Market research runs on a question nobody can verify: do you actually own it? Panels take the respondent's word and 30–40% of the answers are wrong or fraudulent. Here that answer is public record — so we read it instead of asking, and you buy the ones that check out.
Self-serve. Live within minutes of funding. Results public by default.
Holds at least $100 of tokenized NVDA. Wallet older than ninety days. Twenty transactions or more. Every condition is a thing the chain can settle, so none of them can be answered wrongly by the respondent.
The eligible wallets are counted from a published census, not estimated from a panel book. If the audience is too small for the sample you want, you find out before you pay rather than three days into fielding.
Deposit the rewards in USDG. Respondents are screened again server-side, answer, and are paid in the same second. Every response arrives stapled to what the chain said about that wallet at that block.
A panel asks whether you hold the stock and prices your answer as fact. We check the position, its size, when the wallet was created and how much it has traded. None of it is self-reported.
The eligible wallets are a published list frozen at a block, so a result reads “212 of the 1,450 wallets that qualify answered”. No research vendor on earth can state its denominator.
Each response carries the chain reading that admitted it and the transaction that paid for it. You can re-check any row against the chain yourself, months later, without asking us.
Prolific takes 42.8% and says so on its pricing page; expert networks take about 70%. That spread pays for the fraud they absorb. Ours is smaller because the fraud is cheaper to stop.
No seat licence, no annual minimum, no sales call. The whole cost of a study is what respondents are paid plus our share, and both numbers sit on the study page for anyone to read.
One question set, one audience. Enough to publish a number.
Enough to cut the result by position size and wallet age.
Deep positions and long tenure. Small, rare, and worth the price per head.
Illustrative bundles of the same self-serve pricing — the builder prices your exact audience live.
Sentiment surveys ask people whether they own the stock and take the answer on faith. This one does not have to: every respondent provably holds tokenized NVIDIA, and the size of their position is public. So the readings below are not what retail investors say they think — they are what NVIDIA holders think, cut by how much they actually have on the line. Answers are paid in USDG the moment they are accepted.
One person can hold several. No figure on this site is described as a share of a population, and none of them should be read that way.
There is no demographic weighting, because there is no demographic data. National brand tracking needs a different vendor and we will say so before you buy.
Everything a buyer sees is already public on the chain. No name, no email, no document. What you type is the answer to the study and nothing else.
Respondents are paid in USDG, a redeemable stablecoin. Every project that paid its users in a token it minted itself was diluting, not earning.
The full method, the screening vocabulary and the anti-fraud design are written up in the documentation.