GEO is necessary. It is also uncertain, difficult to measure, and carries risks that the SEO industry spent twenty years learning to manage.
This article is not about discouraging GEO investment. It is about investing with open eyes.
1. The Attribution Black Box
SEO had imperfect attribution, but it had attribution. You could track rankings, impressions, clicks, and conversions. The path was visible.
GEO attribution is a black box. When someone asks ChatGPT a question and you are cited, there is no click to track. You may never know the citation happened.
When that person later searches your brand and converts, your attribution system shows branded search. The actual influence, the AI citation, is invisible.
This creates measurement challenges that SEO professionals are not accustomed to. You cannot optimize what you cannot measure. And with GEO, you often cannot measure what matters most.
Current workarounds include manual citation audits (asking AI systems questions and documenting responses), brand search monitoring (using brand search volume as a proxy), third-party share of voice tools (emerging but immature), and user surveys (asking customers how they found you).
None of these are as reliable as SEO metrics were. Accept measurement uncertainty as a cost of entry.
2. The Probabilistic Nature of AI Citation
SEO rankings were relatively stable. If you ranked third yesterday, you probably rank third today.
AI citation is probabilistic. The same question asked twice may produce different responses. Your content may be cited in one response and absent in another.
This is not a bug in AI systems. It reflects how language models work: sampling from probability distributions, incorporating randomness to avoid repetitive outputs.
What this means for GEO: you cannot “rank” in a stable sense. You can increase citation probability, but you cannot guarantee citation.
Success metrics shift from position tracking to probability estimation. Are you cited more often than last month? More often than competitors? These questions require statistically significant sampling, not single-query checks.
3. The Model Collapse Risk
Research by Shumailov et al. in Nature documented model collapse: when AI systems train on AI-generated content, each generation produces lower quality output.
If everyone optimizes for AI citation by producing AI-generated content designed to be cited by AI, the entire system degrades.
This creates a tragedy of the commons. Individual actors have incentive to produce synthetic content. Collective action produces systemic degradation.
The risk for GEO practitioners: if AI systems implement filters to reduce synthetic content influence, GEO strategies based on content volume rather than genuine information value will fail.
The hedge: invest in content that contains genuinely original information: proprietary data, expert perspective, primary research. This is harder and more expensive. It is also more defensible.
4. The Platform Instability Problem
SEO practitioners learned to fear algorithm updates. GEO practitioners must fear model updates.
When Google updated its algorithm, years of optimization work could be invalidated overnight. The same will happen with AI model updates.
OpenAI, Google, Anthropic, and other AI providers regularly update their models. Each update changes how information is retrieved, synthesized, and cited. Content that was cited frequently before an update may be invisible after.
Unlike SEO, where some signals remained stable (quality content, authoritative links), AI model updates can change fundamental retrieval mechanisms. What worked is no guarantee of what will work.
Strategic response: avoid over-optimization for any specific model’s current behavior. Focus on genuine information value that should translate across model iterations.
5. The Trust Erosion Concern
Ofcom research shows only 18% of users trust AI search results. This is surprisingly low and represents both a risk and an opportunity.
The risk: if trust does not improve, AI-mediated discovery may not become dominant as quickly as projected.
The opportunity: the trust gap creates space for trusted entities. If AI citation comes from recognized, trusted sources, it carries more weight than citation from unknown sources.
Trust erosion also affects GEO strategy. If users do not trust AI responses, being cited in those responses has less value. The citation needs to be verifiable for users to act on it.
This argues for visible entity association in cited content. Users who want to verify AI claims can find your organization, evaluate your credibility, and build trust directly.
6. The Economic Sustainability Question
AI queries cost significantly more than traditional search queries. Current estimates suggest 100x more energy consumption per query.
AI companies are currently subsidizing this cost to drive adoption. That subsidy is not sustainable indefinitely.
When economic pressure increases, AI companies will need to reduce costs or increase revenue extraction. This likely means more advertising in AI responses, more aggressive monetization of citation placement, or reduced AI feature availability.
How this affects GEO is uncertain. Scenarios include pay-for-citation models emerging, AI responses becoming more ad-heavy and less trustworthy, reduced AI feature deployment reducing GEO value, and premium AI services becoming the primary GEO target.
Plan for uncertainty. GEO strategy should not assume current AI economics are permanent.
7. The Ethical Considerations
GEO raises ethical questions that SEO largely avoided.
Content origin: If your content trains AI systems, should you be compensated? If AI cites you without clear attribution, is that fair?
Information quality: Optimizing for AI citation without ensuring accuracy contributes to potential misinformation spread.
Competition dynamics: GEO may favor established entities with resources to invest, creating barriers for new entrants.
Transparency: Users may not realize AI responses are synthesized from sources with GEO agendas.
These questions do not have clear answers. They are worth considering as GEO practices develop.
8. Operating Principles for the Post-SEO Era
Given these risks and uncertainties, what principles should guide GEO investment?
Principle 1: Invest in genuine information value, not synthetic optimization.
Principle 2: Maintain measurement humility. Accept that GEO outcomes are harder to track than SEO outcomes.
Principle 3: Diversify across platforms. Do not bet everything on a single AI system’s current behavior.
Principle 4: Build owned audience as the ultimate hedge. Direct relationships do not depend on any platform.
Principle 5: Establish entity identity. Without recognizable source identity, citation has limited value.
Principle 6: Plan for instability. Model updates, platform changes, and economic shifts will happen. Build flexibility.
Principle 7: Maintain SEO for transactional queries. The traffic economy is declining, not dead.
Principle 8: Invest in learning. GEO best practices are still forming. Organizations that learn fastest will have advantages.
The Real Conclusion
GEO is necessary and uncertain. Both of these things are true.
The organizations that succeed will be those that invest in GEO while managing its risks, that pursue citation while building owned audience, that optimize for current systems while maintaining flexibility for future changes.
The worst strategy is pretending nothing has changed and hoping SEO continues to work. The second worst strategy is abandoning SEO entirely for an unproven GEO playbook.
The right strategy is sophisticated: understand the transition, invest appropriately across both disciplines, and build the entity and audience foundations that provide resilience regardless of how platforms evolve.
The future of search is being written now. Your job is to be part of writing it, not just being written about.
Sources:
- Shumailov et al.: “The Curse of Recursion” Nature (2024)
- Ofcom: AI search trust research
- AI energy consumption: IEA and academic studies
- AI economics analysis: Industry reports on model training and inference costs
- Platform economics: Various academic and industry research