Why measuring AI citations is no longer optional
Without measurement, GEO is faith. With measurement, it's an acquisition channel as steerable as any other. Three concrete reasons: (1) prioritise actions — you'll know which engine cites you least and why; (2) defend the GEO budget to leadership with hard KPIs; (3) detect regressions — a model swap (GPT-4 to GPT-5) can wipe your citations overnight.
The typical NEXUS GEO client ratio: 70% of the GEO budget goes to editorial and technical production, 30% to measurement. That's the ratio that gives you solid pilot ability.
Free tools — where to start
Google Search Console — AI Overviews tab (since 2025)
Google added a dedicated AI Overviews report inside Search Console in late 2025. You see the queries where your site has been cited inside an AI Overview, click rate and impressions. Free, official, the first place to look. Limit: covers Google AI Overviews only — not ChatGPT, Claude, Gemini chat or Perplexity.
Bing Webmaster Tools — Copilot tab
Microsoft Bing Webmaster has offered a Copilot tab since 2024 that surfaces citations of your site inside Copilot answers. Since ChatGPT in browsing mode relies on Bing, it's a useful proxy for estimating your ChatGPT visibility. Free, activate it today.
Manual prompt tests
The most basic approach but ruthlessly effective: list 20 to 30 prompts your prospects would type ("best agency X in Europe", "solution Y for SMBs"). Test them every month in ChatGPT, Claude, Gemini and Perplexity. Note: (1) are you cited? (2) at what position? (3) which competitors are cited instead? Cost: €0, time: 2 hours per month.
Paid tools — when to level up
Profound (tryprofound.com) — the US reference
Profound, founded in 2023, is the US leader of GEO measurement tools. Multi-engine tracking (ChatGPT, Claude, Gemini, Perplexity), automated competitor benchmarking, citation change alerts. Pricing: $999/mo entry tier (August 2025). Fit for brands with 50+ tracked prompts and a GEO budget above €5,000/mo.
Goodie AI
Direct competitor to Profound, similar positioning. Goodie AI leans on the depth of semantic analysis of AI answers (who's cited, on what angle, with what sentiment). Pricing: $600 to $1,200/mo depending on volume. Good alternative if you want analytical depth over prompt volume.
AthenaHQ
More recent, AthenaHQ focuses on the AI share-of-voice dimension and provides clean dashboards for marketing leaders. Announced pricing: $750/mo entry tier. Interesting for brands needing a visual monthly deliverable for their leadership.
The DIY approach: in-house tracker with D1 + Gemini API
This is the approach NEXUS GEO built in-house in its /pilote department. Simple, free, scalable stack: (1) a Cloudflare D1 database storing target prompts and citation history; (2) a cron that queries Gemini API and OpenAI API every month on the prompts; (3) a parser that detects whether the client brand is cited and at what position; (4) an internal dashboard plotting the trend.
Monthly cost: €5 to €20 in API calls per client (Gemini Flash at $0.000075/1k tokens, negligible). Advantage over Profound: full control of the prompt scope, ability to integrate models not covered by US tools (Mistral, Le Chat), and marginal cost close to zero. Downside: initial dev time (2 to 3 weeks).
The 3 metrics that actually matter
1. Citation rate
Percentage of prompts in the panel where your brand is cited at least once. The simplest and most telling metric. Realistic 90-day objective for a starting French B2B brand: move from 5-10% to 30-40%.
2. Average position in the answer
When you're cited, in what position? First brand mentioned, or third out of five? Position 1-2 captures most of the attention. Objective: reach position 1-2 on at least 50% of the prompts where you're cited.
3. AI share of voice
Across all brands cited in your prompt panel, what share do you represent? If you get 40 citations and your 3 direct competitors total 120, your AI share of voice is 25%. The reference metric for steering against competition.
Monthly test methodology (45 minutes)
- Define 30 representative industry prompts (10 category A: direct search; 10 B: comparatives; 10 C: indirect recommendations).
- Run the 30 prompts in ChatGPT, Claude, Gemini and Perplexity (120 queries total). 30 minutes manually, or 5 minutes via an automated tracker.
- For each query, log: brand cited Y/N, position if yes, competitors cited, source URLs cited by the AI.
- Compile in a sheet or dashboard: citation rate per engine, average position, share of voice, top external sources mentioned.
- Compare to the previous month. Any variation > 15% deserves a causal analysis (did we ship an action? did a competitor publish?).
This is the methodology [NEXUS GEO applies for its clients](/methodologie) with an automated monthly report. The €1,750 initial audit includes the first full benchmark on 30 prompts.
FAQ
Sources
- Google Search Central — AI Overviews documentation inside Search Console, 2025.
- Microsoft Bing Webmaster — Copilot tab documentation, 2024-2025.
- Profound — Pricing and documentation, 2025 (tryprofound.com).
- Goodie AI — Official site and 2025 pricing.
- AthenaHQ — Official site 2025.
- Aggarwal P. et al. — "GEO: Generative Engine Optimization" — Princeton CS, 2024.
- NEXUS GEO internal tracker — code and methodology to be open-sourced spring 2026.
Ready to see what these levers do for your brand? [Book a NEXUS GEO audit](/audit/new) — 10 days, no commitment.
Want a 30-prompt benchmark of your brand?
The NEXUS GEO €1,750 (VAT included) audit includes the initial monthly benchmark across ChatGPT, Claude, Gemini and Perplexity. Shipped in 10 days, refunded if not profitable.
