The New “UGC”: How to Make Viral Contents
The line between authentic and artificial has never felt thinner.
We are Stella and Amy. We share firsthand stories and perspectives that are either lost in translation or simply inaccessible to you. Together, we bridge cultural divides and bring the world a little closer—one post at a time.
The Conversation
Ian Wu is the CEO of Accucrazy, a marketing tech startup in Taiwan. Previously, he worked at Cheetah Mobile, Musical.ly (predecessor of TikTok), and Meitu. He helped expand overseas markets for Chinese tech and digital media companies.
Ian: We've developed this AI model that mimics Taiwanese internet culture and forum-style conversations. It's been incredibly effective for brand marketing campaigns.
Amy: Is it called digital word-of-mouth marketing?
Ian: Yes. Word of mouth marketing is powerful and we help brands create word-of-mouth contents. With Large Language Models, the focus is now on Content Tech - how to make AI generate engaging, authentic content. Contents that will go viral.
Stella: LLMs like ChatGPT are usually very general-purpose and cannot generate authentic content for the local market. What data did you use to customize the model for the Taiwan market?
Ian: We trained it on successful social media posts and, crucially, the logic behind them - the humor, the cultural references, the meme timing. It can generate entire threads that feel native to the platform.
Amy: I've tried it - it's impressive! The way it writes these posts of personal stories that you often see on Reddit, but instead of Reddit-style, the content is highly localized and looks like something Taiwanese people would write... you don't even realize it's marketing content until halfway through.
Ian: Yes. People get tricked. Many still don't know they are not written by a person. It's all about having that authentic internet vibe.
Stella: So you're basically using AI to scale what your best social media managers can do?
Ian: Think of it like casting the perfect influencer, but in AI form. We carefully curate the training data to capture specific styles and emotional resonance.
Ian: The key difference is that AI doesn't naturally have intuition about what will go viral - we have to carefully design that through data selection and training. It's like casting. You are curating the next superstars and then trying to replicate the content automatically.
The Cocoons
User-generated content (UGC) represents the authentic voice of consumers - content created by real users about brands they interact with. Traditionally, this has included social media posts, reviews, images, and videos that customers create voluntarily, sharing their genuine experiences with products or services. According to Adweek, 85% of consumers find UGC more influential than brand-created content, highlighting its powerful role in modern marketing.
The Rise of UGC
The story of UGC's rise is fascinating, and we can actually trace one of its early success stories through Ian Wu, the CEO of Accucrazy and an early member of Musical.ly (now TikTok). In a recent podcast episode with Stella & Amy, Ian shared how he helped secure Musical.ly's first major commercial deal - a million-dollar campaign with Coca-Cola. The campaign's success wasn't measured in metrics like clicks or impressions, but in user-created videos. For just one dollar per video, they generated a million unique pieces of content - a revolutionary achievement that opened the doors to a new era of UGC campaigns. Brands realized that while they were spending big budgets on commercials, user-generated content could grow rapidly and organically. Users also created a network effect when they used their own profiles to promote the brand.
When AI Joins the Conversation
When everyone wants a solution to something, there's no lack of startups trying to solve the problem. Ian and his team trained an AI that can produce UGC content that's so authentic to Taiwanese internet culture that readers often don't realize it's marketing until halfway through the content. This development raises questions about the future of user-generated content. Is it still UGC if AI is the author? Perhaps the more important question is whether it maintains the authenticity and local relevance that made UGC valuable in the first place.
The Power of Intuition: Predicting Viral Content
In marketing, having an intuition for what content will go viral is a rare and valuable talent. Top marketers who can consistently predict virality of a piece of creative content are highly sought after by brands.
10 years ago, Netflix used data analytics to predict the success of "House of Cards". Now other companies are trying to replicate this ability to predict viral content through AI. However, AI doesn't naturally possess this intuitive understanding - it needs to be carefully designed through data selection and training.
The challenge lies in teaching AI not just to analyze content, but to grasp the underlying cultural nuances - the humor, references, and timing that make content resonate with specific communities. It's an attempt to codify what talented marketers do instinctively.
Is UGC still UGC?
If AI generates content, can we still call it user-generated? Traditional UGC thrives on authenticity—real people sharing genuine experiences. But when algorithms craft posts so seamlessly they mimic human voices, we risk losing the trust and rawness that make UGC impactful. Sure, if it resonates and goes viral, some might say it doesn’t matter, but isn’t that the point? When content blurs into AI-curated storytelling, are we still fostering real conversations, or just feeding a machine-driven illusion? It’s a slippery slope, and the line between authentic and artificial has never felt thinner.
Please subscribe The Cocoons for more exclusive conversation snippets and stories.
We are Stella and Amy. We share firsthand stories and perspectives that are either lost in translation or simply inaccessible to you.
AI Generated Content = AGC ???
I guess it would also depend if it's a user running the AI vs a brand or agency.