Guides
Structured guides for implementing GEO, AEO, and AI SEO workflows. Written for founders, growth teams, and technical operators.
SaaS products live or die by how well they are represented when potential customers ask AI assistants for recommendations. The goal is not just to be mentioned — it is to be described accurately, positioned correctly, and cited with enough detail that the person reading the AI's answer has genuine reason to click through.
Your homepage, product page, and key feature pages should answer three questions in plain text: what does your product do, who is it for, and how is it different from alternatives. AI systems synthesize these answers when users ask "what is the best tool for X." If your page cannot answer those questions in the first 200 words of parseable HTML, you are being misrepresented.
Users frequently ask AI systems to compare products. Write a dedicated page or section that clearly articulates how you differ from the top 3–5 alternatives in your category. Use direct language: "Unlike [Competitor], [Product] does X." This gives AI systems extractable comparison context that feeds directly into generated answers.
AI systems are particularly good at matching intent to product. If someone asks "what tool should I use for [specific use case]," you want content that directly describes your product in the context of that use case. Map your top 5–10 customer use cases to dedicated content sections or pages with clear, parseable descriptions.
Use JSON-LD SoftwareApplication schema on your product and pricing pages. Include name, description, applicationCategory, operatingSystem, and offers. This gives AI systems structured entity data that reduces hallucination and improves citation accuracy.
Ecommerce AI visibility is primarily about product discoverability. When users ask AI systems "what is the best [product category] under $X" or "where can I buy [specific item]," you want your catalog to be in the answer. The challenge is that most ecommerce sites rely heavily on JavaScript rendering — making their product data invisible to AI crawlers by default.
Product names, descriptions, prices, specifications, and availability should be in the HTML response — not loaded via client-side JavaScript. If your product data is fetched client-side, AI crawlers receive an empty shell. Use server-side rendering or static generation for product pages, or use adaptive rendering to serve a structured version to crawlers.
Add JSON-LD Product schema to every product page. Include name, description, image, brand, sku, offers (price, priceCurrency, availability), and aggregateRating if you have reviews. This structured data is directly consumed by AI systems and significantly improves citation accuracy and completeness.
Category pages are often the target of comparative AI queries ("what are the best running shoes for flat feet"). Your category page should describe the category clearly, list key differentiators of the products within it, and include structured content that explains how to choose within the category. Thin category pages with only a product grid are invisible to AI.
When AI systems answer product questions, they pull specific attributes: dimensions, materials, compatibility, included accessories. Make sure these are in plain text on your product pages — not only in images or interactive selectors. Tables with specifications are particularly well-parsed by AI crawlers.
Use this checklist before launching a new page or auditing your existing ones. Each item represents a signal that AI crawlers use to parse and understand your content.
Perplexity is unique among AI assistants: it performs real-time web searches and cites sources inline. When a user asks Perplexity a question, it fetches your page at that moment and decides whether your content is worth citing. This creates a different optimization target than ChatGPT or Claude, which use pre-crawled knowledge bases.
Perplexity looks for quotable, self-contained statements it can use as citations. Start each major section with a direct, complete answer to the implied question — then expand on it below. "Appear works by routing DNS traffic through a Cloudflare Worker that classifies each request and serves the appropriate response" is more citable than a paragraph that gradually builds to the same point.
Perplexity fetches pages in real time. If your page takes 4+ seconds to return content, it may be skipped. Prioritize fast Time to First Byte (TTFB) on pages you want cited. Server-side rendered or static pages with content in the initial response are significantly more reliable for Perplexity citation.
Perplexity cares about recency. Include publication or "last updated" dates on your content pages. Use article schema with datePublished and dateModified. Keep your most important pages updated — Perplexity is more likely to cite fresh content over stale pages that rank well in traditional search.
Perplexity attributes citations to page titles and domains. Make sure your page titles are descriptive and specific — "Appear | Pricing for AI Visibility" is a better citation source than "Pricing." Use consistent brand naming across titles, H1s, and within body content so Perplexity can attribute content to the right entity.
Unlike traditional SEO, there is no standard dashboard for AI visibility. Building your own tracking framework is essential for understanding baseline, measuring improvement, and justifying investment. Here is a practical setup that does not require third-party tools.
Configure your server or CDN to log user agent strings. Filter for GPTBot, PerplexityBot, ClaudeBot, anthropic-ai, Google-Extended, and Bytespider (ByteDance). Record: crawler name, URL requested, timestamp, HTTP status code, and response time. This is your raw visibility data.
Every week, generate a report showing: total crawler visits by AI system, unique pages crawled, most frequently visited pages, pages that returned non-200 status codes to crawlers, and new pages that got crawled for the first time. This becomes your baseline trend line.
Once a month, query ChatGPT, Perplexity, Claude, and Gemini with 5–10 questions your target customers would ask. For each: note whether you are mentioned, whether the information is accurate, how you are positioned relative to competitors, and whether a link to your domain is included. Log this in a spreadsheet.
AI crawlers prefer fresh content. Track the last-modified date of your top 20 most-crawled pages. When a page goes 60+ days without an update, flag it for review. Even minor content updates signal recency to crawlers and can improve citation frequency.
As you improve AI visibility, you may start to see referral traffic from Perplexity and other AI platforms in your analytics. Track referral sources specifically and add UTM parameters to AI-specific landing pages so you can attribute conversions to AI-originated traffic as the channel matures.
Want to implement these playbooks with expert guidance?
Talk to usInstall in minutes. Keep your stack. Improve AI visibility.