SEO

How to Optimize for GenAI Answers

ChatGPT is taking the world by storm. It reported in November 2024 100 million weekly users, despite being only one of the popular generative AI platforms.

No business should ignore those channels, as consumers increasingly turn to genAI for product and brand recommendations.

Yet showing up in AI answers is tricky, and the tactics vary among platforms.

In August 2024, Seer Interactive, a marketing agency, compared the leading “answer engines.”

Answer Engines: AI vs SEO(Seer Interactive) Generative Engine Optimization (GEO) vs SEO AI Models: • Claude, Llama (Training Data) • Perplexity, Google AIO (Search Data) • ChatGPT, Gemini (Hybrid: Training + Search Data) • SEO: Google, Bing, Yahoo How They Generate Results: • Claude, Llama: LLM interprets query and serves information from training data. • Perplexity, Google AIO: LLM interprets query and serves information primarily from web index. • ChatGPT, Gemini: LLM routes response via training data or web index based on query. • SEO: Index & Retrieval: Crawl & Index. How They Serve Results: • Claude, Llama: Primarily text • Perplexity, Google AIO: Text & citation links • ChatGPT, Gemini: Text & citation links • SEO: Search engine serves most relevant indexed webpages (10 blue links, SERP features, Ads). Ability to Influence: • Claude, Llama: Low • Perplexity, Google AIO: Medium • ChatGPT, Gemini: Medium • SEO: High Speed to Influence: • Claude, Llama: Slow • Perplexity, Google AIO: Fast • ChatGPT, Gemini: Medium • SEO: Fast Mechanisms of Influence: • Claude, Llama: Brand marketing, Earned media • Perplexity, Google AIO: Website content, Earned media, Organic social • ChatGPT, Gemini: Content, Brand, Earned media, Social • SEO: Content, Brand, Earned media

“Answer Engines: AI vs. SEO” from Seer Interactive. Click image to enlarge.

Here’s my version.

StructureOrganic SearchChatGPT, Gemini, CopilotPerplexity, Google AI OverviewsClaude, DeepSeek, Llama
Knowledge sourcesSearch index, knowledge graphTraining data + search data + memorySearch indexTraining data
OutputCitations, ads, search featuresAnswers + citationsAnswers + citationsAnswers (few links)
Optimization tacticsContent, backlinks, brandingContent, backlinks, brandingGoogle & Bing indexes + top ranking Branding

Foundational search optimization tactics apply to genAI. Sites should be indexed and visible in Google and Bing to appear in answers on the leading platforms — ChatGPT, Gemini, Google’s AI Overviews, Microsoft Copilot, and Perplexity.

However, some AI-optimizaiton tactics are more important than others in my experience.

Fast, Light, Simple

AI crawlers that access a site will learn about the business and its purpose but may not link to it. Generative AI platforms often repurpose content without referencing the source. Antropic’s Claude, Meta’s Llama, and now DeepSeek rarely include links.

Thus allowing those AI crawlers on a site is debatable. My advice to clients is this: Google has monetized our content for years, but we’ve all benefitted from the visibility. So I usually suggest optimizing for AI platforms rather than blocking them.

The best AI-optimized sites are fast, light, and usable with JavaScript disabled. AI crawlers are immature, more or less. Most cannot render JavaScript and abort crawling slow-loading sites.

No Fluff

For years, Google’s machine learning favored featured snippets from pages with clear, concise, factual answers — even when the page itself wasn’t ranking organically near the top.

Recent case studies prove the point. One comes from search optimizer Matt Diggity, who shared examples on X of the ranking benefits in Google from brevity and clarity.

Matt Diggity✔ @mattdiggityseoQuick story about how I cracked Google's code with NLP... Ran tests on two identical articles. One written normally, one written in NLP-friendly format. The difference? • Position #1 vs Position #12 on Google. • Here's what Google's algorithm loves: • Clear subject-verb-object structure • Zero fluff words • Direct statements • Simple language • Example: "HTML is the core language for creating web pages." What it hates: • Complex sentence structures • X Ambiguous phrases • X Fancy marketing speak • X Abstract concepts •Example: "Revolutionary HTML empowers stunning webpage experiences."

Search optimizer Matt Diggity posted on X the results from his “natural language processing” text. Click image to enlarge.

Matt’s findings apply to all writing, including generative AI platforms.

In short, AI optimization aligns with commonsense organic search tactics. Optimizing for one will likely help the other.

Ann Smarty
Ann Smarty
Bio