Customer Acquisition in the AI Era Is Completely Different from the Search Engine Era
Deep thoughts on AI and aspirations —— ByteDance Deep Thinking Circle
For the past two decades, the underlying logic of online customer acquisition has been SEO: get your webpage to rank high in search engines, then users click through. This logic is being quietly rewritten.
More and more users no longer use search engines to find answers—they ask AI directly. They don’t ask for “a bunch of links,” but rather “just tell me the answer.” The fate of content has changed—it no longer needs to be “clicked,” but to be “cited.” This is the problem GEO (Generative Engine Optimization) aims to solve.
The Fundamental Shift in Rules: From “Being Clicked” to “Being Cited”
In the SEO era, users wanted “a bunch of options.” The higher your webpage ranked, the higher the probability of being clicked. Content was for users to see, and optimization focused on “how to get users to click through.”
In the AI era, users want “one answer.” AI selects the most relevant pieces from countless sources and synthesizes them into a response. Now your content doesn’t need to be clicked by users—it needs to be selected, cited, and synthesized into answers by AI. Content isn’t for people to read; it’s for AI to read.
This difference determines that the direction of optimization is completely different. SEO optimizes for rankings and keywords; GEO optimizes for “can AI understand, remember, and cite it.”
How to Get AI to “Cite” You
Getting AI to select your content comes down to three actions.
First, make your content accessible to AI. Many websites block AI crawlers by default, essentially shutting themselves out. You need to confirm that your content is open to AI engines.
Second, structure your content. AI reads content differently than humans—it prefers clear structure: explicit subheadings, concise paragraphs, clear definitions. One idea per paragraph is more likely to be cited than explaining three things in one paragraph.
Third, write for both humans and AI. Don’t stuff keywords—write content that is clear, accurate, and directly answers a question. AI selects content based on “what best answers this question,” not “what has the most keywords.”
A New Opportunity, But Don’t Rush to Bet Big
This shift does contain new opportunities: a traffic entry point overlooked in the search engine era is being opened up by AI. Those who get it right early might capture traffic others haven’t noticed yet.
But stay clear-headed. AI engine rules are still changing rapidly, and today’s optimization methods might be outdated in a few months. Plus this direction is so new that how to measure results and attribute success isn’t fully figured out yet. So it’s worth attention and experimentation, but shouldn’t be an all-in bet from the start—validate on a small scale first, see if your content actually gets cited by AI, then decide how much to invest.
Putting It Into Action
If you’re doing content-based customer acquisition, three things are worth starting now:
One, confirm your content is open to AI engines (not blocked by crawlers).
Two, restructure your core content: clear subheadings, concise paragraphs, one idea per section.
Three, actively validate: use AI engines to ask about your core topics, see if your content gets cited, and how well it performs.
The fundamental rules of customer acquisition have changed. The old rule was “get people to click through.” The new rule is “get AI to cite you.” Those who adapt to the new rules first will capture traffic from the new entry points.
Key points: Traffic rules shift from “being clicked” to “being cited”; GEO requires making AI understand you (open crawlers/structure content/answer directly); new opportunities worth validating but don’t go all in; test small scale first before investing.