Generative Engine Optimization going mainstream

Last week in Shinjuku, hosted a small but sharp group of 25+ marketers to talk about something that’s been quietly rewriting the rules of discoverability. Yes I am talking about the arrival of Generative Engine Optimization (GEO) where LLMs are influencing the cosnumer behavior beyond simple search experiences. The session was to engage the half curious, half “uh oh, this changes everything.” audience  wanting to understand the trends on intent based experiences and Japan specific content strategy in the age of AI. Here is what we delivered in terms of the messaging.

What are the conversations in marketing circles?

Across Japan’s marketing ecosystem(October 2025), the same five conversations keep popping up no matter who you talk to. First, everyone wonders why traffic from ChatGPT is still so low, almost like the channel hasn’t “switched on” for Japan yet. Then there’s the ongoing debate about whether SEO and GEO are even different anymore, or if they’re just two names for the same fight for visibility. Add to that the growing belief that the future might simply collapse into chatbot-first experiences, making websites feel more like backend content warehouses than destinations. People point to the U.S. as proof that ChatGPT-driven shopping is already eating traditional search alive over there. And finally, the classic Japan refrain: “Yes, we’re behind… but that’s fine, we’ll catch up later.” It’s honest, it’s familiar, and it sets the stage for why these discussions matter right now.

Even though less than 1% of traffic comes from ChatGPT on average, the real action is happening inside the LLMs, not on websites. That’s the subtle shift most teams underestimate. Users aren’t bouncing around links anymore, they’re getting their answers straight from the model. And honestly, this isn’t a problem; it’s a massive opportunity. Consumers have jumped from typing 3-4 keyword snippets to firing off 20-30 word, intent-rich prompts that reveal exactly what they want. Also consumers are having atleast 4 prompts chained before they are satisfied with the answers. 

So although the site traffic from LLM might be low, consumers are having intent rich conversations (average 26 words against 1-3 keywords) which can help brands understand "the real requirements" of consumers. For the first time, brands can optimize for full-sentence intention rather than guessing what a vague keyword might mean. Site traffic has long been a vanity metric, and we challenged the audience to look beyond it and use LLMs to better understand customers. Thats where a citation analysis Adobe LLM Optimizer comes in!

What does the data say : Consumers are finally showing confidence

Now that said, there was also an interesting insight I shared from Adobe Digital Insights report. It shows that users who come to wesites via ChatGPT are high quality leads since they are better informed and convert better! Beyond search LLMs have done some heavy lifting which is causing the customer to shorten the time to conversion and the September data is already showing that. The days of Google search are not over, LLMs are just powercharging decision making.

Designing Consumer Experiences in the age of AI:

In the AI-driven landscape of 2025, consumer expectations have evolved into an "always-on" state where the Zero Moment of Truth (ZMOT) is diversified across countless digital touchpoints. Consumers now navigate a journey heavily influenced by generative AI (like ChatGPT, Gemini, and Grok), which sets a new baseline for speed and convenience. With this new expectation of consumers, brands should aspire to do more than just information provisioing; they must leverage AI-derived Intent Data infused with their own proprietary data to create differentiating "Experiences". What those "differentiations" can be is where the brand managers to study Customer Intent very closely and better align their messaging. That is what Adobe LLM Optimizer is there to address.

The ultimate goal is not just to be present online, but to use digital tools to "positively betray" consumer expectations - delivering a level of personalized care and surprise that surpasses the standardized efficiency of AI. Thats where we introduced to audience that Content strategy will be intent driven. AI to create more slop is not a strategy at all. Rather than creating more content, its important to create Just-In-Time content better aligned with customer intent. Thats the differentiator going forward

Introducing Adobe LLM Optimizer

With the perspective of Content with Intent set out, here is where Adobe LLM Optimizer comes in. Adobe LLM Optimizer is an AI-focused analytics and optimization tool designed to help brands secure visibility within generative AI platforms like ChatGPT, Gemini, and Perplexity, a practice increasingly known as Generative Engine Optimization (GEO).LLM Optimizer also allows for doing competitive analysis with other brands, sentiment analysis and more to help drive content strategy.

Unlike traditional SEO tools that focus on ranking links on a search page, Adobe LLM Optimizer analyzes how large language models (LLMs) interpret and cite a brand's content. It scans a brand's digital presence to identify "visibility gaps" where AI models fail to reference accurate information. My personal use of the tool also surprised me how the citation works, what real needs of the consumer can be uncovered and overall a beahvioral map of the consumer to work on.

Beyond the Owned Media

Media mix is becoming non-negotiable in the AI era, where visibility doesn’t come from just polishing your owned media anymore. From a Japan specific standpoint, platforms like Yahoo Chiebukuro, Note, Abema, and Hatena are stepping up as serious earned-media engines, letting brands tap into real user intent, community signals, and organic authority. When combined, they create a communications stack that isn’t just broader, it’s smarter, more resilient, and far more aligned with how people actually discover and trust information today. At the same time its super important that marketers do not cannibalize these platforms as was seen in case of Reddit which lost trust weight from ChatGPT since marketers used it as a influencing mechanism to bring in traffic without bringing in any additional value from the content and communication. Genuinity of the experience is the differentiator, thats where Google added the additional E (for experience) to the EEAT model (Experience-Expertise-Authority-Trustworthiness)