.

53:50 - The forces a model leaves out are set to zero rather than to unknown

The forces a model leaves out are set to zero rather than to unknown.

"if you imagine that shot around the moon that happened, you know, just a few weeks ago"

— 53:50

Obsessing over the handful of forces you understand while setting the rest to zero rather than to unknown produces confident predictions that miss entirely. Attempting a spaceflight without accounting for gravity is the analogy, and brand building compounds while being judged as if it were linear.

55:16 - Technology marketing abandons the old, and category belief beats category performance

Technology marketing abandons the old, and category belief beats category performance.

"It records It records what you say. I mean, it plays music through your ears here."

— 55:23

Every technology is presented as ideal for the young, which leaves several that would transform an older person's life entirely unpromoted. A wearable, bone-conducting headphones, a larger folding screen — each marketed to the group that needs it least.

57:51 - Serving one strongly held preference the category prohibits is high-margin business

Serving one strongly held preference the category prohibits is high-margin business.

"should be another one which is like you've got a really nice gym"

— 58:03

Hotel groups run dozens of indistinguishable sub-brands while nobody offers the single attribute some guests would choose on. A chain defined by having baths, or soap, or enough hangers, is a positioning that costs nothing and nobody has taken.

1:00:34 - A dollar of usefulness buys a lifetime of recall, and rescue buys delight

A dollar of usefulness buys a lifetime of recall, and rescue buys delight.

"The other thing about hotels is they never have travel adapters or phone chargers"

— 1:00:34

A cheap object given at the moment of need is remembered for decades, because relief is the strongest surprise available. A hotel that sold the one cable a traveller could not work without is still being talked about thirty years later. Usefulness at the right moment buys recall.

Personal Branding Lessons

Two marketers spend an hour describing an industry organised into departments — finance guarding the cost line, research demanding evidence before anything is tried, marketing left holding the only view of the future. A solo operator runs all of that out of one head, which removes the meetings and removes the alibi along with them. What is left is an advantage the whole conversation circles without naming: the asset being discussed here cannot be bought out, reassigned, or defended by anyone but you.

Make the piece before anyone asks for it. When production was slow and costly, work was commissioned first and made afterwards — the jeweller took the order, the potter made the plates and then sold them. Your production cost is an afternoon, which puts you on the potter's side of that line whether you behave like it or not, and you have no client relationship to protect from what you publish. Most of what you make speculatively will go nowhere, which is the shape the returns come in rather than a sign you chose wrong. The consequence is that when someone finally describes the problem out loud, you send a link with a date on it from months ago instead of offering to write something. 00:10

Name the thing before you explain it. A café somewhere worked out that drinking heavily before eleven on a Saturday becomes a legitimate way to spend a weekend once it is called bottomless brunch — and nobody in marketing commissioned that. You have a recurring format, a repeat piece of advice, or a way of working that you currently describe in a paragraph each time. Give it a two-word name and use only that name for six months. The consequence is that someone introduces you at an event using your phrase, and the phrase does the positioning while you are standing there saying nothing. 03:05

Promise slower than you deliver. An insurer was ready to spend heavily on paying claims faster because everybody complained; it already paid in about two weeks on average, so instead it told claimants to expect the cheque within four. The cheques still went out in two, and the complaints turned into thank-you calls. Do the same with your own turnaround: say the draft lands Friday when you know it lands Wednesday, and publish the reply window you actually keep. The consequence is that the chasing email stops arriving, and the identical work now shows up early. 11:31

Run the experiment you could not have argued for. Proceeding only through sequential logic explores roughly 15 to 20% of the available solution space, and that is the part your competitors have already reached by the same route. Reserve one post, one format, or one offer a month that fails the justify-it-first test — no comparable, no precedent in your field, no case for it you could put in an email. It is a door that opens both ways: an unread post can be deleted on Sunday and costs you an afternoon. The consequence is an archive that by next year holds two or three things nobody in your niche has, precisely because none of them could have survived a meeting. 17:22

Raise the price before you widen the audience. Doubling the number of buyers probably earns less than persuading the same number to pay 30% more, because volume drags the entire cost base along with it. Your cost base is hours you cannot manufacture, which makes the arithmetic sharper for you than for any company: the second client costs you a week, the higher price costs you an awkward sentence. Rewrite the number on the next proposal before you write another outreach message, and read what you already know about why people pay what they pay. The consequence is four clients instead of eight and two afternoons a week returned to the work that got you hired in the first place. 25:18

Publish the thing your format is moving away from. There are no trends, only vectors, and every vector generates its counter-movement: retail became automated and impersonal, and farmers' markets filled up. Your feed is compressing toward clipped video and one-line takes, which means the value is draining out of that and pooling in the opposite. Write the fifteen-hundred-word argument nobody else is willing to sit down for — a position somebody is plainly prepared to defend rather than a claim thrown out. The consequence is being the one entry in a scroll of clips that a reader saves to finish on a train. 31:27

Win one room before you chase a thousand strangers. Adoption is a sigmoid because habit and copying drive it, which is why a researcher studying social contagion would rather sell five solar panels on one street than fifteen to people living ten miles apart. Pick a single room — one conference, one industry forum, one company's orbit — and be present in it every week for a quarter instead of spreading the same effort across four platforms. That concentration is what turns readers into an audience that reinforces itself. The consequence is a stranger opening a message with the line that three people on their team already follow you, which scattered reach never produces. 52:22

Address the reader your work actually serves, not the one who flatters you. Bone-conducting headphones are a game changer for someone whose hearing is a bit shot and they are marketed to joggers; the folding phone sells on screen size when the stated benefit is a screen big enough to read on. Your own positioning is probably doing this — written toward the impressive peer you want to be seen by, while the person who most needs the material assumes it is not for them. Name that person in the next three pieces you publish, in the narrowest terms you can stand. The consequence is replies from a group that has never once heard itself described in your field. 55:16

Give away the small useful object with your name on it. A hotel could buy chargers wholesale, print its icon and please take this with you on them, and for about a dollar a guest spends the rest of his life saying: remember the hotel in Istanbul, they gave me this. Your equivalent costs a weekend — the checklist, the pricing spreadsheet, the template you rebuild for every project — handed over with no form in front of it. The effort has to be visible, because a thing that plainly took work carries its own evidence in a way a claim never does. The consequence is a file sitting in a stranger's downloads folder with your name in the filename, still there on the day they finally need to hire someone. 1:00:34

Questions

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

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

How new can something be before people refuse it?

Unfamiliarity is the ceiling, not capability. An object with no recognisable precedent asks the buyer to work out what category it belongs to before judging whether it is any good, and most decline the homework; the same components inside a familiar shape sell, while doing less. Visibility sets the limit — things worn or shown in public face the strictest version of the test, and a tool used privately can afford to look odd. A personal brand arriving ahead of what its market will accept meets the same wall. 55:20

What does efficiency destroy that nobody counts?

The parts of a service nobody wrote down. A cost programme can only remove what appears on a ledger, so the staffed counter goes and the awkward case it quietly solved goes with it, along with the exchange that made a stranger feel dealt with. The mistake is entering the unquantifiable as nought rather than as unknown: the relationships built face to face never showed up as a line item, which is exactly why they were cheap to delete. 34:40

Why does adoption follow an S-curve rather than a line?

Because most buyers wait until someone they know has gone first. Habit and copying, not evaluation, drive the bulk of the curve, so the opening stretch is flat for reasons of risk rather than indifference, and contagion runs geographically — five houses on one street beat fifteen scattered across a county. The common error is judging a good product against an overnight-success expectation and killing it in the flat part, long before the spread through a network begins. 51:21

Why does insisting everything make sense in advance make you worse?

Because a demand for justification up front pre-selects conclusions rivals have already reached, having walked down the same corridor of logic. Sequential reasoning searches only the well-lit fraction of the available options, and whatever cannot be argued for beforehand is never tested at all. The rigour is not wrong everywhere: it earns its keep on commitments that cannot be undone. The error is applying it to the reversible kind, where the cheapest evidence available is simply trying. 18:39

Why is naming your work brand a mistake?

The word hands the problem to one department and excuses every other. Said in a meeting, it files the issue under communications, while the things that actually move perception — the price, the queue, the wording on a sign — sit outside that remit. Better to name the outcome instead: fewer complaints, a higher resale value, a reason to pay more. Inside the team the word is harmless; the damage lands when a remit wider than communications has to be argued for. 08:43

Why is uncertainty cheaper to fix than the thing people complain about?

Because the complaint names the symptom, and the expensive cure is the one everyone assumes is being demanded. Telling a claimant to expect payment in four weeks while still paying in two costs nothing; actually paying faster costs capital. What converts is the removal of not-knowing, not the hours saved. Taken literally, what an audience complains about sends the budget to the named metric. The limit: performance has to be acceptable already, or managed expectations merely relocate the anger. 13:07

Why is value created in the mind rather than in the product?

Two identical cars off one production line stop behaving identically once one wears a badge people trust. Expectation sets the standard the experience is measured against, so the trusted twin collects fewer complaints and higher second-hand prices with no engineering difference between them. That gap is where the money sits: in thin-margin categories, persuading the same buyers to pay around thirty per cent more beats doubling their number. Comparisons that clean are rare, which is why what people will pay gets written off as soft. 27:35

Why does progress usually happen backwards?

Most of what matters was found first and explained afterwards. Cheap, fast production allows the thing to be made and then sold, which is how songs, pottery and most art have always worked; an antibiotic and a naturalist's voyage arrived by the same route. A forward-only method rules out any attempt that cannot be justified before it happens. It holds while a single attempt stays cheap: where a few attempts return everything and most return nothing, volume of tries beats quality of reasoning. 18:04

Sources

Tom Goodwin and Rory Sutherland Fix Marketing (Part 2)

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.