How to Get Cited by ChatGPT in 2026: 7 Concrete Steps
Most websites are invisible in ChatGPT, not because they are bad, but because they are not structured to be cited. Seven specific, technical steps that move a site from invisible to citable, in priority order.
The honest read on ChatGPT visibility in 2026
Eighty-eight percent of small business websites are invisible in ChatGPT for their category-defining queries (Omni Eclipse 2025 study). The reason is not that those businesses are bad. The reason is that they are not structured to be cited. ChatGPT cites a small set of sources for any given question, and the criteria are mostly technical: can the AI read the site, does it have structured data, is the content answer-first, is the entity unambiguous. The seven steps below address each of those criteria in priority order.
1. Allow AI crawlers in your robots.txt
The most basic requirement and the most commonly missed. AI engines crawl with named bots: GPTBot (ChatGPT), ClaudeBot (Claude), PerplexityBot (Perplexity), Google-Extended (Google AI), Bytespider (TikTok), Amazonbot, cohere-ai. If any of these are blocked in your robots.txt, the corresponding AI cannot read your site, period. We audited 10 Canadian Shopify stores recently and 9 of them were blocking GPTBot without knowing it.
Verify: visit yourdomain.com/robots.txt and search for each bot name. None should be under Disallow. If you find blocks, remove them and republish. The change takes effect within 24 to 48 hours.
2. Add Organization and LocalBusiness schema
JSON-LD structured data is how AI engines establish entity identity. An Organization schema block tells the AI exactly who you are, what you do, where you are located, and what other properties you control. LocalBusiness schema is the local-search variant, with geo coordinates and service area. Without these blocks, AI engines have to guess, and they often guess wrong (or skip you).
Verify: paste your homepage URL into Google's Rich Results Test or validator.schema.org. You should see at least Organization (or LocalBusiness for local businesses) and BreadcrumbList. Most sites we audit have only basic site metadata.
3. Add FAQPage schema to your highest-traffic pages
FAQPage schema is the single highest-leverage AEO move. AI engines extract Q&A pairs from FAQPage schema and use them directly in their answers. A page with 5 to 8 well-structured FAQs in proper schema format is dramatically easier for AI to cite than the same content as plain HTML.
Add FAQPage schema to your homepage, your top 3 product or service pages, and your most-trafficked blog posts. Each Q&A should be 40 to 60 words for the answer (the optimal length for AI engine extraction). Validate at validator.schema.org before deploy.
4. Create and publish an llms.txt file
llms.txt is a plain-text file at the root of your site (yourdomain.com/llms.txt) that summarizes your business in a format AI language models can read. The format is similar to robots.txt but for content, not crawling. The standard is new and not yet widely adopted, which is exactly why early movers are easy to cite.
Include: business description (2 to 3 sentences), primary services or products, target customer, three to five links to your most important pages with descriptions. Keep it under 500 words. Update it every 6 months or when your business changes meaningfully.
5. Establish your branded entity across multiple sources
AI engines trust entities they can verify across multiple independent sources. A business that exists only at one URL is harder to confidently cite than one with a Google Business Profile, a LinkedIn company page, consistent listings on relevant directories, and earned mentions on third-party sites.
The cheapest moves: complete your Google Business Profile (categories, services, hours, photos, posts), claim your LinkedIn company page, add yourself to 3 to 5 relevant directories with consistent NAP (name, address, phone). Each consistent listing strengthens entity confidence.
6. Publish category-answering comparison content
The single highest-leverage content move for AEO is publishing comparison or listicle content that directly answers your category's defining query. "Best [your category] in [your location]" or "[your product] alternatives" or "How to choose a [your service]" are all examples. The post should be 1,500 to 2,500 words, include yourself honestly alongside competitors, and structure each entry consistently.
This is exactly what we did with our own Best AI Agencies in Vancouver for Small Businesses (2026) post. Listing ourselves alphabetically alongside competitors. Including our own honest description. The post is now a citation candidate for the category query our SVB found we were invisible for.
7. Build internal-link density toward your highest-authority pages
AI engines use the same authority signals search engines use, with internal linking among the strongest. Pages that are linked from many other pages on your own site are inferred to be more important than orphan pages. Make sure your category-answering content is linked from your homepage, your product pages, and your other relevant blog posts. Use descriptive anchor text, not "click here."
This is also the single most-overlooked AEO move because it does not feel like AI work. It is just good site architecture. AI engines reward it anyway.
How to know it is working
The honest answer: monitoring AI engine citations is harder than monitoring Google rankings, because there is no equivalent of Search Console for ChatGPT. Three pragmatic monitoring practices we have found useful:
Practice 1: monthly query log. Pick five to ten queries your ICP would actually ask ChatGPT in your category. Once a month, copy and paste each into ChatGPT, screenshot the answer, save the screenshot with the date. After three months you will have a clear before/after of whether you are starting to be cited and how the answers are evolving.
Practice 2: branded query monitoring. Each month, ask ChatGPT and Perplexity "What is [your business name]?" and check whether the description is accurate, complete, and cites your properties. If the description is generic ("a company that does X") and cites no specific URLs, your branded entity signal is weak. If it accurately names your services, location, and links to multiple of your properties, your branded AEO is healthy.
Practice 3: referrer log. Some AI engines (Perplexity, Bing Copilot, Google AI Mode) increasingly send referrer traffic that shows up in your analytics under recognizable User-Agent strings or referrer URLs. Check your top referrers monthly. The growth curve will be slow at first then steeper as your citation count compounds. The first AI engine referral is the inflection point worth celebrating.
If you want a structured monthly report instead of doing this manually, the Strategic Visibility Brief ($500/month) runs all three practices plus the trajectory analysis and prescribed actions.
What to skip (for now)
Two things you may have read about that are not yet worth the time:
- Paid AI engine placements. ChatGPT does not currently sell sponsored citations. Perplexity briefly tested ad placements in 2024-25 and pulled back. Until paid placements are a stable, scaled offering, organic AEO is the path.
- Generic AEO tools that promise rank tracking. AI engine citations are too sparse and too query-specific for traditional rank-tracker logic. Spot-check by asking ChatGPT directly the queries your customers would ask, and read what it cites. Tools that claim to "rank you in ChatGPT" are mostly snake oil today.
What to do this week
Three concrete moves you can do in one focused afternoon:
- Audit your robots.txt for AI crawler blocks (5 minutes)
- Run your homepage through validator.schema.org and identify what schema is missing (15 minutes)
- Ask ChatGPT three category questions your ICP would actually type. Note which sources it cites, and what your gap is (30 minutes)
That alone tells you whether you have a robots.txt problem, a schema problem, an entity problem, or a content problem. From there, the priority order in this post tells you which to fix first.
The most common pattern we see across audits: a site has the basics (robots.txt is fine, basic schema is present) but is missing the leverage moves: FAQPage schema on top pages, llms.txt, and category-answering content. The fix order matters because the leverage moves compound; without the basics in place first, the leverage moves cannot land.
If you would rather have the work done for you, our SEO + AEO Optimization applies all 7 steps to your specific site as a done-for-you engagement, From $2,500, in 1 to 3 weeks. Or run the free Website Scorecard to see where you stand on six visibility categories before deciding what to do.
Ready to put AI to work for your business?
From AI consulting to custom integrations, SyncSpark helps Canadian businesses adopt AI practically.
Book a Free Discovery Call