NamedraHow it works
Namedra works in three steps: audit the real questions buyers ask AI assistants, turn the answers into a score and a fix list, then earn the citations that change the answer. Every number is traceable to a stored raw response.
Step 1: audit the prompts buyers actually ask
We write 30 prompts for your category across six stages of a purchase: discovery ("best compliance automation tools for startups"), evaluation ("is Acme good for fintech?"), comparison ("Acme vs Vanta"), alternatives ("Vanta alternatives"), pricing ("how much does compliance automation cost?") and trust ("which vendors are best reviewed on G2?"). You can edit them. Each prompt runs on ChatGPT, Perplexity, Gemini, Claude and Google AI Overviews, three times, because answers vary between runs. Every response is stored with date, engine, model and run number.
Step 2: score, quote, rank, plan
For each answer we detect whether you and each competitor are named (name, alias or domain), the order of mention, the exact sentence used about each brand, and which URLs are cited. From that we compute mention rate, share of voice, average position and citation rate per engine, and the AI Visibility Score overall. The sources engines cite instead of you are ranked by frequency and type (review site, community, blog, competitor). The report ends with a top-10 fix list, each item tied to the prompts and sources that triggered it.
Step 3: earn the citations
AI assistants recommend what they can cite. So the monthly work is about the sources: comparison pages ("Acme vs Vanta"), alternatives pages ("Vanta alternatives: Acme and five others"), a buyer's guide for the category, fresh reviews on the review sites that were cited, genuine answers in the community threads that were cited, pitches to the authors of the listicles that were cited, consistent entity data (one name, one description, Organization and Product schema, sameAs links), and one original data piece a quarter. Each month we re-run the prompts, diff the results, and send a digest under 200 words.
A typical month on the Growth plan
| Week | What happens |
|---|---|
| 1 | Monthly run completes. Digest and report delivered. 30-minute review call. |
| 2 | Two assets drafted and published (e.g. one comparison page, one alternatives page). Review-site campaign brief sent to your CS team. |
| 3 | Community answers and publisher pitches go out. Entity and schema fixes shipped. |
| 4 | Two more assets published. Prompts list refreshed for next month. |
Engines, and what the APIs do and do not reflect
We query official APIs with web search enabled. They share models and search with the consumer apps, but not their system prompts, memory or personalisation. Perplexity's API is closest to its product. Google AI Overviews are fetched from a real results page and vary by location. We state the model used in every report and keep every raw answer, so you can check any number yourself.