Question 1

What queries are being tested?

Accepted Answer

Eight mid-funnel buyer questions across consulting categories: ERP-NA, AI readiness, Power BI, fractional CTO, DevOps maturity, Azure cost, D365 F&O, SaaS architecture audit. The exact query text is on the page next to each category card. We picked queries where Garnet plausibly competes, but the firms being measured are not Garnet — this is a competitive map for the categories themselves.

Question 2

How often does the data refresh?

Accepted Answer

Daily at 22:30 local. Same pollLane.js module that runs for paid GEO subscribers fires the 8 queries × 4 engines (ChatGPT / Claude / Perplexity / Gemini) and writes a per-day snapshot. The page aggregates the last 30 days.

Question 3

What does 'cross-engine consensus' mean?

Accepted Answer

A domain counts as consensus when ≥1 of the 4 engines cited it for that category over the 30-day window. The 'engine pills' next to each domain show which engines specifically cited it. Domains cited by 3-4 engines are the truly dominant brands in that category — those are the ones a buyer's assistant points at regardless of which engine they're using.

Question 4

Is this a ranking?

Accepted Answer

No. Engines disagree, indexes shift week-to-week, and single-day snapshots have noise. The 30-day rolling smooths some of that but it's still a measurement, not an authority. Treat it as a competitive map: 'these are the domains showing up; here's how the four engines weight them differently.'

Question 5

Can my firm get included?

Accepted Answer

Not in this dataset — these are public buyer queries against the open engine indexes, no firm-level filter. To measure your own domain specifically (with full buyer-question set, per-query attribution, and 30-day citation-share over time), the paid GEO subscription runs that on a 12-query × 4-engine × 30-day cadence with monthly PDF.