01Synthetic panelsOne study runs across many markets and demographic segments at once — synthetic respondents, not a recruited panel, with demographic priors anchored on World Values Survey (wave 7) data.
02Cross-model agreementCross-model agreement is scored with Jensen–Shannon divergence — a reliability signal for questions put to more than one provider.
03Cultural validationEach finding clears eight validation walls — cultural, contextual and disclosure — backed by a country-indexed structured knowledge base, before it ships.
04Market-Readiness IndexMarkets ranked on cultural, economic, regulatory and sentiment readiness — one comparable score.
05A proof chain per studyEvery step is hash-chained into a tamper-evident trail — what was run, in what order, and what each step produced.
Interactive demo — the statistics (bootstrap confidence intervals, Cohen’s d effect size, weighted eight-wall aggregation) are computed in your browser; the synthetic cohort and the cultural signals are a seeded simulation. A cross-sectional read of sentiment across markets, not a forecast.
Stimulus
Panel composition
9
25 to 44
62% urban
400 / cell
Regions
Sentiment readIdle
Overall distribution95% confidence interval
Positive—Neutral—Negative—
Emotion breakdownbar · CI whisker
Per-region read—
Panel segments
Cultural layer
Run the panel to surface cultural, contextual and disclosure signals.
Segments polled0Mean agreement—CI width—Markets9
§ 02Who it is for
For buyers whose claims face outside review.
01AI labsDiscipline for model evaluations — pre-registration and confidence intervals brought to your own eval design.
02Regulated technologyAudit-ready research a compliance team and outside counsel can read end to end.
03Category leadersA pre-registered benchmark on your category — the reference your market ends up citing.
04Policy and researchCross-national cultural-values work, built to survive peer review.
§ 03The discipline
Three rules the field has not adopted.
Adopting them means shipping more failures. Spectra ships them anyway.
01No claim without pre-registrationWe pre-register your hypothesis and gate — committed and hash-stamped — before we run your study.
02No number without a confidence intervalYour study ships with bootstrap confidence intervals on every quantitative reading. No false precision.
03No failure quietly droppedWe log every run, passing and failing, and publish our methodology — retractions included.
§ 04Perimeter
Honest about what it will and will not take on.
In scopeWhere Spectra runs
01Pre-registered brand and concept benchmarking
02Cross-market cultural-values research
03Pre-launch concept screening with a bounded go or no-go
Out of scopeWhere it declines
01Forecasting and launch-performance prediction — that is Horizon
02Any number shipped without a confidence interval
03General-purpose survey pretesting
§ 05Offers and terms
Two doors. A study you can buy, a Standing you are brought to.
The Signal Audit is a fixed-scope study with a public price. A Spectra Standing is the recurring relationship — by introduction, priced to you, never to a list.
02Spectra StandingSTANDINGThe operated, recurring relationship — ongoing studies on operated panels, and the roadmap that re-validates them as frontier models change. The Audit credits in full toward year one.By introduction
03EnterpriseMulti-market or sovereign panels, custom scope beyond a Standing.Request a quote
What a Standing holdsHeld for Principals
01The held rateA recurring annual rate on defined scope, held for a bounded term, then a bounded step. Expansion is metered above the base.
02A voice in the methodologyDirect roadmap input with the people building Spectra. The Principal is consulted; NEXI decides. Influence, never authorship.
03The mark and a referenceA named, mutually approved reference, and the right to carry the NEXI verification mark.
04First access, namedFirst production access the day Spectra graduates, and a Principal designation, named with your consent.
Why a Standing recursThe ongoing value
01A study is valid for the models it ran onA synthetic panel mirrors real survey data through the frontier models behind it. When those models change, that correspondence can drift — and a one-shot study cannot flag it.
02Re-validation is the layer a Standing fundsThe recurring product re-runs your study against current models and refreshed ground truth, and re-issues its validity stamp. It is being built; a Standing funds it and is first to it.
03The methodology is open; the operated panel is notProvenance standards and pre-registration templates are published. A Standing is the operated panel and the re-validation built on it.
The fine printRead before you sign
01Build stateSpectra runs today as a research instrument — validated synthetic panels, cross-market study, tamper-evident proof chain. Per-study model-version pinning, hypothesis pre-registration, confidence-interval reporting and re-validation are in development.
02Open and paidThe methodology is published open — the provenance and comparison standards, the pre-registration templates. A Standing is the operated panel and the re-validation built on it.
03Ground truthPanels are validated in a controlled benchmark against real survey data (World Values Survey Wave 7). That is a benchmark result, not a production accuracy guarantee.
04The rateThe Audit price is fixed and public. The Standing rate is quoted per account against modeled value, held for a bounded term, and never published. Earlier accounts hold the lower rate.
05Scope of the workSpectra reads sentiment and culture. It does not forecast launch performance — that is Horizon — and ships no number without a confidence interval.
Rates follow NEXI’s Verify Value Model — modeled value, a pricing estimate, never a discount. Methodology, models and the standard stay NEXI’s sole property. Enquiries: hello@nexiverify.com.
01What is Spectra?Spectra is NEXI’s sentiment and cultural-intelligence product. It runs synthetic research panels — model-simulated respondent studies — and seals each study with a tamper-evident proof chain.
02What is a synthetic panel?A synthetic panel uses AI models to simulate how defined audiences would respond to a question or concept, in place of or alongside a traditional recruited survey panel.
03How does Spectra make synthetic research trustworthy?Spectra scores cross-model agreement, validates findings on cultural and contextual dimensions, and seals every study in a hash-chained proof trail anyone can re-check.
04Who uses synthetic panels?Teams testing concepts, messaging or market fit who need fast results — and need those results to survive outside scrutiny.
Spectra is forming its first Series.
By introduction. The Series is named when it closes, never counted down in public.