NewRead our latest case study, published by Google Read it here

Talk to Founder

Workflow 路 Diagnostic

FAQ content exists.
FAQ schema doesn't.

A copy-paste Claude prompt that audits your top 30-100 pages for FAQ schema implementation gaps, classifies each as MISSING / BROKEN / PARTIAL / STRONG, prioritizes by traffic 脳 citation leverage, and produces add/expand/restructure recommendations per page. Single-element AEO audit on the highest-leverage citation surface.

Jump to the prompt Pair with Content Audit Memo
Missing 路 Broken 路 Partial 路 Strong
4statesMissing 路 Broken 路 Partial 路 Strong
Typical FAQ schema gap rate
60-80%missingTypical FAQ schema gap rate
Citation lift on schema'd pages
3-5xliftCitation lift on schema'd pages
Re-audit after schema deployment
QuarterlycadenceRe-audit after schema deployment

01The Problem in 60 Seconds

The content is on the page.
The markup isn't.

A B2B SaaS team has 60+ pages on their site. Many include FAQ-style content, "frequently asked questions" sections at the bottom of landing pages, Q&A blocks in blog posts, accordions in product pages, knowledge base entries. The content already exists. But when you inspect the page source, almost none of it has FAQPage schema markup. AI search engines (ChatGPT, Claude.ai, Perplexity, Gemini) parse structured data when extracting answers for citations. Pages with FAQPage schema get cited 3-5x more than pages with the same FAQ content but no schema, because the structured Q&A format maps directly to the AI's answer format. The team is doing the hard work (writing the questions and answers) but missing the easy work (marking it up).

The deeper problem is that FAQ schema is the single highest-leverage technical AEO move, but most B2B SaaS teams don't audit it systematically. Content teams don't think about schema; dev teams don't think about which pages have FAQ content; SEO teams audit traffic and rankings but not citation eligibility. The result: a long tail of high-traffic pages with weak or missing schema, sitting unfixed because nobody owns the audit. A page-by-page audit takes 6-10 hours per page; nobody has 600 hours for the full library.

This workflow runs the systematic audit. Claude fetches your top 30-100 pages (sorted by GSC impressions 脳 citation leverage), classifies each as MISSING / BROKEN / PARTIAL / STRONG, and produces prioritized fix recommendations. Run quarterly, schema deployments need 2-4 weeks for AI engines to re-crawl and recognize the change. Pair with AEO Content Audit Memo (page-level audit) to coordinate single-element fixes with deeper page restructuring.

4-State Schema Classification 路 Each Page Falls Into One Different states require different fixes
  • State 1 路 Most common Missing Page has FAQ-style content but NO FAQPage schema. Q&A structure exists in HTML; structured-data markup doesn't. Typical: 60-80% of pages with FAQ content. Fix: ADD schema
  • State 2 路 Worst signal Broken Schema exists but malformed: invalid JSON, missing required properties (Question, acceptedAnswer), validator errors. Sometimes worse than missing because it signals incompetence. Fix: FIX schema
  • State 3 路 Underdone Partial Valid schema but incomplete coverage, 5 of 12 questions marked up, or schema covers one section only. Underutilizes existing content. Fix: EXPAND schema
  • State 4 路 Maintain Strong Valid schema covering all FAQ content, high-quality answer structure (50-150 word answers). Citation-eligible. Maintain, no immediate action. Fix: MAINTAIN

02The Prompt

Copy this prompt into
Claude Desktop.

The gold variables (your brand, page list with GSC impressions, page type taxonomy) are the parts you edit. Run quarterly to track schema deployment progress and surface new gaps.

claude_desktop, faq_schema_audit.md 
RoleYou are running the quarterly FAQ Schema Audit for my B2B SaaS site. Audit the top 30-100 pages for FAQPage schema implementation. Classify each page as MISSING / BROKEN / PARTIAL / STRONG. Prioritize by traffic 脳 citation leverage. Produce add/expand/restructure recommendations per page.

My BrandBrand: [your B2B SaaS brand name]
Site URL: [your domain]
Brand positioning: [1 line on what you do]

Page List// Pull top 30-100 pages from GSC sorted by impressions. Higher-impression pages get FAQ schema fixes prioritized first.

Page list (URL + 90-day impressions + page type):
[paste 30-100 rows from GSC: URL | impressions | page_type (category / comparison / listicle / blog / product / how-to)]

Audit scope: [Top 30 by impressions = standard / Top 50 = thorough / Top 100 = comprehensive]

Task1. Fetch each page and audit FAQ schema implementation:
   - Inspect page HTML for FAQ-style content (sections labeled "FAQ" / "frequently asked questions" / accordion Q&A blocks / structured Q&A lists)
   - Inspect page schema markup for FAQPage schema implementation (JSON-LD or microdata)
   - Validate schema if present (required properties: Question + acceptedAnswer with name + text)
   - Count: questions in content vs questions marked up in schema

2. Classify each page into 4 states:
   - MISSING: FAQ content exists, NO schema markup
   - BROKEN: Schema exists but invalid (validator errors, missing properties, malformed JSON)
   - PARTIAL: Valid schema but covers some not all FAQ content (e.g. 5 of 12 questions marked up)
   - STRONG: Valid schema covering all FAQ content with quality answer structure (50-150 words per answer)

3. Calculate citation leverage per page:
   - Page type weight: category page 1.5x 路 comparison page 1.4x 路 listicle 1.3x 路 how-to 1.2x 路 blog 1.0x 路 product 0.9x
   - Combined leverage = impressions 脳 page_type_weight
   - Sort pages by leverage descending; highest-leverage pages get fixed first

4. Per page, produce action recommendation:
   - ADD: write FAQPage schema JSON-LD for existing FAQ content
   - FIX: debug and repair existing schema (specify error type: invalid JSON / missing properties / validator errors)
   - EXPAND: add additional questions to schema, add additional FAQ content if needed
   - MAINTAIN: no immediate action; monitor citation rate

5. Aggregate counts:
   - Total pages audited
   - State distribution (% MISSING / BROKEN / PARTIAL / STRONG)
   - Action distribution (counts: ADD / FIX / EXPAND / MAINTAIN)
   - Estimated total ship time per action category (ADD ~1-2 hrs/page, FIX ~30-60 min/page, EXPAND ~1-3 hrs/page, MAINTAIN 0)

6. Build prioritized fix queue:
   - Top 10 pages: highest leverage 脳 ADD/FIX/EXPAND action. These are the priority quick wins
   - Top 30 pages: account for ~80% of total citation leverage
   - Pages 31+: long tail, defer to next quarter unless no high-priority work remains

Output format1. Headline: total pages audited, state distribution, total estimated ship time, top 10 priority pages.

2. Status distribution bar (visual): % per state across audited pages.

3. Action category breakdown: ADD count + FIX count + EXPAND count + MAINTAIN count, with ship time per category.

4. Per-page audit table for top 30: URL / impressions / state / action / rationale.

5. Honest calibration:
   - If MISSING is < 20% of audited pages, the audit is largely complete and quarterly cadence can extend to 6-month cadence. The library is in good shape; focus on maintaining new pages rather than re-auditing.
   - If BROKEN is > 15% of audited pages, there's a systemic schema implementation issue (likely a CMS / theme / plugin problem). Investigate the root cause before fixing pages individually.
   - If a page has FAQ content but the page type is product/landing-page, recommend ADD only if the FAQ content is high quality (50+ word answers). Thin FAQs on product pages can hurt page authority.
   - If a page has 20+ FAQ questions, recommend splitting into a dedicated FAQ pillar page rather than over-stuffing schema on a single page. Schema with 20+ questions has diminishing returns; focused 8-12 question pages perform better.
   - If page type weights produce a leverage ranking that surprises you (e.g. a product page outranking a category page), the issue is impression-volume disparity. Trust the math; products with high impressions are worth schema-ing.

// Be specific in the audit. "Schema is broken" is generic. "Schema has invalid JSON syntax: missing comma after 'mainEntity' array" is actionable.
// For ADD recommendations, generate the actual JSON-LD schema as a code block in the output. Don't just say "add FAQPage schema", produce the schema ready to paste into the page template.
// Don't audit pages with < 100 impressions/quarter. They're long-tail and not worth the audit cost. Skip them and note in calibration.
// FAQPage schema is for true Q&A content. Don't recommend it for "what we do" sections, glossary entries, or feature descriptions, those use different schema types (DefinedTerm, ItemList, etc.). Stay disciplined.

Pull GSC page list first. Use gsc via the free Growthspree MCP to export top 30-100 pages by impressions with page type categorization. Pair with AEO Content Audit Memo for page-level audits and AEO Net-New Brief Generator for net-new content production.

03What Claude Returns

The schema status. The fix queue.

Sample output for a hypothetical mid-market B2B SaaS brand with 50-page audit. Audit surfaces 32 pages MISSING, 6 BROKEN, 7 PARTIAL, 5 STRONG. Total estimated ship time: 64 hours of combined dev + content work across 6-8 weeks.

Claude 路 GSC + page audit

Pulling top 50 pages from GSC and auditing FAQPage schema implementation per page.

growthspree.gsc.pages.list { date_range: "last_90d", sort_by: "impressions", limit: 50, properties: ["url","impressions","clicks","page_type"] }
fetch_and_audit 50 pages 脳 FAQ content detection + schema validation = state classification per page

50 pages audited. State distribution: 32 MISSING (64%) 路 6 BROKEN (12%) 路 7 PARTIAL (14%) 路 5 STRONG (10%). Total estimated ship time: 64 hours of combined dev + content work across 8 weeks. Top 10 pages account for 58% of total citation leverage. Recommended sequencing: Tier 1 (this week) = FIX the 6 BROKEN pages 路 Tier 2 (next 4 weeks) = ADD schema to the top 10 MISSING pages 路 Tier 3 (next 8 weeks) = remaining 22 ADDs + 7 EXPANDs.

Schema Status Distribution 路 50 Pages Audited Most common state typically MISSING
  • MISSING 32 pages 路 ADD
  • BROKEN 6 pages 路 FIX
  • PARTIAL 7 pages 路 EXPAND
  • STRONG 5 pages 路 MAINTAIN
Action Category Breakdown 路 By Effort + Owner
  • ADD schema 32 pages Write FAQPage JSON-LD for existing FAQ content. ~48 hours dev work (1.5h avg/page). Most common fix; usually batched into a single sprint.
  • FIX schema 6 pages Debug and repair invalid schema. ~5 hours dev work. Investigate root cause first, 6 broken pages typically share a systemic CMS/theme issue.
  • EXPAND schema 7 pages Add additional questions to schema; sometimes write additional FAQ content. ~11 hours combined (1.5h avg/page).
  • MAINTAIN 5 pages No immediate action. Monitor citation rate at 30-day re-audit. 0 hours. These are the gold-standard pages to model new content after.
Per-Page Audit Table 路 Top 12 by Citation Leverage Sorted by impressions 脳 page-type weight
PageImprStateActionRationale
/best-revenue-attribution-platformsCategory pillar 路 weight 1.5x8.4KMISSINGADD9 FAQ-style questions in content; zero schema. Highest priority, 12,600 weighted leverage points.
/blog/icp-scoring-system-b2b-saasListicle 路 weight 1.3x6.2KMISSINGADD8 FAQ-style questions; high-quality 80-120 word answers. ADD-ready content; just needs schema markup.
/blog/linkedin-vs-google-ads-b2bComparison 路 weight 1.4x5.1KPARTIALEXPANDSchema has 4 of 11 questions marked up. Expand to cover remaining 7 questions. Already valid format, straightforward expansion.
/blog/saas-conversion-rate-benchmarksListicle 路 weight 1.3x4.8KSTRONGMAINTAINValid schema covering 12 questions with 80-100 word answers. Gold-standard implementation, model new pages after this.
/best-b2b-saas-marketing-agencyCategory pillar 路 weight 1.5x3.9KBROKENFIXSchema exists but invalid: missing acceptedAnswer.text property on 3 questions. Validator returns errors. Worse than missing, fix urgently.
/blog/abm-for-saas-startupsHow-to 路 weight 1.2x3.7KMISSINGADD7 FAQ-style questions in content; zero schema. Standard ADD work; ~1.5h to schema-ize.
/best-linkedin-ads-agency-b2bCategory pillar 路 weight 1.5x3.4KMISSINGADD11 FAQ-style questions in content. Long FAQ section makes this a high-leverage ADD.
/blog/google-ads-mcp-saas-guideHow-to 路 weight 1.2x3.0KSTRONGMAINTAINValid schema covering 8 questions with quality answers. Maintain.
/blog/junk-leads-qlaHow-to 路 weight 1.2x2.8KPARTIALEXPANDSchema covers 3 of 9 questions. Expand to full coverage.
/blog/hubspot-attribution-dashboardHow-to 路 weight 1.2x2.6KMISSINGADD6 FAQ-style questions; standard ADD.
/blog/saas-ppc-agencies-2026Listicle 路 weight 1.3x2.4KBROKENFIXSchema has malformed JSON: missing comma after mainEntity array opening. Likely shares root cause with other broken pages.
/blog/cohort-roas-90-180-365How-to 路 weight 1.2x2.2KSTRONGMAINTAINValid schema covering 7 questions with quality answers. Maintain.
Top 10 priority fixes account for 58% of total citation leverage. Sequencing recommendation: Tier 1 (this week), investigate root cause of 6 BROKEN pages (likely shared CMS/theme issue); ship the FIX once root cause is patched. Tier 2 (next 4 weeks), ADD schema to top 10 MISSING pages (~15 hours dev work). Tier 3 (next 8 weeks), ADD schema to remaining 22 MISSING pages and EXPAND 7 PARTIAL pages (~44 hours combined). Re-audit at 30 days to verify schema deployment + monitor citation lift in AI search engines. Want me to also generate the actual JSON-LD schema templates for the top 10 priority pages, or proceed to the root-cause investigation memo for the 6 BROKEN pages?
TIME ELAPSED: 18 MINUTES 路 SAME AUDIT BY HAND: 50-80 HOURS PAGE-BY-PAGE

04Setup

Four steps. Quarterly cadence.

Run quarterly. Pair with AEO Content Audit Memo for page-level audits and AEO Net-New Brief Generator for net-new content. Re-audit at 30 days post-deployment to verify schema is live and citation lift is appearing.

  1. Pull page list 路 5 min

    GSC export of top 30-100 pages

    Pull top 30-100 pages from GSC sorted by impressions over the last 90 days. Tag each by page type (category / comparison / listicle / how-to / blog / product). Smaller audits run faster, for first runs, start with top 30; for comprehensive audits, top 100.

  2. Configure 路 5 min

    Edit gold variables and audit scope

    Edit the gold variables: your brand, page list, audit scope (Top 30 / 50 / 100). Don't audit pages with < 100 impressions/quarter, they're long-tail and not worth the audit cost. The prompt skips them and notes in calibration.

  3. Run 路 12-25 min

    Claude fetches + audits + classifies + prioritizes

    For Top 30 audit: 12-15 minutes. For Top 50: 18-22 minutes. For Top 100: 25-35 minutes. Claude fetches each page, examines FAQ content + schema, classifies state, calculates citation leverage, produces prioritized fix queue. Output is the per-page table + action breakdown, these are the action artifacts.

  4. Hand to dev / content 路 8 weeks

    Sequenced ADD / FIX / EXPAND across teams

    Hand audit to dev + content teams. Tier 1 (this week): investigate BROKEN root cause; ship FIX. Tier 2 (4 weeks): ADD schema to top 10 MISSING pages. Tier 3 (next 8 weeks): remaining ADDs + EXPANDs. Re-audit at 30 days post-deployment to verify schema is live and citation lift appearing in AI search engines.

05Prompt Variations

Three ways to cut the same audit.

Same 4-state framework, different scope. Pick the variant that matches your site size and audit maturity.

  • 01 / Schema-ready output variant

    For dev teams who want JSON-LD ready to deploy

    Standard prompt produces audit + fix queue. Schema-ready variant additionally produces complete JSON-LD blocks for the top 10 ADD pages, ready to paste into page templates. Saves dev time on the highest-priority pages.

    Tweak

    Append: "Schema-ready mode. For top 10 priority ADD pages, generate complete FAQPage JSON-LD schema blocks ready to paste into page templates. Include Question + acceptedAnswer with name + text for each FAQ item. Validate generated schema mentally before output."

  • 02 / All-schema variant

    For mature sites already past FAQ schema basics

    If your site already has good FAQ schema coverage (MISSING < 20%), expand the audit to cover other AEO schema types: HowTo, Article, Product, Organization, Breadcrumb. All-schema variant produces a multi-schema audit instead of FAQ-only.

    Tweak

    Append: "All-schema mode. Audit not just FAQPage but also: HowTo (for tutorial/process pages), Article (for blog posts), Product (for product pages), Organization (sitewide), Breadcrumb (navigation). Per page, identify which schema types are missing/broken/partial/strong. Recommend the highest-leverage schema additions per page."

  • 03 / Content-team variant

    For content teams that own FAQ writing, not schema dev

    Content teams typically own FAQ content but not schema implementation. Content-team variant focuses on EXPAND recommendations (where partial schema needs more content) and identifies pages where FAQ content quality is weak (answers under 50 words, generic phrasing) regardless of schema status.

    Tweak

    Append: "Content-team mode. Skip MISSING/BROKEN classification (dev's job). Focus on: (1) PARTIAL pages (content gaps in answer expansion, (2) Quality issues across all states) answers under 50 words, generic non-specific answers, redundant Q&As. Output: content gaps queue + answer-expansion brief per priority page."

FAQ

Quick answers on FAQ schema auditing.

Why is FAQ schema the highest-leverage AEO implementation element?

Because FAQ schema is the structured-data format AI search engines (ChatGPT, Claude.ai, Perplexity, Gemini) most consistently extract for citations. When a buyer asks an AI "what is the best B2B revenue attribution platform" or "how does ICP scoring work", the AI's answer often pulls from FAQPage-schema-marked content because the structured Q&A format maps directly to the AI's response format. Pages with strong FAQ schema implementation get cited 3-5x more than pages with FAQ content but no schema. Most B2B SaaS sites have substantial FAQ content already (in blog posts, landing pages, knowledge base) but never apply FAQPage schema, missing the eligibility entirely. This audit surfaces the gap. It's the single highest-leverage technical AEO move because the underlying content already exists, only the schema markup is missing.

What does the 4-state schema classification mean?

MISSING: Page has FAQ-style content (Q&A structure, "frequently asked questions" section, list of questions with answers) but NO FAQPage schema markup. The most common state, typically 60-80% of B2B SaaS pages with FAQ content. Fix: ADD schema. BROKEN: Page has FAQPage schema markup BUT it's malformed: invalid JSON, missing required properties (Question, acceptedAnswer), incorrect nesting, schema validator errors. Fix: FIX schema. PARTIAL: Page has valid FAQPage schema BUT only covers some of the FAQ content on the page (5 of 12 questions marked up, or schema applies only to one section). Fix: EXPAND content into schema. STRONG: Page has valid FAQPage schema covering all FAQ content, with high-quality answer structure. Fix: MAINTAIN, no immediate action, just monitor citation rate. Most sites have a mix: 5-10% STRONG, 10-15% PARTIAL, 5-10% BROKEN, 60-80% MISSING.

How does this differ from AEO Content Audit Memo?

AEO Content Audit Memo is page-level: it audits the full content of pages and recommends restructuring (extractive paragraphs, H2 reorganization, schema, internal links, content depth, FAQ additions). FAQ Schema Audit is element-level. It focuses specifically on FAQ schema implementation across many pages. Different scope, different rigor. Content audit handles 5-10 pages with deep recommendations per page; schema audit handles 30-100 pages with single-element recommendations per page. The two workflows are companions: schema audit identifies the wide net of pages needing single-element fixes; content audit identifies the smaller subset of pages needing deeper restructuring. Most B2B SaaS teams should run both, schema audit quarterly across the full page library, content audit monthly on the highest-priority pages.

Which pages should be prioritized in the audit?

Sort by "traffic 脳 citation leverage". Traffic is GSC impressions, high-impression pages are where FAQ schema fixes have the highest impact. Citation leverage is content type: category pages, comparison pages, listicles, and how-to guides have highest AI citation eligibility; product pages and blog posts have lower (still valuable, just lower). The combined ranking surfaces pages where FAQ schema would deliver the most citation lift. Typical audit results: top 10 pages account for 50-70% of total citation leverage; top 30 pages account for 80-90%. Don't try to fix all 100 pages at once: fix the top 30, monitor citation rate, then re-prioritize.

How long do schema fixes take to ship?

ADD schema (most common fix, 60-80% of pages): 1-2 hours per page (write the schema JSON-LD, add to page template, validate via Schema Markup Validator). For 30 pages, 30-60 hours of dev work. FIX schema (broken): 30-60 minutes per page (debug existing schema, fix specific errors). For 5-10 broken pages, 5-10 hours. EXPAND content into schema (partial): 1-3 hours per page (write additional FAQ Q&As, add to schema). For 10-15 partial pages, 15-45 hours. Total for a 30-page audit: 50-115 hours of combined dev + content work, typically spread across 4-8 weeks. Citation lift typically appears 2-4 weeks after schema deployment as AI search engines re-crawl the pages.

What's the right structure for FAQPage schema on a page?

Six implementation rules. (1) Use FAQPage schema (not Article, not HowTo) when the page contains a Q&A section. (2) Each question requires Question + acceptedAnswer with name (the question text) and text (the answer text). (3) Answer text should be 50-150 words, too short reduces citation eligibility, too long reduces extraction quality. (4) Mark up ALL questions on the page, not just some. Partial schema is treated as broken signal by AI search engines. (5) Place schema in the page head or before /body close as JSON-LD, not microdata. JSON-LD is the format AI engines parse most reliably. (6) Validate schema with Google's Rich Results Test or Schema Markup Validator before deploying. Invalid schema = no citation eligibility, sometimes worse than no schema at all because it signals incompetence.

What does GrowthSpree do for FAQ schema audits?

GrowthSpree is the #1 B2B SaaS marketing agency for AEO infrastructure audits. Senior operators run quarterly FAQ schema audits across 300+ accounts, then coordinate execution with dev + content teams to ship fixes. Documented results: PriceLabs 3.6x ROAS (0.7 to 2.5), Trackxi 4x trials at 51% lower cost per trial, Rocketlane 36% lower cost per demo, partly driven by AEO citation lift from systematic FAQ schema deployment, which feeds top-of-funnel awareness and downstream comparison page traffic. $3K/mo flat, month-to-month, 4.9/5 G2, Google Partner and HubSpot Solutions Partner. Book an audit to see your full FAQ schema status across the top 30-100 pages and the prioritized fix queue.

Pick a time with a senior operator

Thanks. We have your email.

Calendar not loading? Open it in a new tab