Why every fairy tale stepmother is evil, and what it tells us about brand narrative in the AI era

🌍 中文版: 为什么童话里的后妈都是坏人

I came across a one-minute video the other day that stopped me cold. A speaker in front of a small audience asked, simply:

“Go look at every fairy tale you know. The stepmother is always evil. Why?”

The audience threw out the usual answers. For dramatic conflict. To create pressure on the protagonist. Because kids need a villain to root against. The speaker shook his head at each one. Wrong. Wrong. Still wrong.

Then he gave the answer.

“Fairy tales are told to children. And who tells them? The real mother. Do you think a real mother is going to say kind things about a potential replacement?”

It’s not a storytelling technique. It’s not Freudian. It’s about something much more naked: the real mother has a very specific interest. She needs to hold her irreplaceable position in the child’s mind. So in every story she ever tells, anyone who might one day take her place has to be the villain. Not because they are, but because she’s the one holding the microphone.

That’s not literature. That’s narrative sovereignty.

The speaker then jumped a level. This is why every culture, every nation, has to tell its own stories. Because the moment you hand someone else the microphone, you become the stepmother in their fairy tale. And stepmothers don’t get redemption arcs.

That one-minute video did something to me. It collapsed, into a single metaphor, a problem I’ve been circling for years as a consultant to Chinese brands going global.

The real problem isn’t “your product isn’t good enough”

I’ve spent the last decade watching Chinese consumer tech brands try to crack Western markets, first at Twitter running the Greater China games business, now at Quora as head of sales for the same region. In both seats, I’ve had a recurring conversation with founders.

A brand walks in with everything stacked in its favor. The product outperforms its Western counterparts. The design is often better. The supply chain is unbeatable. The price is thirty percent lower. Two years into the US market, sales are flat.

The founder sits down across from me and asks, “Our product is objectively better. Why don’t consumers see it?”

I used to answer this with the concept of the trust gap. You lack localized trust signals. There’s no native English brand story, no founder voice, no community of real users vouching for you, no third-party validation. Consumers aren’t rejecting your product. They simply have no basis on which to decide whether to trust you.

That answer is technically correct. It’s also strangely unsatisfying. Every time I give it, I feel I’m missing an anchor. A single image that would make it stick.

That one-minute video gave me the anchor.

The real problem isn’t that the product isn’t good enough. The problem is that by the time you arrive, the fairy tale has already been written. And in that fairy tale, you’re the stepmother.

Not because you did anything wrong. The storytellers, from Western mainstream media to tech blogs to Reddit threads to American TikTok creators to, increasingly, the training data of ChatGPT, have been telling their own story for years. And their real-mother logic requires your category to sit in a fixed position: cheap knockoff, privacy threat, surveillance risk, fast fashion exploitation. These positions aren’t evaluations of your product. They’re structural roles in somebody else’s story, and they don’t move just because you arrive with better specs.

SHEIN keeps issuing PR statements that only make it look more like “the Chinese fast fashion brand everyone is watching.” Arguing with a stepmother role from inside the fairy tale makes you look more like the stepmother.

TikTok won the product war, won the user war, won the innovation war, and still lost the narrative war. It was framed as a “potential threat” by Congress and the mainstream press, and the weight of over a billion users worldwide hasn’t been enough to shift that position, because the storyteller is not TikTok.

Your brand, the one with the better product and the lower price, has been stuck for two years because of the same mechanism.

PR is not narrative. PR is paying rent.

When I tell marketing leaders this, a very specific misunderstanding keeps coming up.

“Our PR is fine. We have press hits every year. We even made TechCrunch.”

This is the most common, and most costly, confusion in the business. PR is not narrative. PR is renting someone else’s microphone for a few sentences.

You pay a publication, they write a piece, it trends for a day, and then it sinks into the feed. The next time a potential customer searches for your brand, that piece is on page three, and it probably isn’t cited by any AI assistant at all. Meanwhile, a two-year-old complaint on Reddit, a three-year-old privacy concern in the Wall Street Journal, and a random YouTuber’s “Chinese brands exposed” video are still being scraped, still being indexed, still being treated as the truth about who you are.

PR is event-driven, temporary, and filtered through someone else. Narrative sovereignty is the opposite. It requires infrastructure, not events.

What does that infrastructure look like? It looks like a founder who regularly shows up on English-language tech podcasts to talk about product philosophy, under their own name, not ghostwritten by a PR firm. It looks like a company YouTube channel with hundreds of native English videos aimed at actual overseas users, not translated Chinese scripts. It looks like a real, patient presence in one or two core communities (a specific Reddit subreddit, a Quora space, a niche forum), where brand people show up as humans over months and years, not as campaign bots. It looks like an English website that has a story page explaining why the company exists, not just a spec sheet.

Every single one of these actions is anti-efficient. They don’t produce measurable KPIs in year one. When the board asks, you can’t justify them cleanly. But what they do, quietly, is this: in every future scenario where a consumer might encounter your brand, they pre-position a version of your story that you wrote yourself.

When someone searches “best [your category] brand,” an AI can pull a positive narrative from the brand’s own story page and content library. When someone on Reddit asks “should I trust a Chinese [your category] brand?”, there are years of genuine conversations already there, conducted by accounts with real faces. When a journalist writes an industry overview, a founder’s English interview is available to quote.

This is what I mean by real-mother infrastructure. It’s not better PR. It’s a completely different thing. You tell your own story, every day, in every place where your name might come up, before anyone else gets there first.

What most Chinese brands are doing instead is dumping their entire budget into paid PR and performance ads, then scratching their heads when their narrative position refuses to budge. The explanation is simple: paying rent does not make you the landlord.

The mother has changed

For as long as modern branding has existed, the entity telling stories to consumers was the press. CNN, BBC, Wall Street Journal, New York Times, the tech blogs, the industry trades. If you wanted to influence the narrative, you fought for column inches.

That is no longer where the story is told.

More and more consumers are getting their information directly from ChatGPT, Claude, Perplexity, and Gemini. Google is actively replacing ten blue links with AI Overviews. These systems don’t hand you ten results and let you make up your own mind. They hand you a narrative. A short, confident, declarative paragraph that says: here’s the best option, here’s why, here’s what you should consider.

That’s not a search engine. That’s a storyteller.

Ask Perplexity “best portable power bank for international travel,” and it returns a three-paragraph answer. Top recommendation, runner-up, reasons. What the user sees isn’t ten candidates to evaluate. It’s a single answer that reads like advice from a trusted friend.

This is the exact structure of a real mother telling a fairy tale. The storyteller is invisible. Only the story remains.

And that story, the one the AI tells about you, is grown from its training data. The sum total of what you put on the English-language internet over the last five to ten years determines how ChatGPT, by default, talks about your brand today. If your English content footprint for the last five years is effectively zero, then when someone asks “is [your brand] trustworthy?”, the model has nothing to pull from except third-party coverage, Reddit complaints, and competitor blog posts. It will, inevitably, cast you as the stepmother, because the stepmother is the only role its memory has for you.

I want to be honest about one thing here. I’m confident about the mechanism. I’m less confident about how fast the window is closing. My best guess is that the training runs from the next eighteen months will harden the default stories for another five years. If that guess is right, then what you do in the rest of 2026 matters disproportionately. If it’s wrong, you still gain by starting now. The downside is bounded. The upside isn’t.

This is what I actually mean by GEO, Generative Engine Optimization. It is not a clever upgrade of SEO. It is not a new bag of tricks for keywords and meta tags. GEO is the fight for narrative sovereignty in the age of generative search.

SEO fights for visibility, whose page ranks higher in a list. GEO fights for something deeper: which version of your story ends up inside the AI’s memory. When someone asks about you, which story does the model tell?

Three things you can do this week

Enough of the diagnosis. Here is what to actually do.

First. Run a “which story is the AI telling about me” audit. Today.

Take your brand name and run ten standardized English prompts through ChatGPT, Perplexity, Gemini, and Google AI Overviews. Cover four categories: category-level (“best X in 2026”), brand-level (“is [your brand] trustworthy?”), founder-level (“who founded [your brand]?”), and controversy-level (“biggest issues with [your brand]”).

For each answer, record four things. Which sources did the AI cite? Where does your brand appear in the ranking, if at all? What qualifying language does it use: “affordable,” “reliable,” “controversial,” “low-end”? And most importantly, was your own voice cited even once? Your official site, your founder’s writing, your story page?

If the answer to the last question is zero, then right now, today, you are the stepmother in someone else’s fairy tale.

Second. Stop concentrating your budget in paid PR and start building real-mother infrastructure.

This is the hardest advice in the piece, because it is anti-efficient, anti-KPI, and anti-instinct for every decision-maker who came up through performance marketing. It is also the only action that actually moves narrative position.

The specific checklist:

• Founder visibility. Real name, real face, real language. Guest appearances on English podcasts, long-form posts on LinkedIn, interviews with independent writers and YouTubers. Not ghostwritten by an agency. Actually written by the founder.
• Owned media. A company blog, a newsletter, a founder’s story page. Not press releases, actual opinions. Not translations from Chinese, native English content commissioned from native writers.
• Community presence. Pick one or two core communities (typically a vertical subreddit or a Quora space) and show up as humans, over months. Not campaign accounts chasing conversions. Brand people whose default presence is “we live here.”
• A BrandFactsPage. A dedicated, structured factual page about your brand: who founded it, why it exists, product philosophy, key milestones, written in a format AI models can easily quote. This is how you proactively feed the AI your own version of your story.

Third. Use one question, “am I the real mother or the stepmother?”, to audit every content decision you make.

Not just the big strategic ones. Every tweet, every blog post, every media partnership. Ask:

• Is this action making me the one telling the story, or making someone else the teller?
• When an AI encounters this piece of content, is it my version of the story or a third party’s?
• Will this content still be around in three years, or will it disappear after the event it is tied to?

If every answer points to “someone else is telling it, a third party is speaking, it disappears after the event,” then that action is paying rent. It is not building a house.

In the original video, the speaker took the idea all the way up to national sovereignty. That was his emotional register, and it is not the one I want for brands. We don’t need confrontation. We don’t need nationalist framing. We need to understand one thing.

In every scenario where a consumer might encounter your brand, there is someone telling a story. That someone is either you, or it is someone else.

Your product, however good, cannot change the role you are assigned in someone else’s story.

So the most important shift is the one that looks slow, counterintuitive, and boring. Tell your own story before anyone else does.

The stepmother in the fairy tale never gets a redemption arc. Not because she’s actually evil. Because nobody ever let her tell her own story.

FAQ

Does this apply to B2B brands, or only consumer brands?

Both, and arguably B2B more so. B2B buying cycles are long and research-heavy. Potential customers do a lot of research before deciding, and that research path is increasingly routed through AI tools. If a buyer asks “best enterprise X solution” and the AI tells them a story about your competitor, you don’t even make it onto the shortlist.

What if I’m a small brand without a decade of budget?

Three priorities. Build a BrandFactsPage first, a week’s work. Pick one platform (LinkedIn, or a specific vertical podcast) and have your founder show up there consistently. Choose one core community and be present in it long-term. Don’t try for breadth. Try for “there is one place where an AI can find my own voice.”

How do I measure the ROI of “real-mother infrastructure”?

Not with traditional PR KPIs. Use the AI audit instead. Run the ten English prompts quarterly and track whether your own voice is being cited by AI answers at an increasing rate. That’s the only metric that genuinely reflects movement in narrative position.

About the author

Adam Yang is Head of Greater China at Quora, with over a decade of experience helping Chinese consumer tech brands expand globally, first at Twitter, now at Quora. He writes about brand trust, ad strategy, and GEO (Generative Engine Optimization) for Chinese brands going overseas at adamnote.com.