logo

The AEO Checklist for Marketers and Developers

2026-05-27

A practical checklist your marketer and your developer run together. Use it to assess where you are with Answer Engine Optimization today, and to agree on what to fix next.

AEO breaks the way most marketing work used to be split. The content half belongs squarely to marketers. The technical half belongs squarely to developers. The two halves only earn citations together, because a beautifully written page nobody can crawl is invisible, and a perfectly accessible page that says nothing specific gets passed over.

This checklist is built to be run as a joint exercise. A marketer and a developer sit down with it, work through every item, mark current state honestly, and leave with a shared roadmap. We use it inside our own studio and with clients, and it is short enough to run in one focused afternoon.

How to use this

Print it, paste it into Notion, copy it into a shared doc. Whatever your team uses. Then work through it row by row, together. For each item:

  • Mark it done if the answer is yes today.
  • Leave it unchecked if it is not done, or if neither of you actually knows.
  • Note the owner if the work falls clearly to one role.

A legend for the tags on each item:

  • [M] Marketer owns it
  • [D] Developer owns it
  • [M+D] Joint, both roles required

When you finish, count the unchecked items. The ones tagged [M+D] are usually the most important to address first, because they are exactly the cross role gaps where things fall through.


1. Access and crawlability

The simplest reason an AI does not cite you is that an AI cannot read you. Rule that out first.

  • [ ] [D] robots.txt allows GPTBot, OAI-SearchBot, ClaudeBot, and PerplexityBot
  • [ ] [D] No staging or development noindex flag is live on production (on WordPress, check Settings, Reading; on Squarespace, check Settings, Crawlers)
  • [ ] [D] Host and CDN, including Cloudflare, are not blocking AI bots by default
  • [ ] [D] Sitemap.xml is published, current, and listed in robots.txt
  • [ ] [D] llms.txt is published at the site root, or through the platform's native feature where available
  • [ ] [D] Important content renders in the initial HTML, not only after client side JavaScript executes
  • [ ] [M] Marketer has personally viewed the page source on at least the homepage and one priority page, and confirmed the real content is there in plain text

2. Content that earns citations

The first half got machines to your content. This half is what makes them quote it. This section is mostly the marketer's domain, with one joint review.

  • [ ] [M] Each priority page leads with a direct answer to the question the page exists to answer, in the first one or two sentences
  • [ ] [M] Headings on priority pages are phrased the way a buyer would actually ask, not as internal product names
  • [ ] [M] Vague marketing claims have been replaced with concrete, dated, sourced specifics across the top ten pages
  • [ ] [M] At least one real statistic or cited source appears on every cornerstone page
  • [ ] [M] An FAQ section is present on the pages where buyers have recurring questions, such as services, pricing, and key product pages
  • [ ] [M] Editorial content beyond product or service pages exists, such as buying guides, comparisons, or explainer articles
  • [ ] [M+D] Marketer and developer have jointly reviewed the top five pages with one question: would an answer engine quote this paragraph, and would it know who to attribute it to

3. Structure and machine readability

Clean structure is the bridge between the content and the access layers. Both roles touch this.

  • [ ] [D] Real semantic heading hierarchy: one H1 per page, then H2 and H3 used correctly, not styled divs that look like headings
  • [ ] [D] Pages are server rendered or statically rendered for the content that matters
  • [ ] [D] Largest contentful paint under roughly 2.5 seconds on mobile
  • [ ] [D] Organization or LocalBusiness JSON-LD is present across the site
  • [ ] [D] Product schema on product pages, Article schema on blog posts, or other relevant type per page
  • [ ] [D] Image alt text is filled in for every meaningful image
  • [ ] [D] Canonical URLs are set on every page
  • [ ] [M] Marketer has read and approved the Organization schema description and confirmed it matches current positioning

4. Identity and consistency

Answer engines build confidence from agreement. When every source describes you the same way, the model becomes confident, and confident impressions get surfaced.

  • [ ] [M] The same one sentence description of what you do is used on the homepage, LinkedIn, Google Business Profile, social bios, and any directories you appear in
  • [ ] [M] Name, location, and core offering are identical across every listing
  • [ ] [M] Earned mentions are tracked in some form, knowing which credible third party sources describe your brand
  • [ ] [M] A short list of target podcasts, publications, and partners for earned mentions has been set for the quarter
  • [ ] [M+D] The sameAs array in your Organization JSON-LD lists every official profile, and every linked profile points back to the site
  • [ ] [M] Reviews on the platforms that matter for your category, such as Google, G2, or Trustpilot, are current and accurately reflect what you actually offer

5. Measurement and monitoring

You cannot manage what you cannot see. The developer sets the instrumentation up, the marketer reads the results, and both review the dashboard.

  • [ ] [D] Server or hosting access logs are reachable and parsable
  • [ ] [D] AI crawler user agents are tagged in logs: GPTBot, OAI-SearchBot, ClaudeBot, PerplexityBot, and Google-Extended
  • [ ] [D] AI referral traffic is segmented in analytics: chatgpt.com, claude.ai, perplexity.ai, and any other tools relevant to your category
  • [ ] [M+D] A priority prompt list has been defined, between five and fifteen real questions a buyer would ask
  • [ ] [M+D] AI share of voice across that prompt list is checked on a fixed cadence, monthly at minimum
  • [ ] [D] An alert fires when AI crawler hits drop toward zero
  • [ ] [D] An alert fires when AI share of voice drops sharply or a competitor overtakes you on prompts you used to own
  • [ ] [M] Conversions and revenue from AI referral traffic are tagged and tracked in your analytics, even if imperfectly

6. Working together

The last section is the one that decides whether any of the above gets done a second time. AEO is a system, not a launch.

  • [ ] [M+D] One person on the team is named as the AEO owner, accountable for the whole list across both roles
  • [ ] [M+D] A quarterly review is on the calendar where the team reads the data and decides what to change
  • [ ] [M+D] AEO items live on the same roadmap as the rest of the site work, prioritized jointly, not handed off
  • [ ] [M+D] The content calendar includes at least one substantive cornerstone AEO piece per quarter
  • [ ] [M+D] New page launches require an AEO sign off, using this checklist before the page goes live

Where to dig deeper

Each section above is a summary of a longer body of work. For the full reasoning and code:

  • Is Your Website Optimized For LLMs? is the foundation, covering the technical setup any site can use.
  • Our platform specific articles on Shopify, Squarespace, Webflow, and WordPress show exactly how to implement these items on your stack.
  • Our piece on AEO monitoring covers the measurement section in more depth, with code.

The COAK take

Most AEO work fails for the same reason most SEO work used to fail: the people who write the content and the people who build the site never sit at the same table. This checklist exists to put them there.

Run it once together, mark what is done and what is not, and leave with a single list. Then run it again every quarter. Score what improved, agree on what is next, and watch the trend. You will be doing what most of your competitors are not, which is treating AEO as a system the team operates rather than a project they finished.

If you want a partner who can run this with you the first time, that is what we do. Cause of a Kind is full stack, full service, on shore and in house. We help cool people build great products, and we help the people who build them work together.

Book a Systems Audit