Run this checklist against a baseline first.
Before working through these items, spend twenty minutes querying ChatGPT, Perplexity and Gemini with the five or six questions your customers are most likely to ask — category plus city, service plus neighbourhood, comparison questions. Document what each assistant says. That baseline tells you which items on this checklist are actively costing you citations, and it gives you a before picture to compare against after you make improvements.
The checklist is organized by the layer of the problem, not by difficulty. Schema is listed first because it is often the fastest fix with the most immediate impact. Content is listed later not because it matters less but because fixing a broken entity foundation makes all subsequent content work more effective. Skipping to content while the schema is wrong is like painting a wall before fixing the damp — the surface improves, the underlying problem remains.
This checklist covers the core items. A full AI visibility audit goes deeper into competitive analysis and cross-source consistency checks. Use this checklist as your own first pass before deciding whether you need professional help.
Is your business machine-readable?
Schema markup gives AI models a structured extract of your business facts. Without it, models have to infer everything from prose — less reliable and lower-confidence.
Check for Organization or LocalBusiness JSON-LD on your homepage.
Use Google's Rich Results Test (search.google.com/test/rich-results) to check your homepage. You should see an Organization or LocalBusiness block. If it is missing, this is your first priority. If it exists, verify that name, address, phone, URL and business type are all populated and accurate.
Verify schema facts match your live website content.
The name in your schema should be exactly the name used in your website header and About page. The address should match what appears in your site footer and Contact page. Discrepancies between schema and on-page content reduce model confidence.
Add Service schema to your key service pages.
Each major service page should have a Service schema block naming the service type, provider and geographic area served. This is separate from the Organization schema on your homepage. It signals to models that you provide specific services, not just that you exist as a business.
Add FAQ schema to pages with FAQ sections.
If your service pages or About page include FAQs, wrap them in FAQPage JSON-LD schema. This formats the question-answer pairs in a way models extract reliably and is one of the most cited content formats in AI answers.
Does every source tell the same story?
AI assistants that search the live web cross-reference multiple sources. When your name, address and phone number conflict across those sources, model confidence drops.
Audit your Google Business Profile for completeness.
Every field should be filled: business name (exact match to your canonical name), address, phone, website, hours, primary and secondary categories, services list, business description (250 characters minimum). Check that photos are present and recent. An incomplete GBP is invisible in both Maps ranking and AI recommendation queries.
Check your business name across the top five directories.
Google Business Profile, Apple Maps (via Apple Business Connect), Yelp, Bing Places and the Better Business Bureau. Your business name should be spelled and formatted identically on all five. Search your business name in each platform and note any variations — even 'LLC' versus nothing, or a comma placement, is worth standardizing.
Verify phone and address consistency across all sources.
Run a BrightLocal or Moz Local citation scan to see everywhere your business appears online. Look for outdated phone numbers (common after a number change), old addresses (common after a move) and name variations from early registrations. Flag and update every discrepancy.
List your business on any relevant industry directories.
Depending on your sector, this might mean the Canadian Bar Association directory for lawyers, the Royal College of Dental Surgeons of Ontario directory for dentists, or the local Chamber of Commerce for service businesses. Each consistent listing adds corroborating evidence for your entity.
Does your site have answers AI can quote?
Structural fixes make you citable in principle. Content fixes make you citable in practice — they give models specific passages to lift when someone asks a question you should be answering.
Check the first sentence of each service page.
The opening sentence of every service page should state the service, your business category and your location in plain factual terms. Not a marketing hook — the actual information. Rewrite any opening that starts with an adjective, a rhetorical question or a brand promise.
Verify your geographic coverage area appears on every key page.
The city, neighbourhood or service region should appear in the first paragraph of every major page. AI assistants filter recommendations by location. A page that never mentions where it operates cannot appear in location-filtered queries. This is a common gap — and a five-minute fix per page.
Add or improve FAQ sections with complete, standalone answers.
Write out the ten most common questions your customers actually ask, then answer each one in two to four plain sentences. Each answer should be complete on its own — meaning it makes sense without the question. Wrap the whole section in FAQPage schema. See the full framework at <a href='content-structure-for-ai-citations.html'>content structure for AI citations</a>.
Add an llms.txt file to your site root.
An llms.txt file at yourdomain.com/llms.txt summarizes your site for AI crawlers: who you are, what you do, where you operate, and links to your key pages. It is a low-effort, high-signal file that most businesses have not yet created. Write it in plain Markdown using the open spec at llmstxt.org.
The authority layer that tips competitive queries.
For competitive local queries — where multiple businesses have clean entities and decent content — review signals and third-party mentions are often the tiebreaker.
Do you have enough recent reviews?
Check your Google review count and the date of your most recent review. If your last review is months old, implement a review request system — an automated follow-up to customers after service completion that includes a direct link to your Google review page. Models weight recency, not just total volume. A business with fifty reviews all from three years ago competes poorly against one with twenty reviews from the last ninety days.
Is your business mentioned on credible external sites?
Search your business name in quotes. Look at the first two pages of results. Are you mentioned in local news, industry publications, neighbourhood blogs or business association features? If not, there are accessible ways to change that: press releases for newsworthy events, local business features, guest content on industry sites, Chamber of Commerce member spotlights. Unlinked mentions on trusted sites still add corroborating evidence for your entity.
Are you active on Google Business Profile?
Regular Google Posts — at minimum one per week — signal an active, maintained business. Models that retrieve live GBP data weight activity. Posts that name a specific service, include a photo and contain a clear call to action contribute more than generic announcements. This is a small investment for a meaningful freshness signal, particularly important for AI recommendation in local queries.
Related, and what to do next.
Questions, answered.
What is an AEO checklist?
An AEO checklist is a structured set of items covering the signals that determine whether AI assistants cite and recommend your business. It covers schema markup and entity data, NAP consistency across directories, answer-shaped content on service pages, FAQ schema, review signals and third-party mentions. Working through it in order gives you the fastest path from AI-invisible to AI-cited.
How long does it take to work through this checklist?
Schema and entity checks can be done in a few hours. The directory audit typically takes a day, with corrections spread over a week. Content rewrites depend on how many service pages you have — budget one to two hours per major page. The full checklist can be completed over two to four weeks for most small businesses.
Which item on the AEO checklist has the most impact?
Organization or LocalBusiness schema on your homepage, if it is missing, is consistently the highest-impact single fix. It gives every AI model a structured extract of your business facts immediately. For businesses that already have schema, the highest-impact remaining item is usually NAP consistency across directories.
Do I need to do all of this myself, or can someone do it for me?
You can do the checklist audit yourself with no tools beyond Google's Rich Results Test and a few manual directory checks. The fixes vary — schema implementation requires some technical knowledge, directory cleanup is time-consuming but not complex, content rewrites require care. TRIVOX handles the full process as part of GEO engagements, including the initial audit and prioritized implementation.
How do I know when my AEO work is paying off?
Run the AI visibility baseline check — query ChatGPT, Perplexity and Gemini with your core category and location queries — before starting, and again four to six weeks after completing the structural items. Track whether your mention rate improves and whether the descriptions become more accurate. Schema and entity fixes often show results within a few weeks. Content and review improvements take longer.
Does this checklist apply to businesses in Canada or the GTA specifically?
The checklist applies anywhere, but all the same signals apply to Canadian and GTA businesses. The Canadian Business Directory, the Better Business Bureau Canada, and the local Chamber of Commerce are all worth prioritizing for citation building. Google Business Profile optimization is especially impactful in the GTA because query volume for local services is high and the competitive density means clean entity signals are the difference between appearing and not.