Context: Since Google AI Overviews deployed at scale in France (Q4 2025), brands cited in AIO responses capture a growing share of informational traffic. At the expense of classic organic results. According to Search Engine Land (2025), CTR on results below an AI Overview drops 15-42% depending on position. A classic SEO audit doesn't detect this problem. This guide explains how to audit it, and fix it.
Here's something I observe regularly on French brand websites: decent Google rankings, flat traffic, and zero presence in AI responses. The technical audit reveals nothing alarming. Backlinks are acceptable. Content exists. And yet, competitors start capturing the new traffic sources. AI Overviews, ChatGPT answers. While the site remains invisible on these channels.
This guide structures a method across 9 dimensions covering both classic SEO and AI visibility. You'll find a complete evaluation scorecard, performance thresholds, and examples of corrections that have produced measurable results on French brand sites.
Why audit SEO AND GEO in 2026, not just SEO
A SEO+GEO audit simultaneously analyzes your positioning in classic Google results AND in AI-generated responses (ChatGPT, Google AI Overviews). The two channels have distinct evaluation mechanics. Ignoring one means leaving 20-30% of traffic opportunities undiagnosed.
Dimension 1, Technical audit: the foundations that determine everything else
The technical audit evaluates whether Google can crawl, understand and index your site correctly. It's the most basic dimension. But errors here block the effectiveness of everything else.
Most common technical SEO issues on French brand sites, in order of frequency:
- Duplicate or missing title tags and meta descriptions, present on 40-60% of audited sites.
- Orphaned pages, not linked from any other page on the site. Google discovers them only via sitemap but their internal PageRank is zero.
- Chained redirects, sequences of 3, 4 or even 5 redirects. Each hop dilutes link juice and slows crawling.
- Pages canonicalized to the wrong URL, a particularly destructive signal: you tell Google « index this page » but your canonical tag points elsewhere.
- JavaScript-injected content not rendered, if your main content (H1, body text, JSON-LD schemas) is injected by JavaScript after load, Google may not index it correctly.
- Inconsistent sitemap.xml, URL list that doesn't match actually available pages (404s included, robots.txt-blocked pages in the sitemap).
| Metric | OK threshold | Critical threshold |
|---|---|---|
| Indexed / crawled pages | > 85% | < 70% |
| 4xx errors in GSC | < 1% of pages | > 5% |
| Chained redirects | 0 | > 5 URLs |
| Duplicate title tags | < 2% | > 8% |
| Missing meta descriptions | < 5% | > 20% |
Dimension 2, Google coverage (GSC)
Search Console reveals the gaps between pages you believe are visible and those Google actually indexes. This is often where unexplained traffic losses hide.
Key GSC data to analyze: coverage report (valid / with warnings / excluded / with errors), queries with impressions > 100 and CTR < 2% (high-ROI quick wins), pages at positions 4-15 on volume queries (push-content opportunities), and the discovery vs. indexed ratio.
Concrete example: On a 1,200-page e-commerce site audited in March 2026, 340 pages were « excluded » from the index. Including 120 with original, unique content. Root cause: canonical tags pointing to sort variants (color, size) instead of category pages. Fixed in 3 days, +18% impressions 8 weeks later.
Dimension 3, AI visibility: are you cited by ChatGPT and Google AI Overviews?
AI visibility measures whether your brand, content or URLs appear in responses generated by ChatGPT and Google AI Overviews. This is the newest dimension, and the most undervalued in current audits.
For Google AI Overviews: enable AIO in Google Search settings, test your 10-20 target queries, note which trigger an AIO, and check whether your domain is cited among the sources.
Signals that favor citation in AIO based on analysis of cited pages:
- Direct answers at the top of sections (LLMs extract the first sentences)
- Valid JSON-LD FAQPage and HowTo schemas
- Numbered or bulleted structured lists (more extractible than prose paragraphs)
- Sourced statistical data (AIO more readily cites content with verifiable stats)
- Domain authority on the topic (multiple pages on same theme > one isolated page)
For ChatGPT: with web browsing enabled, ask questions your prospects actually ask. Note if your domain appears in responses or cited sources. ChatGPT preferentially cites sources with: strong quality backlink profile, presence on Trustpilot/G2/Capterra, content structured with clear definitions, thematic consistency of the domain.
Does your brand appear in ChatGPT and Google AI Overviews? Get a personalized AI visibility diagnostic in 48h.
Request my GEO auditDimension 4, Content gaps
The content audit identifies high-traffic queries in your sector where your site has no page, and existing pages that cover their subject insufficiently to rank.
4-step method: (1) List your 20-30 priority business queries. (2) Crawl top 10 competitors, who ranks where, what format, what depth? (3) Compare with your existing pages. (4) Prioritize by ROI: query volume × commercial intent × KD × production time.
Dimension 5, Schema.org: structured data that triggers rich results and AI citations
Schema.org allows search engines and AI generative tools to directly extract structured information from your pages. Without it, even the best content remains less accessible to extraction bots.
Priority schemas for French brands: Organization (homepage + About page), Article/BlogPosting (every blog post), FAQPage (question-answer content), Person (author schema. Critical for E-E-A-T), BreadcrumbList (improves CTR), and HowTo/Product/LocalBusiness depending on sector.
Audit tool: paste your key page URLs into the Google Rich Results Test. Critical errors (missing required properties) block rich results.
Dimension 6, E-E-A-T: does your site deserve to be trusted?
E-E-A-T (Experience, Expertise, Authority, Trustworthiness) is Google's qualitative evaluation framework. And implicitly that of AI generative engines. To decide whether your content deserves prominence. It's not a direct ranking score, but its components directly influence rankings.
E-E-A-T audit checklist:
- Identifiable author on each article, name, bio, photo, LinkedIn link, Person schema in JSON-LD. Absent on more than 70% of audited sites.
- Substantial About page, who are you? What's your experience? What evidence of results?
- Cited sources in content, every factual claim should point to a verifiable source: Google Search Central, published study, sector report.
- Verifiable external reviews, Trustpilot, Google Business Profile, G2, Capterra depending on sector.
- Information accuracy, up-to-date content, visible modification dates, correction of factual errors.
Dimension 7, Core Web Vitals
Core Web Vitals (CWV) measure the user experience quality perceived on your site. Since 2021, they're an explicit ranking factor for Google. In 2026, INP (Interaction to Next Paint) replaced FID as the interactivity metric.
| Metric | Measures | « Good » threshold | Critical |
|---|---|---|---|
| LCP | Largest Contentful Paint load time | < 2.5s | > 4s |
| INP | Responsiveness to user interactions | < 200ms | > 500ms |
| CLS | Visual stability (shifting elements) | < 0.1 | > 0.25 |
Dimension 8, Backlink profile
The backlink profile reflects the trust other sites grant yours. Quantity, quality, diversity and anchor text are the 4 audit axes. A poorly built profile can actively harm rankings.
Positive signals: diversity of referring domains, thematic relevance, natural anchor text mix. Negative signals: mass links from non-thematic domains, abnormally fast growth, links from penalized sites, over-optimized anchor text (>20% exact-match).
An undervalued risk in 2026: automatic link exchanges between users of the same platform. This model, sold as « automatic netlinking » by some tools. Corresponds to what Google calls a « link scheme » in its spam policies. Sites involved risk algorithmic or manual penalties.
Dimension 9, Competitive benchmark
The competitive benchmark compares your SEO+GEO performance with 2-3 direct competitors on your target queries. It reveals exploitable gaps and strengths to preserve.
| Dimension | You | Competitor A | Competitor B |
|---|---|---|---|
| Indexed pages (GSC) | To complete | Estimated via SERP | Estimated via SERP |
| Top 10 on target queries | To measure | To measure | To measure |
| Active JSON-LD schemas | To audit | To audit | To audit |
| Presence in AIO | To test | To test | To test |
| Mobile LCP | To measure | To measure | To measure |
Complete evaluation scorecard
The evaluation scorecard translates the 9 dimensions into a score out of 100. It allows prioritizing actions and measuring evolution between two audits.
| Dimension | Weight | Score (0-10) | Weighted score |
|---|---|---|---|
| Technical (crawl, indexation, redirects) | 15 pts | X / 10 | X × 1.5 |
| Google coverage (GSC, impressions, CTR) | 15 pts | X / 10 | X × 1.5 |
| AI visibility (ChatGPT, AIO) | 15 pts | X / 10 | X × 1.5 |
| Content gaps | 10 pts | X / 10 | X × 1.0 |
| Schema.org | 10 pts | X / 10 | X × 1.0 |
| E-E-A-T | 15 pts | X / 10 | X × 1.5 |
| Core Web Vitals | 10 pts | X / 10 | X × 1.0 |
| Backlinks | 5 pts | X / 10 | X × 0.5 |
| Competitive benchmark | 5 pts | X / 10 | X × 0.5 |
Score interpretation: 80-100 = healthy profile, detail optimizations; 60-79 = significant improvements available (quick wins in 4-8 weeks); 40-59 = structural issues to fix before any content production; 0-39 = deep SEO rebuild needed.
Cicéro methodology: how to prioritize after the audit
The audit's value lies in the prioritized recommendations it generates. Cicéro's methodology organizes corrections across 3 horizons: quick wins (0-30 days), foundations (30-90 days), growth (90-180 days).
Horizon 1, Quick wins (0-30 days): Fix chained redirects and incorrect canonicals. Optimize title + meta description on pages with high impressions but low CTR in GSC. Add FAQPage schema on the 5-10 most important pages. Submit priority pages to Google indexing. Add/correct authors on articles.
Horizon 2, Foundations (30-90 days): Improve mobile Core Web Vitals. Enrich thin-content pages. Build or strengthen the About page and author pages. Implement coherent internal linking (semantic cocoon). Fix critical schema.org errors.
Horizon 3, Growth (90-180 days): Content production on identified gaps (priority to commercial queries with medium KD). GEO strategy: structure content for AIO and ChatGPT. Build editorial backlinks. Monthly position + AI visibility monitoring.
On our clients where these 3 horizons were respected, the median impact at 6 months is 60-100% increase in organic traffic on target queries. Not from an « SEO trick, » but because the method systematically addresses what specifically holds each site back.
What this audit doesn't cover
- Local SEO: if your activity is geolocated (restaurant, medical practice, retail), the audit must include an additional dimension: Google Business Profile, NAP citations, local reviews.
- E-commerce SEO: sites with thousands of product/category pages have specific issues (pagination, facets, thin content at scale) that warrant dedicated treatment.
- Perplexity or Gemini visibility: this guide focuses on ChatGPT and Google AIO. The two AI channels with significant volume in France. Other AI engines exist but their French market share remains marginal for most sectors.
- Guaranteed results on precise timelines: the timelines given (6-16 weeks) are observed medians. A site with an active Google penalty or an ultra-competitive sector may require longer timelines.
Go further:
Growth and SEO content strategist, I founded Cicéro to help businesses build lasting organic visibility — on Google and in AI-generated answers alike. Every piece of content we produce is designed to convert, not just to exist.
LinkedInFAQ, SEO + GEO Audit
What is a SEO GEO audit for a French brand?
A SEO GEO audit analyzes a site's visibility on Google (organic search) AND in AI-generated responses like ChatGPT and Google AI Overviews. For a French brand, it covers 9 dimensions: technical, Google coverage, AI visibility, content gaps, schema.org, E-E-A-T, Core Web Vitals, backlinks, and competitive benchmarking.
What's the difference between a standard SEO audit and a SEO+GEO audit?
A standard SEO audit focuses on Google organic visibility: technical, backlinks, content, keywords. A SEO+GEO audit adds the AI dimension: is your brand cited in ChatGPT? Do you appear in Google AI Overviews? These new traffic sources already account for 15-28% of visits in some French B2B sectors in 2026.
How much does a complete SEO GEO audit cost?
A complete SEO GEO audit for a French brand typically costs between €800 and €3,000 depending on site size and depth required. Cicéro offers a free initial diagnostic to identify priorities before scoping a full audit.
How long does it take to see results?
Technical fixes produce visible effects in Search Console within 2-6 weeks. Content and E-E-A-T improvements show results in 6-16 weeks. Visibility in AI generative engines can improve faster if content is well-structured with FAQPage schema.
What are the 9 dimensions of a complete SEO GEO audit?
Technical, Google coverage (GSC), AI visibility (ChatGPT/AIO), content gaps, schema.org, E-E-A-T, Core Web Vitals, backlink profile, competitive benchmark. Each is scored 0-10 and weighted to produce a total score out of 100.