>
Eleven free instruments that read public behaviour and report their own limits. Each one names what it proves — and refuses when the data cannot carry the claim.
Eleven free instruments, in three moves. Read what an audience already feels, find where a campaign rubs, then make the call — before the market makes it for you. See the suite as a system — problem → use case → output →
Sentiment, emotion and momentum from public comments, read in seconds.
Open instrument → Echo 03The distance between how a brand speaks and how it is actually heard.
Open instrument → Echo 05What rose, what fell and what changed tone between two snapshots.
Open instrument → Echo 06The moment a conversation stops weighing a thing and starts moving on it.
Open instrument → Echo 09What the whole audience can believe, when only the vocal few spoke.
Open instrument → Echo 10Thousands who decided alone, or a few who staged it. The trace looks the same.
Open instrument → VectorWhich frame is winning, who carries it, and where it drifts.
Open instrument →A working miniature of the suite, running on this page. Every number below is computed from your own text and shows the words it counted — nothing is invented, nothing is uploaded.
Not one number dressed as certainty. The width is the finding, and it is printed first.
Every verdict is paired with what it does not establish, written before the result.
When your stated assumptions contradict the data, no estimate is produced at all.
Everything runs in the browser. No account, no upload, nothing sent anywhere.
What the trace can prove: the interval of worlds still compatible with it. What it cannot: which of the two produced it. That is why every instrument here reports an interval — never a single number.
Each tool is the operational half of a written argument. Every paper is open access and states its own conditions of failure.
The theory every other paper here extends. A shout made of a thousand voices cannot be taken apart: from an observed trace alone, coordinated amplification and genuine consensus are not distinguishable. Six propositions on what observational data can and cannot reveal about who is really speaking — and the reason every instrument in this suite reports an interval instead of a number.
Why observational uniqueness does not establish person-specific structure. Nearly-full uniqueness in a database is compatible with zero real individuality — the percentage is not a weak signal, it is not a signal. Open the three in-browser demonstrations, in English or Romanian.
Action address as a missing variable in digital brand analytics: the same click can be addressed to the brand, to the platform, or to the feed itself — and standard funnels cannot tell which. What that costs, and what would identify it.
Brand authenticity as a cost inference, not a property consumers detect. If generative AI drives the cost of the carrying signal toward zero, which signals still separate an authentic sender from an inauthentic one? The paper derives that single condition — the signal floor — for identity goods.
The costed protocol for the signal-floor account: what an honest test would actually require, priced out, so the claim can be falsified rather than merely argued. Supplements The Story Comes Cheap.
Authentic work can carry a provenance penalty with no lie anywhere in the chain. Names the cost — the ambient provenance discount — the asset it depletes, and the only defence that survives: external corroboration, not what the image shows.
A euro withheld from someone who would have bought anyway is worth roughly ten of new revenue. An operator’s guide that routes promotions on customer and stock state together, and puts price reduction last behind four cheaper levers.
Follows one question — who is really speaking? — down the length of a listening pipeline. Collection, deduplication, classification, aggregation: each stage quietly discards some of the evidence that would separate coordinated amplification from genuine consensus, until a dashboard reports an agreement the raw trace never established.
A dashboard reports 68% positive. That is a fact about the people who spoke, read aloud as a fact about everyone. This paper bounds what the silent majority can actually believe: sharp assumption-free intervals, a one-parameter bridge, and the exact ratio at which a majority claim loses its sign — plus a refusal when your stated beliefs and the data are jointly impossible.
One row per study. The left column is what a society says it wants; the right is what the record shows it does. The gap between them is the story. Read the series →
Poet first, researcher second. Three multimedia volumes published in 2026 — every poem free to read on the page, each with an original soundtrack composed and produced by the author.
Five acts, one for each season, narrated from first page to last by a cat who watches a love story arrive, settle, sour and leave.
Volume 02 · bilingual RO / ENOne tower block, forty floors, forty days of mourning. A narrator climbs and reads the doors — one poem per floor.
Volume 03 · Bilingual RO / EN · ● NewFifty poems about dating apps, distance and the metrics of loneliness — the profile grid read as a building of closed doors. With the original soundtrack “Still Typing”.
The wing itselfBoth volumes, the audio, and the prose that sits outside them. The literary side of the same listening practice.
Alex-George Adam — poet, researcher and strategist in Bucharest, working where online language becomes behavioral evidence.
My work sits at the intersection of social listening, cultural strategy, influencer ecosystems, creative performance and applied AI. Over a decade across leading global technology and consultancy environments, building intelligence frameworks for decoding online behavior — and the noise beneath it. Fifty industry awards, five Effies.
The Listening Room is an independent practice. Echo is the instrument suite. The research defines what the data can support. The writing asks what the numbers leave behind — and none of the three is decoration for the others.