How Satsuma Creative makes ads
It starts with that lunch during the "Sha Hen Da" (殺很大) campaign. In thirty years of advertising, I've relied not on methodology, but on selling directly to bold bosses and on observing people. AI can help — but making an engineer at lunch furrow his brow and mutter "what the hell?" — that's still hard for AI.
—— Inside the 4A advertising world (Part 4)
TL;DR
- The day the "Sha Hen Da" ad aired in 2009, a stranger — an engineer at lunch — glanced up, frowned, and said "what the hell?" That was the moment I knew the ad had landed.
- The reason I've made hit ads for thirty yearsisn't because I'm better than anyone else — it's because I was lucky enough to meet bosses willing to pull the trigger
- Pitches should be sold tothe boss who has decision-making power— not the CMO, not the brand manager. Real decisions don't get made in meeting rooms; they get made over dinner.
- A method an elder taught me: to build a client's brand, first spend a week with the boss.Design first, then back-fill the methodology to package it
- AI can do slogans, KVs, scripts. But "making an engineer at lunch furrow his brow and mutter 'what the hell?'" — that's still hard for AI. The heart of advertising isa precise touch from one living person to another」
- Easter egg: "Satsuma" (薩摩) said in Taiwanese sounds like "san-bóo" (三某), not "san-siâu" (三小).
This is the final piece in the four-part series "Inside the 4A advertising world."Finale · Series navigation ↓
In the previous three pieces I've covered the industry's inside story, how the 4A system packages its work, how Big Ideas get ground down in meeting rooms, and how the industry is being eaten by AI and consulting firms.
The tone has been a bit cold.
This one shifts the tone. I want to talk about my own company — Satsuma Creative — and how we've made ads for the past thirty years.
Not as a sales pitch. It's just that after three pieces explaining "why this industry struggles," somebody ought to talk about "whether there's another way."
It starts with that lunch during "Sha Hen Da"
In 2009, I was working on a TV ad for an online game called "Sha Online."
Back then online game ads were everywhere, and they all looked the same — a handsome guy with a big blade, a beautiful woman in armor, exploding effects in the background, a voiceover saying "the most thrilling Three Kingdoms game, the most spectacular battles, join now for limited-edition treasures." Every single one. Every single one forgotten.
I told the client we couldn't do the same thing again.
I wrote a script — get Yao Yao (back when she wasn't famous yet) sitting on a coin-operated rocking horse, shaking, with the line "Sha hen da, sha bú yòng tsîⁿ" ("Kill it big, killing's free"), followed by "Don't die, don't die."
Even our own team thought those three characters were strange.
"Sha hen da" wasn't a catchphrase at the time.It became one because of this ad.But before it aired, no one knew it would. When we pitched it, even our team had butterflies — could we really do this? Wouldn't people complain? What does this even have to do with the game?
The day the ad aired, I was having lunch with the team downstairs from our office. The TV happened to be playing our ad.
An engineer next to us — not from our company, just at the next table — glanced up, shook his head, and said two words:
"What the hell?"
Our whole table burst out laughing.
In that moment I knewthe ad had landed。
Not because he got it, but becauseit hit him.After watching it, he couldn't help turning to the person next to him to say "what the hell?" That's the contagion of advertising — the moment an ad makes a stranger unable to keep quiet, its job is done.
The rest is history. "Sha Hen Da" became one of the most famous ads of that year, Yao Yao became a star, the phrase "sha hen da" became a national catchphrase, and the NCC convened more than once to discuss it (I'm guessing — because the NCC later said the ad was fine). The ad even has a Wikipedia page.
But every time I think back on it, the most vivid image isn't awards night — it's that lunch. That moment when a stranger engineer glanced up, frowned, and said "what the hell?"
That's what advertising should look like.
Why I've been able to make these ads
I've been in advertising for thirty years. The reason I could make "Sha Hen Da," do Taiwanese-Japanese ads, make "Master, let's go fetch the scriptures" with the male/female lead roles flipped — a female Wukong protecting a male Tang Sanzang —
isn't because I'm better than anyone else.。
It's because I was lucky. I met a group of bosses who were willing to let me run.
In the first piece I said: clients say they want a Big Idea but in their heart they want nothing to go wrong. That's true 90% of the time. But there's still that 10% of clients out there — the ones who'll bet, who'll let you go wild once, who believe "looking like everyone else means losing."
Over thirty years, I've happened to meet a few of those bosses.
They don't care how thick your methodology deck is, don't care whether you cite Ogilvy or Leo Burnett, don't care whether your strategy doc has a SWOT analysis attached. They care about one thing —Will this ad make my product memorable?。
The boss behind "Sha Hen Da" was that kind of person. In an era when no one dared use Taiwanese-Japanese as ad copy, the boss who greenlit it was that kind of person. The version of "Master, let's go fetch the scriptures" with the gender flip — the one that made viewers pause for three seconds before they realized — happened because the boss said, "I don't quite get it, but I find it interesting. Let's do this version."
These bosses share a common trait —They trust their own judgment.。
They don't need to see a concept test report before nodding, don't need unanimous agreement from marketing, don't need to explain their choice to a board. They see something, their gut says "yes," and they do it. If it goes wrong, they take the hit. If it goes right, they enjoy it.
These bosses are getting rarer.
But as long as some still exist, advertising is still interesting.
Who are pitches actually sold to?
In the second piece I asked — who is a pitch actually sold to? I went on about CMOs, brand managers, cross-departmental meetings.
But I wasn't telling the whole truth.
The truth is: pitches should be sold to the boss with decision-making power.。
Not the CMO. Not the product team. Not the brand manager.The person who pulls the trigger.。
During my years working on game ads, I made the rounds at every drinking session during gaming expo week. The daytime VC dinners, player meet-and-greets, media events — I went. The private after-hours drinks hosted by each game company — I went.
Not because I like socializing. Ihatesocializing.
But I went.
Because I knew one thing —Real decisions don't get made in the daytime meeting rooms.The people in those rooms are there to execute what the boss has already decided. Real decisions are made over dinner that night, at the second round of drinks, in the third-round KTV room.
I went to those gatherings not to drink. I went tolet the game company bosses get to know me.。
To let them see how I drink, how I talk, how I describe the ads I've made. To let them feel "this guy is reliable," "this guy gets us," "what this guy says, I can trust."
Once they feel that way, the next day in the office they tell their team, "Send the ad work to Satsuma" —
That single sentence is worth more than a hundred decks and a hundred cross-departmental meetings from our side.。
This isn't an unwritten rule. It's the nature of the advertising business.
Advertising is a high-risk decision — expensive, uncertain outcome, blame if it fails. Any rational mid-level employee making this decision will lean toward "the safest option" — the largest agency, the most standard methodology, the ad least likely to go wrong.
Only the boss will make the "unsafe but potentially huge" choice.Because only the boss can absorb the failure.。
So if you want to make brave ads, you have to sell directly to brave people.
A method an elder taught me
When I was just starting out, an elder taught me how to do a CIS (corporate identity system) for a company.
He said: "Don't show up at the client's office with a stack of books, methodologies, and color psychology theories. Just spend a week at the company, hanging around with the boss."
"Hanging around for what?" I asked.
"See what he likes to eat. Whose songs he likes. Which movies left the deepest impression. What books are on his shelf. What paintings hang in his office. Where he likes to travel. How he spends his weekends.How this person lives is how his company will be built.。」
"And then?"
"Then you design the CIS based on these observations. After it's done,you use 4A methodology to package the design."。」
I was in my early twenties. I was stunned for three seconds.
Back then I thought design was derived from methodology — from consumer research, brand positioning, market analysis, competitive scans, SWOT, 4P — step by step, until eventually a logo, a palette, a typography system emerged.
The elder told methe order was backwards.。
Design grows directly from understanding a person. Methodology gets layered on afterwards, to explain the design, to get the client to sign off.
I've used this principle for thirty years.
When building a brand for a client, I don't start with market reports. I start with dinner — several dinners, long ones. I listen to him tell his founder's story, why he started the company, what he's proudest of, what he fears most, what he wants to leave to his daughters and sons.
By the end of listening,the shape of the brand has grown inside me.。
The rest — the strategy doc, positioning analysis, brand personality matrix — those are for his team, for his board, for his own peace of mind.
Not useless.Just don't get the order wrong.。
After becoming a client
I also spent a stretch as an in-house client.
During that time, every time an agency came to pitch, I got impatient. I'd cut them off and say, "Skip the methodology part. Just tell me what your idea is."
The agency team would freeze. Their strategy doc was already printed, the methodology framework already in the deck, the planner already prepared to talk for twenty minutes — and suddenly they were stopped.
But I really didn't want to hear it.
It's not that methodology is useless.It's boring.It's bloated. I already know what you're going to say — you'll start with market overview, then consumer insight, then SWOT, then brand positioning, then communication strategy, and only at the very end, the creative.
Every agency does this. Every agency's deck is structured this way. Every agency pretends they arrived at their unique conclusion by following the process.
Butthe conclusion was there first.The methodology was back-filled.
During my time as a client, I learned one thing —Good agencies don't need to perform like this.A good agency sits down and lays out the core idea in the first five minutes, then uses the rest of the time to discuss feasibility, revisions, and execution. A bad agency needs to perform for a full hour, because they're afraid you'll think they "haven't done any work."
So later, on the Satsuma site, I wrote a line — "We brief fast."
What it means: tell me what you want to do, and in five minutes I'll tell you how I'd do it.
Because I don't waste words.。
Wasting words wastes your time, my time, and the world's time.
AI and creativity
In the third piece I said AI is eating the first piece of advertising's territory — the system for observing human decision behavior. I'm building AI systems myself now.
So someone might ask — if you're doing AI too, what value does Satsuma Creative still have?
My answer is:
AI can help everyone organize things, produce things, and do all the repetitive, regular, structurable work faster and better.。
But creativity —is really hard.。
By "creativity" I don't mean "thinking up a slogan," "designing a KV," or "writing a script." AI can do all of these now, and is getting better at them.
What I mean by creativity is —An engineer at lunch glances up at the TV, frowns, says "what the hell?", and can't help but turn to the person next to him to say so.。
That kind of thing is hard for AI.
Because the essence of that thing isa precise touch from one living person to another.Someone has to have been there, on the ground, feeling some air in Taiwanese society — some anxiety, some absurdity, some emotion that hasn't been named yet — and then crystallize that feeling into a sound, an image, a moment.
AI hasn't "lived." It has no memory of crouching by the curb eating a bento, no memory of watching someone get drunk in a KTV room, no memory of being still at it at 3 a.m. during gaming expo week. It has no flesh-and-blood experience of that "what the hell?" moment.
It can learn how to use the phrase, learn in what context "what the hell?" appears, learn to mimic the syntax. Butno one was actually there in that moment.。
The heart of advertising has never been the deck, the methodology, the strategy doc, the execution timeline.
The heart of advertising is —One person created a moment, and another person caught it.。
That's still hard for AI.
At least, still hard today.
So Satsuma Creative will keep going. We'll use AI for the repetitive, structurable, boring parts — so we have more time for the things only living people can do.
By the way, the name "Satsuma"
One last easter egg.
A lot of people ask why we're called Satsuma.
Satsuma is an ancient domain in Kyushu, Japan — the Satsuma Domain, one of the strongest powers at the end of the Edo period. Saigō Takamori and Ōkubo Toshimichi were both from Satsuma. I love that spirit — "on the periphery but bold, not at the center but able to change the center."
But that's the official reason.
The real reason is an easter egg.。
The characters "Satsuma" (薩摩), said in Taiwanese —
sound a lot like "san-bóo" (三某).
Not "san-siâu" (三小).
It's "san-bóo" (三某).
One character different. A world of difference.
Those who get it will laugh. Those who don't, no problem.
That's advertising —Some things only land for those who get it.But it doesn't lose an ounce of value just because some don't.
It belongs only to those who happen to catch it.
2026, Taipei