Service · 03
Retainers
Single-Base $1,500 · Core $3,000 · Growth $6,000 · Enterprise from $12,000 per month (scoped to $20,000) · three-month initial term, then month to month (Enterprise: then quarterly)
Ongoing measurement and placement work sized to your segment, run on one shared ledger you can open at any time.
The Citation Ledger
Exhibit
Prompt panel → Logged runs → Gaps → Fixes → Placements → Re-runs
Every engagement runs on one ledger, visible to you in a shared dashboard with the run logs downloadable where provider terms permit. It is a working method, not a trademark; the methodology page explains it in full. How the ledger is kept
Four tiers
| Item | Single-Base · $1,500/mo | Core · $3,000/mo | Growth · $6,000/mo | Enterprise · from $12,000/mo |
|---|---|---|---|---|
| The work | One location fact page built and kept current, three placements attempted each quarter on the directories and lists engines cite for your airport, listing errors corrected monthly. Twenty prompts logged every month, variance shown; a written quarterly report. | Two quotable pages a month, one placement secured or three attempted, entity and listing errors corrected, crawler access checked. Forty prompts logged every month against five competitors, variance shown. | Four quotable pages a month, two placements secured or six attempted, a YouTube asset or trade-press pitch, executive LinkedIn drafts, disclosed Reddit and Quora participation. Eighty prompts logged every month against ten competitors, variance shown. | Eight quotable pages a month, four placements secured or twelve attempted, an engine-correction program across Wikidata, Crunchbase and LinkedIn, a quarterly original-data story for trade press. 150+ prompts logged every month, with region and brand cuts, variance shown. |
| For | One location: an independent FBO at one airport, a single-shop MRO, or an operator flying from one base | Operators under 10 aircraft with more than one base; independent brokers; FBOs and single-shop MROs that want new pages every month | Fleets of 10–40 aircraft; FBO chains; multi-shop MROs; management and sales firms; most software vendors | Top-70 fleets; MRO networks; fuel suppliers; larger software vendors |
| Monthly measurement (API) | 20 prompts × 4 API engines × 5 repeats (400 API runs) vs 3 competitors, with variance | 40 prompts × 4 API engines × 5 repeats (800 API runs) vs 5 competitors, with variance | 80 prompts × 4 API engines × 5 repeats (1,600 API runs) vs 10 competitors, with variance | 150+ prompts × 4 API engines × 5 repeats (3,000+ API runs), region and brand cuts |
| Checked separately | Licensed Google AI Overviews check on your prompts, own column | Licensed Google AI Overviews check on your prompts, own column | Same | Same, with region cuts |
| Consumer apps | Not included | Quarterly hand-run spot check, 10 prompts, across the ChatGPT app, AI Overviews and Copilot — an agreement note, not a ranking. Copilot is measured at no tier: it has no API. | Same | Same |
| Quotable pages | One location fact page, built in month 1 and kept current | 2 per month (new or rebuilt) | 4 per month | 8 per month |
| Placements | 3 attempted per quarter on the directories and lists engines cite for your airport | 1 secured or 3 attempted per month; 1 outreach action weekly | 2 secured or 6 attempted per month; 2 outreach actions weekly | 4 secured or 12 attempted per month; 4 outreach actions weekly |
| Mentions program | — | — | 1 YouTube asset or 1 trade-press pitch per month; 1 executive LinkedIn draft weekly; disclosed Reddit/Quora participation | 2 YouTube assets per month; 2 LinkedIn drafts weekly; quarterly original-data story pitched to trade press; PR-wire release per quarter |
| Entity and accuracy | Up to 5 listing and entity corrections a month; crawler access checked monthly | Entity and trust-facts refresh monthly; crawler access checked monthly | Same, plus Google visibility check | Same, plus engine-correction program (feedback submissions, Wikipedia-eligibility review, Wikidata/Crunchbase/LinkedIn consistency) and an issue log for engine misstatements |
| Cadence | Monthly note (what shipped, hours used vs budget); quarterly report and 30-min call | Monday status note (what shipped, what's next, blockers, hours used vs budget); 60-min monthly review; quarterly panel review and Index position note | Monday status note; 90-min monthly review; quarterly review plus competitor teardown | Weekly 30-min status call; monthly executive briefing deck; quarterly board-ready review |
| Always | Shared dashboard; run logs downloadable (derived data for all engines; verbatim text where the engine's terms allow); named strategist; 24-hour response | Shared dashboard; run logs downloadable (derived data for all engines; verbatim text where the engine's terms allow); named strategist; 24-hour response | Same | Same, plus a named analyst |
| Term | 3 months, then monthly; billed monthly in advance, net-30 | 3 months, then monthly; billed monthly in advance, net-30 | 3 months, then monthly; billed monthly in advance, net-30 | 3 months, then quarterly |
Measured through official APIs OpenAI API + web search · Gemini API + Search grounding · Perplexity Sonar API · Claude API + web search — the models' retrieval-augmented APIs, not the consumer products. How each surface is measured
- Excluded at every tier
- Keyword stuffing, programmatic page volume, llms.txt as a lever, fake or incentivized reviews, Wikipedia editing on your behalf, automated querying of consumer AI apps, and guaranteed rankings or citations. 123
- Add-ons (published)
- YouTube interview production $1,500 per episode · Additional prompt panel (region or brand) $400 per month · Disclosed Reddit/Quora presence program $750 per month.
Questions about retainers
Why a three-month initial term?
Can we start on a retainer without a Pilot?
Yes if you already have a baseline we can audit; otherwise the first month of Core, Growth or Enterprise includes the Diagnostic. On Single-Base the first month's logged panel is the baseline; the Diagnostic is not included.
Do you use your own tracking tool?
We run and log prompts ourselves through official APIs. A licensed tool, named on the methodology page, supplies the Google AI Overviews check for your own prompts; Copilot and the ChatGPT app are covered only by the quarterly hand-run spot check, because we do not query consumer apps by automation. The interpretation and the prompt design are the product; the monitoring is not.