Echo for business · Sample reports
Real reports, in full. Nothing held back.
Commissioned reports are private — unlisted, never published on Echo. These two are the exception: genuine runs on real topics, with no client attached, published so you can judge the work before you buy it.
Beauty · consumer
Are beauty shoppers trading down from prestige brands to dupes, what actually triggers the switch, and which categories are proving resistant?
A prestige founder’s anti-dupe statement is the organising event of the month — and it backfired, handing creators a ready-made premise for side-by-side rebuttals that reached millions of views. The report traces what actually triggers a switch, which categories hold their premium, and why the Chinese 平替 conversation runs on completely different logic to the Western one.
The report
Open in a new tab →Reports are delivered as a private, unlisted web page plus PDF. Download this one as a PDF to see the delivered format.
What to look for
The topic above happens to be beauty. The structure is the same whichever topic you commission.
Every claim is clickable
Numbered receipts throughout — 94 of them. Each one opens the actual post, with its platform, date and engagement count listed in the evidence index at the end.
Confidence is stated, not implied
Each finding carries a high / medium / low rating tied to how many independent sources support it. A section resting on one post says so.
The numbers are counted, not recalled
Volume by week, share of voice, community sizes and every figure in the corpus are computed by counting — so they can be checked, and cannot be invented.
The methodology names what was dropped
Scroll to the end. Sources that were queried but did not contribute enough on-topic evidence are listed with the reason, rather than counted toward the total.
Health · consumer · second sample
Peptide mania
A $1–3bn gray market caught mid-shift by an FDA advisory vote, read across Reddit, TikTok and YouTube alongside WeChat and Xiaohongshu — where the upstream supply chain actually lives. The deeper of the two reports, and the one that shows how the format handles a contested, fast-moving story.
How reports are reviewed
The checks are built into the pipeline, not promised in a brochure.
Every rule below is enforced when the report is produced, and every one of them leaves a trace you can find in the documents above.
Sources earn their place, or they are dropped
A platform is admitted only if at least a quarter of what it returned is on topic, or it contributed at least twelve relevant items outright. Anything below that is cut from the findings and from the coverage count — and published in the appendix with the reason. Having been queried is not the same as having been cited.
The numbers are computed before anything is written
Volume by week, share of voice, community sizes, the stance split and every figure quoted from the corpus are counted mechanically, ahead of the writing step. Where a draft claim and the counts disagree, the counts win and the disagreement is stated in the report.
A broken citation cannot hide
Claims reference evidence by id, and those ids are resolved to real links when the report is rendered. An id that fails to resolve prints a visible “unresolved” marker in the delivered document rather than quietly vanishing — so a missing receipt shows up as a defect you can see, not a sentence that reads fine.
Confidence is tied to evidence, not to how good the story is
Each finding is rated high, medium or low against the number of independent items supporting it. A section resting on a single post is rated low however striking that post is.
Inference is labelled as inference
Reports do reason past what a post literally says — that is most of the value. Where they do, it is marked as inference rather than presented as something observed.
The report is required to name its own gaps
“What this report could not establish” is a mandatory section. If the evidence did not settle a question, the report says which question and why, instead of writing around it.
Your topic, same treatment.
$149 for one report, delivered within 3 business days. If it misses the brief we revise it; if it still misses, you don't pay.