Six GEO Business Opportunities and WICHI's Choice

Where the GEO Market Stands Today

GEO (Generative Engine Optimization) was academically defined at ACM SIGKDD in 2024, and VC capital began flowing into commercial tools in H2 2025. As of March 2026, confirmed cumulative funding for GEO-specific SaaS exceeds $174.5M. The signal that a market is forming is unmistakable.

But a single number is insufficient to judge market maturity. The ecosystem that SEO built over 20 years — infrastructure, analytics tools, optimization software, agencies, certifications, specialized media — does not yet exist for GEO. Semrush’s market cap sits at roughly $3B. Google Analytics certification holders number in the hundreds of thousands. Search Engine Journal draws millions of monthly visitors. Compared to this mature SEO ecosystem, GEO is just beginning to take shape at the first one or two stages of its value chain.

This nascency creates both opportunity and risk simultaneously. The opportunity: many seats are empty. The risk: which of those seats will actually generate revenue remains unproven.

timeline
    title GEO Market Maturity Timeline
    2024 : Academic definition (ACM SIGKDD)
         : Aggarwal et al. propose GEO-Bench
    2025 H1 : Early tools emerge (Profound, Peec)
            : VC interest begins
    2025 H2 : Series A rounds concentrate
            : Cumulative funding passes $100M
    2026 Q1 : 10+ dedicated SaaS confirmed
            : Agencies begin adding GEO services
            : Education/certification is a blank space
    2027~ : Unpredictable
          : Fork between standardization and fragmentation

This post analyzes six identifiable business opportunities in the GEO market, evaluating each against estimated market size, competition level, entry barriers, revenue models, and time to revenue. It then documents why WICHI chose a specific direction and the decision-making structure behind that choice.

One premise to state upfront: market size estimates are indirect projections based on analogous SEO market segments. No direct survey data for the GEO market exists as of March 2026. All figures carry inherent uncertainty and should be used as reference points only. The core of this post is not precise market sizing but the structural differences among the six opportunities and the decision logic built on them.

Analytical Framework

Five evaluation axes are established to compare the six opportunities:

Axis	Definition	Measurement Basis
Market Size	TAM estimate	Ratio to corresponding SEO market + independent estimation
Competition	Current player count and capital	Confirmed companies, cumulative funding
Entry Barrier	Technical, capital, and network requirements	Minimum Viable Product (MVP) build cost and timeline
Revenue Model	Revenue generation structure	Subscription / per-transaction / retainer / license
Time to Revenue	Duration until first paying customer	Product launch > first payment

This framework is imperfect. Market size estimates carry significant uncertainty because the GEO market is still being defined. SEO market analogues serve as references, but the assumption that those ratios transfer directly to GEO is not made. Each estimate includes both its rationale and its limitations.

quadrantChart
    title GEO Business Opportunity Positioning Map
    x-axis "Low Entry Barrier" --> "High Entry Barrier"
    y-axis "Low Scalability" --> "High Scalability"
    quadrant-1 "High Barrier + High Scalability"
    quadrant-2 "Low Barrier + High Scalability"
    quadrant-3 "Low Barrier + Low Scalability"
    quadrant-4 "High Barrier + Low Scalability"
    "SaaS Platform": [0.55, 0.80]
    "Agency": [0.25, 0.30]
    "Education/Certification": [0.30, 0.50]
    "Data Infrastructure": [0.85, 0.90]
    "Vertical Specialization": [0.60, 0.55]
    "GEO + PR": [0.50, 0.45]

Detailed Analysis of Six Business Opportunities

1. GEO SaaS Platform

Software that measures and tracks brand visibility across AI search engines. The GEO equivalent of what Semrush, Ahrefs, and Moz represent in SEO.

Market Structure

Item	Detail
Market size estimate	SEO analytics tool market ~$7B (2025). Assuming 5-15% GEO conversion rate: $350M-$1.05B (2028-2030)
Current competition	Profound ($23.5M), Peec AI ($21M), Scrunch AI ($19M), Gumshoe AI ($2M), Writesonic, Otterly + 5 others
Gap	No platforms specialized in Korean, Japanese, or Chinese AI search
Target AI engines	Global: GPT, Gemini, Perplexity, Claude / Local: Naver AI, Clova X, domestic LLMs
Revenue model	SaaS subscription (monthly/annual). SMB $49-$199/mo, Enterprise $500-$2K/mo
Core moat	AI response data accumulation + local AI engine coverage
Entry barrier	Medium (API integration, data pipeline, UI/UX required. Not yet an infrastructure scale competition)
Time to revenue	6-12 months (MVP launch > early adopter acquisition > paid conversion)
Risk	Global platform i18n expansion, AI engine policy changes, absence of measurement standards

Competitive Landscape

Global GEO SaaS players ranked by funding:

Company	Funding	Round	Core Positioning	Primary AI Engine Coverage
Exa	$85M	Series B	AI search infrastructure (API)	Proprietary search engine
Bluefish	$24M	Series A	Retail/e-commerce specialized	GPT, Gemini, Perplexity
Profound	$23.5M	-	Enterprise comprehensive	GPT-5.2, AI Overviews
Peec AI	$21M	Series A	Multi-platform monitoring	ChatGPT, Perplexity, AIO
Scrunch AI	$19M	Series A	Brand narrative + AXP	Multiple engines
Gumshoe AI	$2M	Pre-seed	Undisclosed	Undisclosed

These players share a common trait: all target English-language markets first. Their analyzed AI engines — ChatGPT, Gemini, Perplexity, AI Overviews — are English-based. No platform systematically covers Asian local AI search engines as of March 2026.

Notably, these companies’ positioning strategies are already diverging. Exa occupies the infrastructure layer, Bluefish targets the vertical (e-commerce), Profound pursues enterprise comprehensive, Peec AI specializes in analytics, and Scrunch AI focuses on content delivery optimization (AXP). This differentiation reveals that the GEO market is not a monolithic “GEO tools” category but is segmenting into subcategories, just as SEO did. This segmentation pattern provides entry paths for later entrants to focus on specific segments.

Technical Components

The core technical building blocks required for a GEO SaaS platform:

Component	Function	Technical Difficulty
Multi-AI engine query execution	Send identical queries to multiple AI engines, collect responses	Medium (API integration)
Response parsing and structuring	Extract citation sources, brand mentions, entities from AI responses	High (NLP required)
Visibility score calculation	Compute quantitative scores combining citation frequency, position, and context	High (methodology design)
Time-series data storage	Track visibility changes on daily/weekly/monthly basis	Medium (time-series DB)
Dashboard visualization	Customer-facing analytics dashboard UI	Medium (frontend)
Competitor benchmarking	Compare competitor visibility for identical queries	Medium (data collection expansion)
Alert system	Real-time alerts on sudden visibility changes	Low to medium

The hardest component is “visibility score calculation.” Since GEO measurement standards have not been established, each SaaS must develop its own scoring methodology. Aggarwal et al.’s (2024) GEO-Bench provides an academic foundation, but translating it into a commercial-grade scoring system requires separate design work.

Why SaaS

The structural advantages of the SaaS model come down to three factors:

First, marginal cost is low. After infrastructure is built, the cost of adding one more user is limited to incremental server expenses. Consulting costs scale linearly with headcount. Education requires ongoing content production costs. SaaS is a fixed-cost structure.

Second, data network effects apply. As the user base grows, tracked queries and AI response data accumulate, improving analytical accuracy. Higher accuracy attracts more users, creating a virtuous cycle.

Third, it is measurable. The service is built around a quantitative metric: the GEO Score. Agencies and education programs struggle to prove “this strategy worked.” SaaS can show on a dashboard: “this number rose 12% versus last month.” Measurable services have higher retention rates, and value demonstration is straightforward.

The core value proposition of SaaS is “repeatable measurement.” Customers renew monthly subscriptions because “the numbers are visible.” SaaS is the only model in GEO that can deliver these numbers.

2. GEO Agency/Consulting

A service business that develops and executes GEO strategies on behalf of clients. Just as thousands of SEO agencies exist, GEO agencies could proliferate at scale as the market matures. But for now, it is early days.

Market Structure

Item	Detail
Market size estimate	Global SEO agency market ~$80B (2025). Early GEO conversion 1-3% = $800M-$2.4B (2028)
Current competition	Dozens of GEO-specialized agencies. Genevate, Intero Digital, etc. Rapid growth in existing SEO agencies adding GEO services
Revenue model	Monthly retainer. SMB $500-$3K/mo / Mid-market $3K-$10K/mo / Enterprise $10K-$20K/mo
Core moat	Case study accumulation (industry-specific GEO performance data), personal networks
Entry barrier	Low (GEO knowledge + sales ability. No technical infrastructure required)
Time to revenue	1-3 months (fast. Revenue starts immediately upon landing the first client)
Risk	SEO agencies are expanding into GEO services rapidly. The differentiation window is narrowing

The Agency Service Spectrum

Not all GEO agencies offer the same depth of service. Three distinct levels:

Level	Service Scope	Price Range (monthly)	Required Capabilities
Level 1: Diagnostic	AI search visibility audit, competitor benchmark report	$500-$1.5K	GEO tool proficiency
Level 2: Strategy	Visibility audit + GEO strategy development + content optimization guide	$1.5K-$5K	Domain expertise + content strategy
Level 3: Execution	Strategy + content creation/modification + continuous monitoring + performance reporting	$5K-$20K	Full-service team (strategy + production + analytics)

Most early GEO agencies start at Level 1. Since Level 1 essentially means interpreting SaaS tool outputs and packaging them as reports, entry is easy. But this ease is precisely the problem. As SaaS tools provide more intuitive dashboards, Level 1 agencies lose their reason for existing. Delivering sustained value requires moving to Level 2 or above, which dramatically increases the talent and expertise requirements.

Structural Weaknesses

The agency model is easy to enter but hard to defend. Three reasons:

First, it is labor-dependent. Revenue is proportional to consultant headcount. When clients grow 10x, staffing needs grow nearly 10x. Margins plateau at a certain level.

Second, switching costs are low. The cost for a client to switch from Agency A to Agency B is minimal. SaaS has data migration and learning curve friction. Agencies can be replaced at the end of a contract.

Third, incumbent SEO agencies enter fast. Existing SEO agencies with established client bases, sales channels, and brand awareness can add “GEO services” to their menu within months. Independent GEO agencies’ time advantage erodes quickly.

Comparison	SEO Agency Expanding to GEO	Independent GEO Agency
Existing client base	Yes (upsell opportunity)	None (must acquire from scratch)
Brand recognition	Established in SEO market	Starting from zero
GEO expertise	Learning (may lack depth)	High (dedicated specialization)
Price competitiveness	Can offer bundle discounts	Must justify premium pricing
Expansion speed	Fast (existing infrastructure)	Slow (building everything new)

Agencies are useful for proving “the GEO market exists,” but their defensibility as standalone businesses structurally weakens over time.

3. GEO Education/Certification

A business providing GEO training programs and professional certifications. In SEO, Google Analytics certification, HubSpot Academy, and Moz Academy built the education market over many years.

Market Structure

Item	Detail
Market size estimate	Global digital marketing education market ~$12B (2025). GEO-specific education is near $0
Current competition	Official GEO certification programs = 0. Only individual blog/YouTube content exists
Revenue model	Online course sales ($49-$499), certification programs ($199-$999), B2B corporate training ($2K-$10K per engagement)
Core moat	First-mover advantage. The first official GEO certification could become the de facto standard
Entry barrier	Low (content creation ability + domain expertise)
Time to revenue	3-6 months (content production > platform listing > sales)
Risk	No established GEO standards mean short curriculum shelf life. Quarterly updates needed

Opportunity and Constraints

The biggest opportunity in GEO education is the void itself. As of March 2026, searching for “GEO Certification” returns zero accredited programs. Whoever creates one before HubSpot launches an SEO-style certification gets to define the market.

But the constraints are equally clear:

Constraint	Explanation
No standards	GEO measurement methodology is still in the academic stage. Education must be designed without a definitive “correct answer”
Content obsolescence	AI engine response algorithms change frequently. Training content from six months ago may already be inaccurate
Authority challenge	SEO education had Google as a single standard. GEO has multiple AI engines operating differently — whose criteria constitute the “answer” is unclear
Demand uncertainty	Whether enough people want to learn GEO to constitute a viable market is unvalidated

Capturing the education market early is attractive, but running an education business in a field where “what to teach” changes quarterly means high content maintenance costs.

Precedent from SEO Education

SEO education has historically updated its curriculum with each Google algorithm change — Penguin, Panda, BERT. Programs that couldn’t keep pace were culled.

GEO education faces an even more complex situation. SEO had one standard: Google. GEO has ChatGPT, Gemini, Perplexity, Claude, and AI Overviews, each using different response generation logic. A single engine update can invalidate parts of the curriculum. Tracking multiple engines simultaneously could multiply maintenance burden several times over compared to SEO education.

Still, when the education market forms, its ripple effects are substantial. More marketers trained in GEO means more demand for GEO SaaS tools. The education and SaaS markets are complementary: education creates demand, SaaS fulfills it.

4. AI Response Data Infrastructure

Like Moz and Ahrefs providing web crawling data as infrastructure for SEO, this is the business of systematically collecting, refining, and distributing AI engine responses. It occupies the very bottom layer of the GEO value chain.

Market Structure

Item	Detail
Market size estimate	SEO data infrastructure market (Ahrefs, Moz, Majestic, etc.) ~$2B. GEO data infrastructure potential: $500M-$1.5B (2028-2030)
Current competition	Exa ($85M cumulative). No other dedicated infrastructure companies confirmed
Revenue model	API-based billing (per call or data volume). Enterprise annual licenses
Core moat	Data accumulation, crawling infrastructure scale, API reliability
Entry barrier	High (large-scale crawling infrastructure, multi-engine API costs, data refinement pipeline)
Time to revenue	12-24 months (infrastructure build > data accumulation > API release > customer acquisition)
Risk	Initial infrastructure investment $1M+ required. Vulnerable to AI engine ToS changes. Crawling may be blocked

Entry Barrier Analysis

Data infrastructure has the highest entry barrier among the six opportunities. Breaking down the required components:

Component	Requirement	Estimated Annual Cost
Multi-AI engine API calls	High-volume calls to GPT, Gemini, Perplexity, Claude, etc.	$100K-$500K+
Crawling infrastructure	Distributed crawlers, proxy infrastructure, scheduling systems	$50K-$200K
Data refinement pipeline	NLP processing, entity extraction, citation source mapping	$50K-$150K (including labor)
Storage and serving	Large-scale time-series data storage, low-latency API serving	$30K-$100K
Legal/compliance	AI engine ToS review, data licensing	$20K-$50K

Total annual operating costs run $250K-$1M, with initial build-out requiring a minimum of 6-12 months. Bootstrapping teams entering this space face a realistic impossibility. That’s why Exa raised $85M.

Data infrastructure is the most technically attractive opportunity with the strongest defensibility, but it follows a “money makes money” structure. Without starting capital, you cannot begin.

Growth Path Reference from SEO Data Infrastructure

Moz was founded in 2004, Ahrefs in 2011, Semrush in 2008. All three built web crawling data as their core asset and layered analytics tools on top. Each took 5-7 years to accumulate meaningful data, operating at a loss or low profitability during that period.

GEO data infrastructure will likely follow a similar trajectory. But there is a key difference. SEO crawling targeted the public web, making technical and legal barriers relatively low. GEO data collection requires calling AI engine APIs, creating direct exposure to engine operator policy changes. If OpenAI raises API prices or Perplexity restricts high-volume calls, data collection costs spike immediately. This dependency risk is a new variable that did not exist in SEO infrastructure.

5. Vertical-Specialized GEO

GEO solutions optimized for specific industry domains. While general-purpose GEO tools address “AI visibility across all industries,” vertical GEO goes deep into the unique AI response patterns of a particular industry.

Market Structure

Item	Detail
Market size estimate	Varies by vertical. E-commerce GEO ~$200M, healthcare GEO ~$100M, legal GEO ~$50M (2028 estimates)
Current competition	E-commerce: Bluefish ($24M). Healthcare, legal, finance, travel = no dedicated GEO solutions
Revenue model	SaaS or consulting. Industry-specific premium pricing (1.5-3x of general-purpose)
Core moat	Industry-specific AI response pattern data, domain expertise, regulatory understanding
Entry barrier	Medium (requires both GEO technical capability and industry domain expertise)
Time to revenue	6-12 months (domain-specific MVP > industry early adopters > paid conversion)
Risk	TAM is narrow. General-purpose platforms may add vertical features

Industry-Specific AI Response Pattern Differences

The rationale for vertical GEO rests on the structural differences in how AI generates responses across industries. Bagga et al.’s (2025) E-GEO framework empirically demonstrated AI response specificity in e-commerce, and the principle applies across other verticals.

Industry	AI Response Characteristics	Primary Citation Sources	GEO Optimization Direction
E-commerce	Product comparisons, pricing, reviews-centric	Review sites, comparison platforms, official sites	Product spec structuring, review data optimization
Healthcare	Academic evidence emphasis, conservative language	PubMed, Mayo Clinic, WebMD	Academic citation acquisition, authoritative source linkage
Legal	Case law citations, jurisdictional distinctions	Legal databases, government sites	Case-based content, jurisdiction-specific optimization
Finance	Regulatory disclaimers, past performance caveats	Financial news, regulatory agency sites	Compliance-ready content, performance data structuring
Travel	Location data, seasonality, real-time pricing	Travel reviews, booking platforms	Location entity optimization, seasonal content

General-purpose GEO tools attempt to handle these differences within “one dashboard.” But the optimization strategy for “how AI cites academic papers” in healthcare is fundamentally different from “how AI cites reviews” in e-commerce. This difference is the raison d’etre of vertical specialization.

Bagga et al.’s (2025) E-GEO is the first framework to separately define AI response product recommendation patterns, citation structures, and visibility metrics for e-commerce. This approach is replicable across other verticals, suggesting that “industry-specific GEO frameworks” may become standardized over time.

The Strategic Dilemma of Vertical Specialization

Vertical specialization carries an inherent dilemma: the deeper you go, the smaller the TAM becomes. “Healthcare GEO” is inevitably smaller than “general-purpose GEO.” But the premium pricing and higher customer retention that come with depth can compensate for the smaller addressable market.

Bluefish securing $24M in funding for e-commerce GEO specialization is evidence that VCs validate the vertical strategy. E-commerce was a natural first vertical since it’s the industry most directly affected by AI search. Healthcare, legal, and finance verticals are more regulatory-complex and have higher accuracy demands for AI responses. This raises entry barriers but also creates stronger moats for those who succeed.

Whether to start with a general-purpose platform and expand into the highest-demand vertical based on data, or to go all-in on a single vertical from the start, remains an open question. For small teams, starting general and letting early customer data reveal naturally concentrating verticals carries lower risk.

6. GEO + PR/Communications Fusion

A business combining PR and media exposure with GEO, leveraging the tendency of AI search engines to prefer third-party sources (news, reviews, wikis, forums) as trusted citation sources.

Market Structure

Item	Detail
Market size estimate	Global digital PR market ~$20B. GEO fusion segment $500M-$1B (2028 estimate)
Current competition	Genevate, Intero Digital, and other early players. PR agency GEO awareness is still low
Revenue model	PR agency model (monthly retainer $3K-$15K) + GEO performance measurement reports
Core moat	Media network, PR-GEO linkage data, citation causation analysis capability
Entry barrier	Medium (requires both PR network and GEO technical understanding)
Time to revenue	3-6 months (if existing PR capabilities are in place) / 12+ months (for new entrants)
Risk	Proving PR-to-GEO citation causation is difficult. “This article led to this AI citation” is hard to establish

Citation Preference Structure

Chen et al.’s (2025) research found that AI search engines tend to prefer third-party media (news, reviews, forums) over brand-owned channels (official websites, corporate blogs) as citation sources. This provides the logical basis for combining PR and GEO.

Source Type	AI Citation Preference	SEO Effect	GEO Effect
Official website	Medium	High (direct ranking)	Medium (cited but not preferred)
News/media articles	High	Medium (backlinks)	High (cited as trusted source)
Review platforms	High	Medium	High (especially e-commerce)
Academic papers/reports	High	Low	High (especially healthcare/tech)
Wikipedia	Very high	Low (nofollow)	Very high
Social media	Low	Low	Low to medium (varies by platform)

This structure redefines PR’s value in GEO terms. In SEO, PR’s primary value was “acquiring backlinks.” In GEO, PR’s value becomes “getting the brand mentioned in sources that AI trusts.” A shift in perspective is required.

However, the biggest challenge right now is attribution. A methodology for quantitatively proving that “this media placement led to this AI citation” has not been established. Correlations can be shown, but causation is hard to prove. Until this limitation is resolved, GEO + PR businesses struggle to move beyond “it seems to work.”

Practical Execution Structure for GEO + PR

Despite attribution challenges, GEO + PR operates as follows: first, monitor which sources AI search engines cite for a specific brand. When cited sources are insufficient or negative, produce or generate exposure for media articles, expert reviews, and industry reports that contain positive brand mentions. After a period, measure changes in AI response citations.

In this process, temporal sequence — “AI citations increased after media article placement” — can be demonstrated. But causation — “AI citations increased because of the media article” — is hard to prove. Other factors (competitor negative news, AI engine algorithm changes) could have been at play simultaneously. This attribution problem was never fully solved in traditional PR either, and it becomes more complex in GEO.

GEO + PR has long-term growth potential, but as of 2026, the measurement infrastructure is insufficient for it to stand alone as an independent business. Once GEO SaaS tools mature enough to automatically track correlations between PR activities and AI citation changes, this segment’s business model will become clearer.

Comprehensive Comparison of Six Opportunities

Quantitative Comparison

Opportunity	Market Size Estimate (2028)	Entry Barrier	Scalability	Defensibility	Time to Revenue	Measurability
SaaS platform	$350M-$1.05B	Medium	High	Data moat	6-12 months	High
Agency/consulting	$800M-$2.4B	Low	Low	Case studies	1-3 months	Medium
Education/certification	Difficult to estimate	Low	Medium	First-mover advantage	3-6 months	Low
Data infrastructure	$500M-$1.5B	High	High	Infrastructure scale	12-24 months	High
Vertical specialization	$50M-$200M (per vertical)	Medium	Medium	Domain expertise	6-12 months	Medium
GEO + PR	$500M-$1B	Medium	Medium	Media network	3-12 months	Low

Fit by Team Type

Not every opportunity suits every team. The right opportunity depends on team size, capital, technical capability, and domain expertise.

Team Type	Best Fit	Second Best	Poor Fit
Bootstrapped 1-3 person technical team	SaaS (niche market)	Vertical specialization	Data infrastructure
VC-funded tech startup	Data infrastructure, SaaS (general)	Vertical specialization	Agency
SEO agency pivoting	Agency (add GEO services)	GEO + PR	SaaS (if technical capacity is lacking)
Marketing/PR professionals	GEO + PR	Agency	Data infrastructure
Education/content specialists	Education/certification	Agency	Data infrastructure
Industry domain experts	Vertical specialization	Agency	Data infrastructure

WICHI’s Choice: SaaS + Asian Market

WICHI selected opportunity #1: the SaaS platform. The target market is Asia, starting with Korea. This section documents “why SaaS” and “why Asia” separately.

Decision Structure

flowchart TD
    A["6 GEO Business Opportunities"] --> B{"Team: small<br/>Capital: bootstrapped"}

    B --> C{"Entry Barrier Filter<br/>High = exclude"}
    C -->|Excluded| D["Data Infrastructure<br/>Initial capital $1M+"]

    C -->|Pass| E{"Scalability Filter<br/>Low = secondary channel"}
    E -->|Secondary| F["Agency<br/>Secondary channel"]
    E -->|Secondary| G["Education/Certification<br/>Secondary channel"]

    E -->|Pass| H{"Measurability Filter"}
    H -->|Low| I["GEO + PR<br/>Attribution difficulty"]

    H -->|Pass| J{"Market Differentiation Filter"}
    J --> K["SaaS Platform<br/>Local market gap"]
    J --> L["Vertical Specialization<br/>Limited TAM"]

    K --> M["Final Choice:<br/>SaaS + Asian Market"]

    style M fill:#2563eb,color:#fff,stroke-width:2px
    style D fill:#fee2e2,stroke:#ef4444
    style I fill:#fee2e2,stroke:#ef4444
    style F fill:#fef3c7,stroke:#f59e0b
    style G fill:#fef3c7,stroke:#f59e0b
    style L fill:#fef3c7,stroke:#f59e0b

Why SaaS: Three Criteria

The decision applied three criteria: measurability, scalability, and defensibility.

Criterion 1: Measurability

SaaS is built around a quantitative metric — the GEO Score. It can show customers, right on a dashboard, “this number went up or down compared to last month.”

Agencies must argue through reports that “our strategy produced these results.” Education struggles even harder to prove “completing this course improves GEO competency.” GEO + PR faces the most difficult attribution challenge: proving “this PR activity led to this AI citation.”

Model	What’s measurable	What’s hard to prove
SaaS	GEO Score changes, citation frequency, SOV	Score-to-revenue causation
Agency	Volume of work executed	Work-to-outcome causation
Education	Completion numbers, satisfaction scores	Education-to-skill improvement
GEO + PR	Media exposure volume	PR-to-AI citation causation

Measurable services have two structural advantages. First, retention is higher. Canceling a service where “the numbers are visible” carries greater psychological cost than canceling one where they aren’t. Second, upselling is natural. “Want to track more keywords? Upgrade to the next plan” is a seamless conversion path.

Criterion 2: Scalability

Model	Resources Needed When Clients 10x	Marginal Cost Structure
SaaS	Incremental server costs	Low (infrastructure scaling only)
Agency	~10x headcount	High (proportional to labor)
Education	Content updates + platform maintenance	Medium (partial economies of scale)
Data infrastructure	Infrastructure expansion	Low to medium

After infrastructure is built, the cost of adding one more SaaS user is limited to incremental server expenses. The service delivered to the 100th customer and the 1,000th customer is identical. Agencies must hire a separate consultant for the 100th client.

For bootstrapping teams, scalability is the question of “can revenue eventually outpace headcount?” Agency margins plateau. SaaS margins improve dramatically once fixed costs are exceeded.

Criterion 3: Defensibility

SaaS defensibility comes from data accumulation. As users engage with the platform, query-level AI response data, time-series visibility changes, and competitor benchmark data accumulate. This data appreciates in value over time, and late entrants need the same elapsed time to match the dataset.

Local AI engine (Naver AI, Clova X, etc.) response data is a category that global platforms are not systematically collecting. Accumulating this data first creates a gap that global platforms cannot quickly close even upon entering the Korean market.

flowchart LR
    A["User growth"] --> B["Query/response data accumulation"]
    B --> C["Analysis accuracy improves"]
    C --> D["User satisfaction increases"]
    D --> E["Retention + new user acquisition"]
    E --> A

    B --> F["Time-series data depth"]
    F --> G["Data gap vs. late entrants"]
    G --> H["Defensive moat strengthens"]

    style A fill:#dbeafe,stroke:#2563eb
    style H fill:#dcfce7,stroke:#16a34a

Why Asia: Market Gap Analysis

After choosing SaaS, the next question is “which market to target?” WICHI is starting with Korea and planning expansion to Japan and Southeast Asia.

Regional Competition Density for Global GEO SaaS

Region	Confirmed GEO SaaS Count	Major Players	Competition Density
United States	8+	Profound, Exa, Bluefish, Scrunch, Gumshoe	Very high
Europe	3+	Peec AI (Germany), Otterly, etc.	High
Korea	0 (dedicated platforms)	-	Vacant
Japan	0	-	Vacant
China	Undisclosed (hard to confirm)	-	Unknown
Southeast Asia	0	-	Vacant

The US and European markets already have $170M+ in VC capital deployed. A bootstrapping team entering these markets to compete head-to-head against Profound or Peec AI is unrealistic — outmatched in capital, talent, and brand recognition on every dimension.

The Asian market, particularly Korea and Japan, differs structurally in three ways.

Difference 1: Local AI Search Engines Exist

In the US, AI search is dominated by ChatGPT, Perplexity, and Google AI Overviews. In Korea, Naver AI Search and Clova X hold independent positions alongside these global engines. Japan has a similar structure.

These local engines exhibit response generation patterns distinct from their English-language counterparts. A Naver AI response to a Korean-language query and a ChatGPT response to the same query in English differ in citation source composition, response structure, and entity recognition methods. Global GEO platforms are not designed to track these differences.

Comparison	Global AI Engines (GPT, Gemini)	Korean Local AI Engines (Naver AI, Clova X)
Primary citation sources	English web, Wikipedia, Reddit	Naver blogs, Knowledge iN, Korean news
Response language	English (mixes English sources even for Korean queries)	Korean sources prioritized
Entity recognition	English-based entities prioritized	Korean entities (e.g., brand names in Hangul)
Citation structure	Inline citations + bottom source list	Varies by engine (not standardized)
Data accessibility	APIs available (paid)	Limited or private

Difference 2: Language Barriers Become Entry Barriers

For global GEO platforms to enter Korea and Japan, they must solve three problems simultaneously:

1. Korean/Japanese AI response NLP processing: Morphological analysis, entity extraction, and sentiment analysis require language models different from English
2. Local AI engine API integration: Accessing Naver AI, Clova X, etc. involves authentication and billing systems different from global services
3. Local market understanding: Knowing which industries in Korea feel the need for GEO, and at what price point they buy, requires on-the-ground market knowledge

Solving all three simultaneously is not a priority for global platforms. The English-language market provides more than sufficient TAM. There may be a 12-24 month gap between Profound adding Korean language support to its roadmap and actually shipping a product optimized for the Korean market. That time gap is the local player’s opportunity.

This pattern is repeatedly observed in SaaS history. Even after Salesforce dominated US CRM, Sansan built an independent market in Japan around business card-based CRM. Even as Semrush dominates global SEO tools, local tools like AceCounter and Naver Search Advisor play unique roles in Korea. Global SaaS localization requires deep understanding of local platform ecosystems, user behavior patterns, and regulatory environments — not just translation. This is a domain where local teams have a structural advantage.

Difference 3: Lower Competition Density Than the US/Europe

Metric	US GEO SaaS Market	Korean GEO SaaS Market
Dedicated players	8+	0
VC capital deployed	$170M+	$0
Agencies offering GEO services	Dozens	Fewer than a handful
GEO-related content (blogs/media)	Abundant	Minimal
Customer GEO awareness	Early adopter stage	Initial awareness stage

Low competition density in Korea is a double-edged sword. The upside: market preemption is possible. The downside: the market itself may not yet have formed. “No competitors” could mean “no market.” This risk is acknowledged as part of the decision.

That said, the structural conditions for GEO demand in Korea are taking shape. First, as Naver introduces AI search features, Korean companies that invested in Naver SEO are starting to care about “visibility in AI search.” Second, Korean users adopting global AI search (ChatGPT, Perplexity) are growing rapidly, raising corporate interest in how Korean-language content is treated in global AI search. Third, SEO agencies are adding GEO to their service menus but lack measurement tools to deliver meaningful services. This “tool gap” could be a direct demand driver for SaaS.

“A market with no competitors” carries two meanings: either it is an undiscovered opportunity, or it is a market nobody wants. The only way to distinguish the two is to enter and find out.

What WICHI Chose Not to Do

Documenting what was not chosen, and why, is as important as documenting what was chosen.

Opportunity	Reason Not Chosen	Future Possibility
Agency/consulting	Labor-proportional scaling. Hard to defend	Can serve as SaaS secondary channel
Education/certification	No standards makes curriculum shelf life short	SaaS data can generate derivative educational content
Data infrastructure	$1M+ initial capital required. Incompatible with bootstrapping	Long-term, SaaS data could transition into infrastructure value
Vertical specialization	Limited TAM. Starting general and adding vertical features later is more rational	Can pivot to a vertical if specific industry customers concentrate
GEO + PR	Attribution difficulty. A tech team has no comparative advantage in PR networks	Can connect via partnerships

Agency and education are not “things to avoid” — they are “things that are not the main focus.” Data and insights accumulated from running the SaaS platform become raw materials for consulting and educational content. But the business focus stays on the platform. Even if agency and education contribute revenue, the core metric is the platform’s MRR (Monthly Recurring Revenue).

flowchart TD
    A["SaaS Platform<br/>(Core Business)"] --> B["AI Response Data Accumulation"]
    B --> C["Analysis Insights Generated"]

    C --> D["Consulting Services<br/>(Secondary Channel 1)"]
    C --> E["Educational Content<br/>(Secondary Channel 2)"]
    C --> F["Industry Reports<br/>(Secondary Channel 3)"]

    D --> G["Case studies > SaaS marketing asset"]
    E --> H["Trainees > potential SaaS customers"]
    F --> I["Report downloads > lead generation"]

    G --> A
    H --> A
    I --> A

    style A fill:#2563eb,color:#fff,stroke-width:2px
    style D fill:#fef3c7,stroke:#f59e0b
    style E fill:#fef3c7,stroke:#f59e0b
    style F fill:#fef3c7,stroke:#f59e0b

Risks and Prerequisites

For this choice to remain valid, several prerequisites must hold. Contingency plans for each are documented alongside.

Prerequisite	If Not Met	Contingency
Korean enterprise GEO demand materializes in 2026-2027	Market stays in SEO and does not transition to GEO	Pivot to English-speaking SMB market. Maintain local specialization but change target customer segment
Global platforms delay Korean language support by 12+ months	Profound or Peec rapidly add Korean	Reinforce local AI engine (Naver AI, Clova X) coverage as differentiator
AI engine response structures remain broadly similar to today	AI engines switch to citation-free responses or undergo fundamental structural changes	Pivot required. In this scenario, the entire GEO market gets redefined
Bootstrapping can reach PMF (Product-Market Fit)	Capital constraints delay product development	Explore angel/seed funding or supplement with agency revenue

The greatest risk is the third: AI engine response structures changing fundamentally. This is not WICHI-specific but a risk shared by the entire GEO market — including the $174.5M-funded companies. Since it is uncontrollable, the response is monitoring rather than hedging.

Categorizing risks by type:

Risk Type	Controllability	Response	Monitoring Frequency
Market risk (demand doesn’t materialize)	Partially controllable (can stimulate demand through education/content)	Secure early adopters, then expand	Monthly
Competitive risk (global platform entry)	Uncontrollable	Strengthen local data moat	Quarterly
Technical risk (AI engine structure changes)	Uncontrollable	Monitor + rapid pivot	Weekly
Capital risk (development delays)	Partially controllable	Agency revenue supplement or funding	Monthly
Regulatory risk (data collection restrictions)	Uncontrollable	Operate within legal bounds + legal review	Quarterly

Among these, technical risk scenarios warrant closer examination. AI engine response structures could change in several directions. First, AI engines could eliminate source citations entirely. This would remove GEO’s core measurement target and redefine the entire market. However, the current trend runs counter to this — legal and ethical pressure is pushing AI engines toward more citation transparency, not less. Second, AI engines could introduce proprietary “brand profile” systems, in which case GEO would shift to “optimizing AI engine brand profiles.” Third, AI search could consolidate around a single dominant engine. Currently multiple engines compete, but if one dominates, GEO complexity decreases while dependency on that engine increases.

Regardless of which scenario materializes, the core value proposition of GEO SaaS — “measuring and managing brand visibility in AI search” — is likely to persist. Implementation methods may change, but the underlying problem does not disappear.

Summary

The GEO market as of March 2026 is at an early stage comparable to SEO circa 2005-2008. Six business opportunities are identifiable, each with distinct market structures, entry barriers, and revenue models. Which one is the right answer — nobody knows yet.

WICHI applied three criteria — measurability, scalability, and data defensibility — to select SaaS platform + Asian market. This choice is grounded in a structural judgment: “a bootstrapped small team can avoid direct competition with globally VC-funded companies while leveraging local data as a moat.”

Agency and education serve as secondary channels, while the core focus remains on platform MRR. Data infrastructure may eventually derive value from SaaS data, but is not realistic at the initial stage.

Whether this choice is correct is something only the market can answer. This post records “why this direction was chosen” and documents contingency paths for each scenario where the underlying assumptions prove wrong. Strategy is not something you set once and forget — it is something you refine by periodically validating whether its premises still hold.

Finally, what this analysis intentionally does not cover should be stated. WICHI’s specific technology stack, cost structure, pricing policy, and internal metrics are outside the scope of this post. This post addresses only the strategic framework: “what opportunities exist in the market, and why this direction was chosen.” Implementation details will be covered in separate build logs.