Editorial
Guide
8 CX Survey Templates Built to Get Answers, Not Just Scores
The Gist
- Two environments, one strategy. Content now has to perform in traditional search and in LLM-mediated answers simultaneously — the signals that drive each overlap significantly.
- Structure is the foundation. Clean heading hierarchy, schema markup, and descriptive internal linking are the highest-leverage technical investments for AI citation visibility.
- Measurement has changed. Tracking "share of model" in LLM responses is now as important as tracking keyword rankings — and requires different tools and a different mindset.
Earlier this year, I sat down with my team and asked a question that would have sounded strange just 24 months ago. Are we writing content that AI can actually read, understand and cite?
The answer was more complicated than I expected. We had solid SEO fundamentals in place. Our content was performing reasonably well in traditional search. But when I started asking how often we were being surfaced in ChatGPT, Perplexity, Claude, or Google’s AI Overviews when buyers were asking category questions like “what's the best digital experience platform for financial services,” the answer wasn’t so clear.
That conversation kicked off an initiative to build a content and SEO playbook designed for a world where your content needs to perform in two distinct environments simultaneously. Traditional search engines still hold tremendous value, and the signals that drive rankings there haven’t disappeared. At the same time, LLMs are now doing their own form of content evaluation, pulling from what they’ve been trained on and what they can retrieve, and deciding whether your brand deserves to be mentioned in a response.
We built a guide for our teams. What follows is an adapted version of that playbook that is usable for both marketing teams and anyone in a company who’s writing for the web.
On-Page Fundamentals: The Structural Layer That Everything Else Depends On
The foundation of all of this is on-page structure. LLMs parse content the same way a careful human reader would. When the title, headings and logical flow from section to section are clean and predictable, the content is more extractable. When it’s inconsistent, the model has a harder time knowing what the page is actually about.
Title tags are where this starts. A title tag under 60 characters with the primary keyword as close to the front as possible communicates to both Google and language models what the page covers. Generic titles get rewritten by Google and ignored by AI. Specific, accurate titles appear in search results and responses. We standardized a format for our team: Primary Keyword — Secondary Keyword or Value Proposition | Liferay.
Heading structure is the part of this conversation that most content creators underestimate. There should be one H1 per page, containing the primary keyword, aligned closely with the title, H2s for major sections, H3s to break up long H2s when needed and no skipping levels. Screen readers rely on this hierarchy for accessibility, and LLMs use heading hierarchies to build their understanding of what a page covers and how it’s organized. If your content outline is logical and clearly signposted, it becomes significantly more useful as a training and retrieval source.
Meta descriptions are a lower-priority item since Google rewrites them frequently anyway. However, the underlying discipline matters. Every page needs a unique, specific description that leads with value rather than a description of the page. “Learn how to improve customer retention with a digital experience platform” is more useful than “This page covers our digital experience platform capabilities.” The former tells a user and a model what they’re going to get.
URLs and Content Length Are Signals, Not Afterthoughts
URLs deserve more attention than they typically receive. Descriptive, lowercase, hyphen-separated URLs that reflect the content hierarchy help users decide whether a result is relevant before they click, and they tell search engines how your content is organized. We group topically related content into directories, which helps Google learn how often different sections update and adjust crawl frequency accordingly. We also never change live URLs casually. When a URL has to change, a 301 redirect from the old URL to the new URL is non-negotiable. Every link pointing to that page represents equity you don't want to throw away.
One change we made explicitly in our playbook is in how we think about content length. We use SurferSEO to optimize for semantic completeness, targeting NLP terms that should naturally appear in a piece on a given topic. But we stopped chasing word count targets mechanically. If a question can be answered well in 1,400 words, we write 1,400 words. AI citation data increasingly tells us that well-organized, authoritative, specific content outperforms padded-out content designed to hit an arbitrary count.
Related Article: Brands Are Having a 'Crisis of Faith.' AEO Isn't Making It Easier.
Schema and Structured Data: Making Your Content Legible to Machines
Schema markup is one of the areas where I’ve seen the clearest evidence that small technical investments produce meaningful returns in AI visibility. In an analysis of 1,000 AI overviews, Digital Applied found that schema-marked pages are cited 2.3 times more than unstructured equivalents. The research on this is still developing, but the directional signal is consistent: structured content gets cited more.
The basic idea behind schema is that it gives machines explicit context about what your content is and who produced it. Without it, a language model has to infer context from surrounding text. With it, you’re telling the model directly: this is a blog post, written by this author, published on this date, about this topic. That helps the model determine whether your content is authoritative and citable.
Page Type Determines Which Schema to Implement
Every page on your site should carry Organization schema and BreadcrumbList schema. Beyond that, the requirements vary by page type. Blog posts need Article schema with author, publication date and modification date. Product and capability pages need Product or Service schema. Pages with FAQ sections, which are high-value for both featured snippets and AI Overviews, need FAQPage schema. Webinar replay pages should carry VideoObject schema including the transcript, because Gemini reads full transcripts and ChatGPT and Perplexity rely on titles and snippets for retrieval.
All of this is validated through Google's Rich Results Test, which should pass before any page goes live. Schema that’s implemented incorrectly is worse than no schema at all, because it can create errors in Search Console that require cleanup.
AEO Readiness Checklist
Use this checklist to evaluate whether your content, technical infrastructure and measurement practices are prepared for answer engine optimization (AEO), AI Overviews and LLM-driven discovery.
| Area | Requirement | Status |
|---|---|---|
| Content Structure | Every page has a clear, keyword-focused title tag under 60 characters. | ☐ |
| Content Structure | Each page contains a single H1 aligned with the page topic. | ☐ |
| Content Structure | Heading hierarchy follows H1 → H2 → H3 without skipping levels. | ☐ |
| Content Structure | Content directly answers common buyer and customer questions. | ☐ |
| Content Structure | Articles are written for completeness, not arbitrary word-count goals. | ☐ |
| Content Quality | Subject matter experts are identified and credited where appropriate. | ☐ |
| Content Quality | Statistics, claims and research are supported with credible sources. | ☐ |
| Content Quality | Pages are updated regularly to maintain freshness and accuracy. | ☐ |
| Schema Markup | Organization schema is implemented sitewide. | ☐ |
| Schema Markup | BreadcrumbList schema is implemented sitewide. | ☐ |
| Schema Markup | Article schema is applied to blog and editorial content. | ☐ |
| Schema Markup | Product or Service schema is applied to solution pages. | ☐ |
| Schema Markup | FAQPage schema is used on FAQ-rich content. | ☐ |
| Schema Markup | VideoObject schema is applied to video and webinar content. | ☐ |
| Schema Validation | Structured data passes Google's Rich Results Test. | ☐ |
| Internal Linking | Every article links to priority internal pages. | ☐ |
| Internal Linking | Anchor text is descriptive and keyword relevant. | ☐ |
| Internal Linking | Internal links support topic clusters and content hubs. | ☐ |
| Technical SEO | URLs are descriptive, lowercase and hyphenated. | ☐ |
| Technical SEO | 301 redirects are in place for retired or changed URLs. | ☐ |
| Technical SEO | XML sitemaps are current and submitted. | ☐ |
| AI Accessibility | robots.txt allows GPTBot access. | ☐ |
| AI Accessibility | robots.txt allows OAI-SearchBot access. | ☐ |
| AI Accessibility | robots.txt allows Google-Extended access. | ☐ |
| AI Accessibility | robots.txt allows Anthropic-AI and ClaudeBot access. | ☐ |
| AI Accessibility | robots.txt allows PerplexityBot and CCBot access. | ☐ |
| Measurement | Traditional SEO rankings and organic traffic are tracked. | ☐ |
| Measurement | AI referrals are segmented and measured in analytics. | ☐ |
| Measurement | Brand mentions in ChatGPT, Gemini, Claude and Perplexity are monitored. | ☐ |
| Measurement | "Share of model" or AI visibility metrics are tracked against competitors. | ☐ |
| Governance | A documented AEO strategy exists across content, SEO and PR teams. | ☐ |
| Governance | Content creators are trained on AI-friendly content structure. | ☐ |
| Governance | A recurring process exists for reviewing AI citation performance. | ☐ |
Internal Linking: The Frequently Neglected High-Value Habit
Internal linking is, in my experience, the SEO practice that gets the least attention relative to its impact. It’s also directly relevant to AI visibility because it affects how authority flows through your site and which pages get treated as most important.
The framework we built around is a 70/30 split for informational content: 70% of links in an article pointing to internal Liferay pages, 30% pointing to credible external sources. The external links demonstrate that your content engages with the broader knowledge ecosystem, instead of existing in isolation. The internal links do the work of routing authority to the pages you most want to rank.
A Priority URL Matrix Makes Anchor Text Consistent Across Every Channel
We maintain a priority URL matrix that maps target categories to specific pages, preferred anchor text and acceptable variations. Every article that goes live links to at least one page from that matrix. The anchor text is descriptive and keyword-rich text that tells both the reader and the search engine exactly what they’re going to find. We also coordinate this matrix with our PR and link-building work, so the anchor text strategy is consistent whether a link is coming from an internal blog post or an external publication.
For anyone outside the marketing team who’s writing content for the web, the most important habit to develop is linking to specific, relevant internal pages using descriptive language.
Related Article: AEO, GEO, SEO: What's the Best Search Playbook?
AEO Readiness FAQ
Editor's note: Five questions every marketing, SEO and content leader should ask when preparing their organization for answer engine optimization.
Measurement: Tracking Performance in Both Worlds
This is the area that’s changed most dramatically for our team over the last year. Traditional SEO measurement is well-established. Those metrics still matter, and we track them consistently.
What’s newer is AI visibility measurement, and the tooling is still maturing. We use Peec.ai to track brand mentions in generative AI platforms and to measure what we think of as “share of model,” which is how often Liferay is named in LLM responses to category queries relative to competitors. We also track generative referral traffic in GA4, looking at source and medium data for traffic originating from AI platforms.
Webinar
Jun
18
The 5-Question CX Audit: Benchmark Your CX Operations for 2026
Built around insights from the 2026 CX Outsourcing Report, this live session puts the audit into practice.
Register
Webinar
Jun
25
The Hidden Cost of Fragmented Customer Communication
Discover why growing businesses are rethinking the systems, workflows and communication habits shaping customer experience.
Register
Webinar
Jun
30
How Modern Marketing Is Exposing the Limits of Legacy CMS
Why marketers are rethinking CMS workflows that slow publishing, personalization and campaign execution.
Register
Webinar
Jul
14
Content Leaders Collective: Proving Content’s Business Impact
Join us as top content leaders look beyond the buzzwords to share how they actually prove ROI and scale what works.
Register
Webinar
On demand
From Legacy to Launch-Ready: How Gainbridge Made Its Website a Marketing-Led Growth Engine
Join in to learn how a D2C annuity brand gave marketing full website ownership — without slowing down or risking compliance.
Watch Now
Webinar
On demand
Content Strategy Leaders Live: Managing Risk, Compliance & AI in Financial Services
Learn how financial services leaders are modernizing content systems without disrupting trust, compliance or experience.
Watch Now
Your robots.txt file needs to explicitly allow AI crawlers if you want those platforms to index your content. The list we maintain includes GPTBot, OAI-SearchBot, Google-Extended, Anthropic-AI, ClaudeBot, PerplexityBot, and CCBot. These crawlers won’t ignore a disallow directive, so if your robots.txt is blocking them your content isn’t being read by the platforms you want to be cited in.
A Priority URL Matrix Makes Anchor Text Consistent Across Every Channel
We’ve also shifted from thinking about “rankings” to thinking about “presence.” In traditional search, you’re tracking positions on a results page. In AI-mediated search, you’re tracking whether your brand is being mentioned at all in relevant conversations. That’s a different metric with different implications for what content you produce and how you structure it.
We’re still figuring out the right cadence and benchmarks for AI visibility measurement. Anyone who tells you they have this fully dialed in is overstating it. What I can say is that measuring it consistently, even imperfectly, gives you a feedback loop that you didn’t have before. And that feedback loop is already informing decisions about where we invest our content efforts.
The question I started with, whether we're writing content that AI can actually read and cite, doesn't have a clean answer yet. What I can say is that asking it changed how we work. The signals are real, the tooling is improving, and the teams that build the habit now will have a meaningful head start. The playbook is still being written, and that's what makes it worth paying attention to.
Learn how you can join our contributor community.