The measurement layer behind our consulting work is becoming a product.
Sentiment across AI answers, brand reputation risk, a comparative LLM visibility score and lead scoring — instrumentation we currently run inside engagements, packaged so a team can run it themselves.
We are opening a limited beta. Places are capped, because early access includes direct access to us.
Join the waitlist →Four instruments, one evidence base
Each module answers a question that conventional analytics cannot: not whether someone arrived on your site, but what an AI system told them beforehand — and whether that representation is moving for or against you.
Sentiment Analysis
Multi-axis rather than a single polarity score: evaluative tone, confidence, commercial framing and brand voice can all point in different directions on the very same mention.
In the betaBrand Reputation Risk
Systematic detection of the failure modes inside AI answers — hallucinated claims, negated mentions, phantom citations and sentiment drift — scored by exposure, not anecdote.
In the betaLLM Benchmark Visibility Score
Where you stand against a defined peer set across models, prompts and buyer situations — one comparable score with the underlying evidence attached, tracked over time.
In the betaLead Scoring
Which signals actually precede qualified pipeline, and which only correlate with traffic — scoring built on observed outcomes rather than assumed intent.
In the betaEarly, and therefore unfinished
- Free during the beta. No payment, no commitment, no contract while the product is in testing.
- You shape what gets built. The order in which modules mature will follow what beta participants actually need.
- Rough edges are the deal. You get access before the product is polished — in exchange for telling us where it breaks.
- Invitations go out in waves. Joining the list is not an automatic place; we match invitations to the situations each module is ready for.
Where the method is already documented
None of this starts from scratch. The concepts the product measures are defined publicly in the Compendium, the reputation side is laid out in the AI Brand Reputation Risk guide, and the inference standards we hold ourselves to are on the methodology page. Read those first if you want to judge whether the measurement is sound before you sign up.
Join the waitlist
Tell us where you sit and which module matters most to you. We use that to decide who gets invited in which wave — nothing else.