Tools tell you you're invisible.
We fix it.
Profound, Peec, Otterly, Scrunch — every AI visibility tool reaches the same conclusion: they diagnose and do not fix. They surface a citation rate and leave the content, schema, entity work, and authority-building to someone else.
The AI Citation Programme is that someone else. A named product, a published methodology, a written performance target, and a team that does the work. From $2,500/month with no lock-in.
What Tools Give You
- ✗Your citation rate is 12%
- ✗Competitors average 34%
- ✗You appear on 4 of 60 prompts
- ✗Suggested fix: improve your content
What We Give You
- ✓Your citation rate is 12% → target 28%
- ✓6 content assets published this quarter
- ✓3 editorial citations placed this month
- ✓Schema fixed, llms.txt configured, entity clean
Client Zero
We ran it on ourselves first
Before selling this programme, we ran it on Reveal Rank itself — 60 prompts, 6 engines, 90 days. Here is what moved.
Lighter bar = baseline. Solid bar = 90 days. Citation rate = % of 60 prompts returning brand in top-3 citations, 3-run average.
6 long-form assets
2 per month covering category, comparison, and problem-led prompts
11 editorial citations
Placed across industry publications that Perplexity and Gemini index heavily
Entity + schema overhaul
Organization schema, llms.txt, Knowledge Graph corrections, NAP consistency
Transparent Pricing
Three tiers, each converting into the next
No lock-in on any retainer. The $997 audit is credited in full against month one.
One-time
AI Visibility Audit
Credited in full against month one of any retainer
- 60 prompts across 6 AI engines
- 3 named competitors benchmarked
- Baseline citation rate report
- 60-minute recorded strategy readout
- Prioritised fix list with effort estimates
- Methodology appendix (prompt set v1)
Monthly retainer
Citation Programme Core
No lock-in. Cancel any time.
- 60 tracked prompts, 6 engines, 3 competitors
- Monthly citation-rate report (section 4 format)
- 2 citable long-form content assets per month
- Schema and entity optimisation
- llms.txt and AI-crawler technical configuration
- 4 editorial citation placements per quarter
- Written citation-rate target at month one
- 6-month performance review
Monthly retainer
Citation Programme Scale
No lock-in. Multi-location and multi-product.
- 150 tracked prompts, 6 engines, 5 competitors
- Multi-location or multi-product prompt sets
- Quarterly competitive deep dive
- Sentiment and framing analysis
- PR-led citation campaigns
- Knowledge Graph and Wikipedia-adjacent entity work
- Everything in Core, expanded
Enterprise / Tier 3: Multi-brand, multi-market, regulated verticals. Custom scoped on prompt volume and execution load. Get in touch →
Published Methodology
The prompt set is the intellectual property. We publish it openly.
Every GEO tool runs prompts. The value is in which prompts, held stable over time. A drifting prompt set produces noise and destroys the trend line — which is the one thing clients pay for.
We publish the methodology because transparency is the differentiator against agencies making vague GEO claims — and because it is itself citable content. You should be able to evaluate the rigour of our measurement before you spend a dollar.
Fixed denominator
Same 60 prompts, same wording, same locale every cycle. Changes are versioned and annotated, never silently swapped.
Consumer interface, not the API
Claude.ai is not the Claude API. Gemini chat is not AI Overviews. We test the surface a real buyer uses.
Repeated sampling
Three runs per prompt per cycle, reported as a rate — never a single observation.
Locale locked
Prompts run from a fixed US location per client, noted in the methodology appendix of every report.
Engines tracked
Grok and Meta AI appear as commentary in reports, not tracked metrics, until coverage is reliable enough to sell.
60-Prompt Set Composition
| Category | Count | Example Shape |
|---|---|---|
| Category-defining | 12 | "best [service] companies for [segment]" |
| Local intent | 12 | "best [service] in [city]" — one per priority metro |
| Comparison | 10 | "[client] vs [competitor]", "alternatives to [competitor]" |
| Problem-led | 12 | "how do I fix [problem the client solves]" |
| Buying criteria | 8 | "what should I look for when choosing a [service]" |
| Branded | 6 | "what is [client]", "is [client] any good" |
| Total | 60 | Fixed at onboarding. Versioned on change. |
Why branded prompts matter most
“What is [client]” and “Is [client] any good” reveal how AI models describe your brand — where reputational risk lives and where nobody else is looking. Factual errors in AI-generated brand descriptions are common, often traced to an old press mention, and invisible until you specifically test for them.
The Deliverable
Monthly report — same shape every time
Consistency is the product. A client should be able to flip through 12 months of reports and read the trend without a guide.
The Number
Citation rate this cycle, last cycle, delta. One figure, top of page one.
Share of Voice
Client vs. 3–5 named competitors, per engine. The slide that renews contracts.
Movement
Prompts newly won, lost, never won. The 'never won' list is the roadmap.
Framing
How models describe the brand verbatim. We flag factual errors — clients rarely know they exist.
Source Analysis
Which domains the models cite for these prompts. The outreach target list.
Work Done / Next
What shipped this cycle, what ships next — tied to specific prompts.
Methodology Appendix
Prompt version, engines, sample size, locale, run dates. Every cycle.
Sample Report
See a real report before you buy
An anonymised Core-tier report showing all 7 sections, the citation-rate trend chart, share-of-voice breakdown, and the framing analysis section.
📄 Request Sample ReportSent within 1 business day. No sales call required.
Honest about what this is — and isn't
This is not a dashboard
It is a commitment to move a number on someone else's dashboard. Use whatever tracking tool you prefer.
Citation rate ≠ revenue
We are careful about implying it is. Sophisticated buyers will press on attribution — as they should.
Measurement has variance
AI answers vary by run, model update, and interface. We manage it with methodology. We do not eliminate it.
Training data has lag
ChatGPT and Claude base-mode improvements take 3–6 months to surface. We track leading indicators while we wait.
FAQ
Questions before you commit
How is this different from buying Profound, Peec, or Otterly?
Those are measurement tools. They tell you your citation rate. They do not write content, build links, fix your schema, clean up your Knowledge Graph entries, or place editorial citations. Several of them explicitly say their tool only works if an agency or internal team does the optimisation work. We are that team. You can keep your existing tracking tool and add us as the execution layer, or let us handle both.
What if citation rates don't improve?
We state a target citation-rate lift in writing at month one, reviewed formally at month six. If we're significantly off-track, we either agree a revised plan or you have grounds to leave — no lock-in enforces you to stay. We're structured this way because the risk of doing nothing is real and growing, and we're confident in the execution.
How long until I see movement?
Platforms that search the live web in real time — Perplexity, Google AI Overviews, Microsoft Copilot — can show citation improvements within 4–8 weeks of content and authority work. Platforms that rely on training data — ChatGPT, Claude in base mode — update on training cycles, typically 3–6 months. We track both separately and set expectations per engine from day one.
Why is measurement unstable? Doesn't that undermine the whole thing?
AI answers vary between runs, model updates, and interfaces. This is a real limitation — and every agency that doesn't acknowledge it is misleading you. Our methodology manages it: three runs per prompt per cycle, reported as a rate not a single observation, same prompt wording locked at onboarding, same consumer interface (not the API), same locale. The trend line over 3–6 months is reliable. A single month's number is a data point, not a verdict.
Can you work alongside our existing SEO agency?
Yes. The AI Citation Programme is additive. We can run the full execution, or scope to the AI-specific work (content, schema, entity, citations) while your existing agency handles traditional SEO. We define the handoff clearly at onboarding so there's no overlap or gap.
Do you work in regulated verticals — legal, healthcare, finance?
Yes, under the Enterprise tier. Regulated verticals require additional compliance review on content and claims, and often need deeper entity and Knowledge Graph work because AI models are cautious about recommending unverified providers. Pricing is scoped to the compliance load. Contact us to discuss.
Ready to move the number?
Start with the $997 AI Visibility Audit — a full baseline report, 60-minute recorded readout, and prioritised fix list. Credited in full against month one of any retainer.
No credit card required · Free 30-min strategy session · Response within 24hrs