audit report · sample format · data from the real 2 July 2026 run
What the report you get looks like.
Every number below comes from the real 2 July 2026run, so nothing here is invented. Your audit swaps in your firm, your buyers’ questions, and names names.
1 · The question set
One question in this sample. Your audit runs ten, phrased the way your buyers phrase them.
Q1 · “who's the best interior designer in singapore?” · 22 runs · Mitra engine
2 · Who got named, how often
most-cited firm 1 (named in your report)10/22 · 45%
most-cited firm 2 (named in your report)10/22 · 45%
most-cited firm 3 (named in your report)9/22 · 41%
most-cited firm 4 (named in your report)9/22 · 41%
most-cited firm 5 (named in your report)7/22 · 32%
most-cited firm 6 (named in your report)6/22 · 27%
distinct firms named121
named exactly once105
3 · The finding
Concentration. Four firms sat at the top of almost every answer; 105firms surfaced once and vanished. An answer engine doesn’t spread enquiries around. It routes them to whoever it already trusts.
4 · What your audit adds to this
- Your firm tracked by name across every run, unredacted
- 48 buyer questions instead of one, sampled to 240 answers
- The sources each engine leaned on when it named firms
- The gap list: what we'd fix first and why, in order
How this was measured
- engine
- Mitra engine
- date
- 2 July 2026
- runs
- 22
- counting
- a firm counts once per answer
- raw transcripts
- all 22 published at /real-sample
The free audit produces this exact document for your firm: request it →