Service
We measure share-of-voice in LLM answers the way SEO teams measure rankings—then ship the fixes that move the needle.
Fifteen prompts, four engines, five business days.
Start with a visibility checkDiagnosis
Traffic from assistants does not appear as branded search, so teams discover they are invisible only when a deal asks “why weren’t you recommended?” Rank trackers never flagged the loss.
Approach
We maintain a commercial prompt library with named / cited / absent labels per engine. Work prioritizes the prompts where competitors appear and you do not, then ships the smallest page or proof change that could change the next answer.
Process
Days 1–5
Baseline
Approve prompts, engines, and competitors. Record today’s answers — labelled baseline, not a forecast.
Week 2+
Ship
Publish answer blocks, schema, internal links, and disclosed mentions. Strategists approve customer-facing copy.
Weekly
Report
Ship log plus named / cited / absent on the agreed set. Next sprint targets the largest gap.
Deliverables
Baseline and weekly prompt library with sentiment and rank
Competitor citation share for priority questions
Action queue tied to content, technical, and authority work
Included work
Custom prompt sets by funnel stage and persona
Included work
Engine coverage: ChatGPT, Perplexity, Gemini, Claude, Copilot, AI Overviews
Included work
Alerting when competitors gain citations on core prompts
Sample artefact
A strategist-readable table tied to your prompt list. We do not paste invented win rates into this layout.
Sample weekly excerpt · layout only
LLM visibility optimization
| Prompt | Status | Engines |
|---|---|---|
| Best vendor for [job]? | Cited | |
| Who is trusted in [category]? | Named | Sample row |
Figures appear only when measured for your account. This is not a client export.
Measurement
Prompt status
Named, cited, or absent on each agreed buyer question.
What shipped
Pages, schema, links, and disclosed mentions that week.
Sources cited
Which URLs or domains the assistant attributed when it answered.
Next sprint
The smallest change queued to address the largest gap.
Engine angle
LLM answers differ by engine update cadence. We label each prompt named, cited, or absent per run so a Perplexity change does not hide a ChatGPT win.
Strong fit
Usually not a fit
FAQ
Start with a complimentary AI Visibility Check, or commission a $997 audit if you already know this is a priority.