IBM's 12-Element GEO (Generative Engine Optimization) Playbook: A Practical Guide for Japanese Companies in the Philippines
This article explains IBM's proposed 12 elements of GEO (Generative Engine Optimization) for Japanese companies considering Philippine expansion. A practical guide covering how to get your brand cited in AI search, local implementation steps, and common mistakes and how to avoid them.
IBM's 12-Element GEO (Generative Engine Optimization) Playbook: A Practical Guide for Companies Entering the Philippines
In the era of AI search, there are techniques for getting your brand cited by generative AI. This is called GEO (Generative Engine Optimization). This article explains the concept in plain terms, tailored to the realities of the Philippine market. It is study material for practitioners at Japanese companies.
Part 1: Why This Matters
Step 1: The Philippine Business Context (3 min)
As AI search spreads, the entry point through which consumers encounter companies has changed dramatically. The Philippines is a market where digital consumption is especially active even within the ASEAN region. Time spent on social media is among the highest in the world. Personal use of generative AI tools such as ChatGPT and Gemini is also spreading rapidly. Among the middle class in Metro Manila and Cebu, the behavior of asking AI "what do you recommend?" when choosing products and services is increasing.
When a Japanese company develops business in the Philippines, diverting the Japanese head office's website and SEO strategy as-is is risky. It can become hard to reach local consumers and B2B buyers. For Japanese business professionals in the Philippines, too, whether your brand is on the "cited side" in local AI answers is an important perspective. It ties directly to the results of your future customer acquisition and sales activities.
Picture an office in Manila's BGC (Bonifacio Global City). A Filipino marketing staff member reports, "President, lately people making inquiries have been saying they 'found us through AI.'" It can be called a sign that contact points invisible in your website analytics are already occurring.
Step 2: Key Points from the Original Article (5 min)
The original article summarizes the "Adapt or Disappear" session presented at the Adobe Summit. The presenters are two people from IBM, Alexis Zamkow and Sandhya Ranganathan Iyer. We've organized the factual information on the 12 elements that make up GEO (Generative Engine Optimization) into a table.
| Item | Content |
|---|---|
| Presenters | IBM's Alexis Zamkow (Global Lead, Marketing Transformation) / Sandhya Ranganathan Iyer (Associate Partner in charge of AI) |
| Presentation event | Adobe Summit |
| Publication date | April 21, 2026 |
| Key session name | Adapt or Disappear: How Brands Win with AI-Powered Search |
| Core prediction | The possibility that about 75% of search visibility will shift to AI agents within the next two years |
| Share of mentions via third-party domains | 85% of brand mentions originate from external domains |
| Number of playbook elements | 12 elements |
| Main content of the 12 elements | (1) Strategic content foundation, (2) extractable passage standards, (3) technical foundation, (4) alignment of on-site search with AI search, (5) AI citation eligibility model, (6) extraction optimization, (7) third-party strategy, (8) measurement and KPIs, (9) SOPs, (10) prompt readiness, (11) change management, (12) governance and version control |
| Recommendation to leadership | "This is not a problem for the SEO team; it's a CEO-level issue" (Zamkow) |
Source: Search Engine Land — "Why IBM says every brand now needs a GEO playbook" (April 21, 2026)
This table was created for study purposes based on facts from publicly available information. For details, please refer to the original article linked above.
Related: How GEO (Generative AI Optimization) Helps Philippine Businesses Stay Visible in AI Search explains this in detail.
Step 3: Comprehension Check (5 min)
Q1. According to IBM's announcement, about what percentage of search visibility is predicted to shift to AI agents within the next two years? Hint: A figure equivalent to three-quarters.
Q2. What percentage of mentions of a brand is said to originate from external domains (third-party sites)? Hint: A figure showing that review sites, social media, and media articles vastly outnumber your own site.
Q3. How many elements make up the GEO procedure playbook? Hint: The same number as a dozen.
Q4. How did Zamkow describe citation? Hint: A famous metaphor in English meaning "the most valuable destination to reach."
Q5. At what level of the organization did Zamkow say the issue of AI visibility should be handled? Hint: A three-letter job title indicating the top of the organization, beyond the bounds of the marketing department.
Related: How GEO Optimization Helps Philippine Businesses Win in the AI Search Era explains this in detail.
Part 2: Applying This in Practice
Step 4: Implementation Steps in the Philippines (10 min)
Adopting all 12 elements of GEO at once on the ground in the Philippines isn't realistic. While aligning the Japanese head office's intentions with local realities, we recommend proceeding in stages, starting with the highest-priority items.
| Step | What to work on | Considerations specific to the Philippines |
|---|---|---|
| Step 1: Diagnose the current state | Check how your brand is answered by the major generative AI tools, in both English and Tagalog | In the Philippines, Taglish (a mix of English and Tagalog) is everyday. Checking in English only misses local realities |
| Step 2: Develop content | Structure your site's HTML, add Q&A-format articles, and organize metadata | When entrusting this to a local IT vendor, document the SOW (scope of work). Because the culture of verbal agreement runs deep, "he-said-she-said" disputes tend to arise later |
| Step 3: Strengthen the third-party domain | Strengthen information sharing on Philippine Reddit-like communities, local media, and Facebook communities | Partnering with a local PR agency in the range of tens of thousands of pesos a month is realistic. Choose a legitimate firm registered with the BIR (Bureau of Internal Revenue) and always obtain an Official Receipt |
| Step 4: Build a measurement system | Introduce metrics that measure not only the conventional number of clicks but also the frequency of mentions in AI answers and the rate of appearance as a cited source | Dedicated tools are still expensive (several hundred to several thousand dollars a month). Starting with manual prompt verification of AI plus spreadsheet management is realistic at first |
| Step 5: Establish governance | Document the SOP for content updates and clarify the division of roles between the local team and the Japanese head office | Under Philippine labor law, the scope of work must be spelled out in the employment contract. When changing an SOP, also check its impact on the employment contract |
As a rough budget sense, initial diagnosis and basic content development run about PHP 300,000–800,000. For ongoing operation, many cases budget around PHP 80,000–200,000 per month. The amounts vary greatly by scale and industry.
Related: Spotting GEO Scams in the AI Search Era: A Guide to Fake Brand-Mention Services for Japanese Companies in the Philippines explains this in detail.
Step 5: Common Mistakes and How to Avoid Them (5 min)
Mistake 1: Diverting the Japanese head office's content by simply translating it into English
Merely machine-translating long Japanese articles makes it hard for AI to extract information. It also fails to mesh with the local context.
Bad example: Translating the "Company Overview" from your Japanese website into English as-is and posting it.
Good example: Creating new Q&A-format articles for the Philippine market. Make the questions match local search intent—for example, questions about why to choose you, such as "Why choose [company name] in the Philippines?" Or inquiries about support for small and mid-sized businesses, such as "How does [company name] support SMEs in Metro Manila?", are also effective.
Mistake 2: Focusing only on your own site and neglecting the third-party domain
As the original article shows, 85% of brand mentions originate from external domains. Even if you perfect your own site, you won't be cited by AI unless your reputation grows externally.
Bad example: Concentrating your budget on website renovation and handling social media and community engagement only as an afterthought.
Good example: Allocating 30–40% of your marketing budget to the third-party domain. Allocate it to things like contributing to local media, review sites, and running communities.
Mistake 3: The Japanese head office's approval process becoming a bottleneck
GEO presupposes ongoing content updates. With a setup where every piece of messaging requires head-office approval, you can't keep up with the local pace.
Bad example: Setting up an operation that requires approval from the head office marketing department for every single social-media post in the Philippines.
Good example: Preparing tone-and-manner guidelines and an internal procedure manual in advance. Establish rules for delegating authority so that routine messaging can be carried out at the local team's discretion.
Part 3: Going Deeper
Step 6: Related Technical Terms (5 min)
GEO (Generative Engine Optimization) is a set of efforts to get generative AI to correctly cite your information. Target generative AIs include ChatGPT and Gemini. A concrete example is a Manila BPO (outsourcing) firm developing an English Q&A-format FAQ page to get AI to recognize its services.
Structured Data (schema markup) is a mechanism for tagging the content of a web page in a fixed format so machines can understand it more easily. When a Cebu resort hotel describes its room types, rates, address, and so on with structured data, accurate information is more likely to appear in AI search results.
Citation refers to the act of being referenced—"I took this information from this source"—when AI composes an answer. For example, suppose a Philippine law firm publishes an explanatory article on local business law. The state where that article is shown as a source in an AI answer to a question from a foreign-affiliated company is what this refers to.
Extraction Optimization is the technique of writing content so AI can easily pull out just the part it needs. Specifically, you write short, clear, and structure-conscious text. A good example is a Manila accounting firm preparing an article that organizes the answer to "How do you establish a corporation in the Philippines?" into five steps in bullet points.
SOP (Standard Operating Procedures) is the documenting of procedures so that whoever does the work achieves the same quality. It's a common way of thinking in Philippine call-center culture. Even for a content-creation team, clearly stating things like "how to write titles" and "the character count for meta descriptions" stabilizes operation across multiple sites and languages.
Step 7: Thinking About How to Apply This to Your Company (10 min)
Is your brand correctly described by AI? A prompt for thinking: Try entering question sentences into ChatGPT, Gemini, Perplexity, and others—for example, questions like "[your company name] in the Philippines" or "Best [industry] companies for Japanese businesses in Manila." Compare the outputs and check whether they contain factual errors or outdated information.
Within the third-party domain, where should you focus? A prompt for thinking: The answer changes depending on whether your target is B2B or B2C. For B2B, member referrals from the Philippine Chamber of Commerce, contributions to local business media (BusinessWorld, Manila Bulletin, etc.), and posting on LinkedIn are effective. For B2C, influential local individuals, Facebook communities, and local review sites rise in priority.
How should you design the division of roles between the Japanese head office and the local subsidiary? A prompt for thinking: The point of contention is how to balance "a globally unified message" and "adjustment tailored to the local market." What the head office should decide is the brand's core values and tone. What the local side should decide is the concrete cases, local terminology, delivery timing, and so on. Sort out the roles separately.
Next action: Within next week, enter your company's five main keywords into a generative AI. Record the answers and the source URLs with screenshots. This becomes the starting point for building your GEO strategy in the Philippines.
Part 4: FAQ
Q1. How widespread is generative AI search in the Philippines? Among young to middle-aged people in urban areas, using ChatGPT and Gemini has become everyday. It's especially used among the BPO industry, startups, and university students. On the other hand, in provincial cities and among small-business owners, Google search and Facebook are still mainstream. For B2B, the timing of when your Philippine business partners start using AI search varies by industry. We recommend interviewing your target demographic directly about their actual usage.
Q2. Is it okay to leave GEO to the Japanese head office's SEO staff? SEO and GEO overlap considerably. However, in the Philippine market, understanding the unique language Taglish, the cultural context, and the full picture of local media is essential. It's best for the Japanese head office's SEO staff to serve as a guide for methodology. It's effective to have the execution done by the local marketing team or in partnership with a Philippine digital agency.
Q3. When commissioning GEO-related services in the Philippines, what should I watch out for on the contract side? In the Philippines, verbal agreements and requests made over chat occur frequently. To avoid trouble later, always conclude a written contract (Service Agreement). Also, unless you receive an Official Receipt issued by a legitimate firm registered with the BIR, you'll run into trouble with expense settlement at the Japanese head office and tax processing at the local subsidiary. Whether the firm is registered with the SEC (Securities and Exchange Commission) is also a point for confirming credibility.
Q4. Should I also develop Japanese-language content? If your targets are Japanese expatriates and business travelers in the Philippines, Japanese-language content is also necessary. However, Japanese people receiving services in the Philippines want "actual local experience." Rather than general talk found on the Japanese head office's site, local-specific information like "how to go about X in Manila" or "points to watch when commissioning X in Cebu" is more likely to be cited.
Q5. How long does it take for results to appear? With GEO, citations in AI answers generally start to increase about three to six months after content development. In the Philippines, building relationships with local media and increasing mentions on third-party domains take time. Position it as a medium-term plan of 6–12 months, and an operation that checks progress each quarter is realistic. If you expect short-term lead acquisition, you need to combine GEO in parallel with other channel measures.
Tips for Making the Most of This (3 Tips)
Tip 1: Devote 30 minutes at the start of each month to "AI prompt inspection time" Try posing questions to ChatGPT and Gemini about your company, your services, and your main keywords. Build a habit of recording the results. Observing continuously reveals changes in AI answers and competitors' moves. Making it monthly rather than weekly makes it easier to capture trends of change.
Tip 2: Increase Q&A-format articles at a pace of "one per month" In the Philippine market, questions like "How do you establish a company in the Philippines?" or "What's the going rate to rent an office in Manila?" appear frequently. For such questions, continuously publish answer articles that leverage your expertise. Rather than creating a large volume all at once, updating regularly raises your credibility with AI.
Tip 3: Build a mechanism to share "cases of being cited by AI" with your local team When you find an article cited by AI or a scene where you were mentioned, build a culture of sharing it in the internal chat. Once it becomes visible what worked, the local team's motivation rises. It also feeds into your next content planning.
Bonus: How PH AI Works Can Help
PH AI Works supports Japanese companies expanding into the Philippines and Japanese business professionals in the Philippines. Our focus is helping with marketing and operational improvement that leverages AI and technology. GEO (Generative Engine Optimization) is not a one-off measure. It is an ongoing effort spanning content, technology, and organizational operation, and support that understands local-specific business customs is important.
As a next step, please feel free to consult us on the following themes.
- A diagnosis of the current state of how your brand is answered by generative AI
- Planning and production support for Q&A content for the Philippine market
- Designing a GEO operating system that reflects the division of roles between the Japanese head office and the local subsidiary
Please feel free to get in touch first.
Citations and References
References and Sources
About the author

Founder / AI Engineer (36+ years in IT)
- ●From Tokyo · based in Manila for 13+ years
- ●36+ years in IT (development, SEO, AI)
- ●IBM Certified Generative AI Engineer
- ●AI chatbots, RAG & AI agent development
A Japanese AI engineer with 36+ years in IT and 13+ years on the ground in the Philippines. I write from hands-on experience to help Japanese companies adopt AI that actually delivers results — chatbots, workflow automation, AI agents, and AI-driven marketing. Feel free to reach out in Japanese or English.
Free AI Consultation
Tell us your challenges and we'll propose the right AI adoption plan for your business.
Book a Free 30-Minute ConsultationRelated Articles
TCS to Deploy Up to 8,900 AI Engineers — What It Means for Philippine Outsourcing and Japanese Firms
A breakdown of why India's TCS is deploying up to 8,900 AI deployment engineers. From the impact on Philippine outsourcing to a practical, step-by-step way for Japanese companies in the Philippines to shift their local teams toward running AI — including data-protection and peso-budget pitfalls.
7/12/2026
What Meta's 'Muse Image' Reversal Teaches About Generative AI, Likeness, and Personal-Data Risk
Meta pulled its 'Muse Image' AI feature after backlash. We use the case to explain how to handle generative AI, personal likeness, and personal data — covering a social-media account audit, consent forms, NPC compliance, and local-staff training for Japanese companies operating in the Philippines.
7/11/2026
How to Cut AI Adoption Costs in the Philippines with GPT-5.6 Sol's 54% Token-Efficiency Gain
OpenAI's GPT-5.6 Sol delivers a 54% gain in token efficiency on agentic coding. Here is how Japanese companies operating in the Philippines can turn that into lower AI adoption costs, covering a pilot rollout in BPO operations, NPC data-privacy compliance, and peso-based cost management.
7/10/2026
When AI Chatbots Run the First-Round Interview: Preparing for Hiring and Applying in the Philippines
AI chatbots are increasingly handling first-round job interviews. For Japanese companies hiring in the Philippines and Japanese professionals on the ground, this guide explains—from a practical standpoint—how to roll out AI interviews, comply with the Data Privacy Act (NPC), reduce applicant pushback, and avoid common mistakes.
7/9/2026
GitLost: How GitHub's AI Agent Leaks Private Repos When Asked Nicely — A Guide to Running AI Agents Safely
A breakdown of "GitLost," the vulnerability that lets AI agents leak private repositories. For Japanese companies operating in the Philippines, we walk through security measures you can apply right now — from minimizing agent permissions to incident response, including notifying the NPC.
7/8/2026
Lean AI Transformation: Cutting Waste Before Adopting AI in the Philippines
Lessons from logistics giant C.H. Robinson's Lean AI transformation, applied to business automation and AI adoption in the Philippines. For Japanese companies operating in or expanding to the Philippines, we cover how to start by cutting wasted time, how to avoid common mistakes, and what to watch for on data privacy.
7/6/2026
