Tools
HUM.05 · HUM△NS ES
See it live

TDI / HUM.05 · simulates · The four instruments

HUM△NS™

Test your concept and your price before paying for fieldwork.

Before paying for fieldwork, focus groups or a new execution, HUM△NS lets you put an idea in front of profiles built on consumer survey data: a concept, a label, a price, a consumption occasion. It is not a substitute for a real person. It's meant to rule out weak hypotheses before you spend money testing them.

And every study goes through six tests and publishes the ones it fails: parameter recovery, out-of-sample validation, seed stability, circularity, assumption sensitivity, and external anchoring. One demo run failed two and wrote, on its own, «do not deliver as conclusive without correcting». Without an external data point to check it against, the anchoring test fails by design — and that gets printed too.

Concept test
which of your creative routes wins, with which cohort, and on which lever
PSM · pricing
what range the market lets you be in, and where you earn most
Quali
in what words they say it, and which objection surfaces first
Conjoint
which lever wins when something has to be given up

Who it's for, and what question they bring

Brand director«which of these three routes?»
Which one wins, with which cohort and on which lever — with its confidence interval. And if the top two fall within the noise, it tells you that instead of crowning one.
Creative director«are they going to kill my idea?»
No. It tells you which part of the idea carries the weight: in one real run, territory carried 30.9% and the name just 2.6%. It's there to defend what matters and let go of what doesn't.
Sales / commercial«what price range can I be in?»
The range it can move in and where it earns the most, before testing it in the market. Starting at $15,000 and within five business days.
Insights«is this representative?»
No, and it prints that. Six tests per study, the ones that fail get published, and without an external data point to check against, the report ships marked «not calibrated». It ranks hypotheses before you pay for fieldwork; it doesn't replace fieldwork.
HUM△NS screen with its four instruments —Concept Test, PSM/Pricing, Quali and Conjoint— each with its panel size and turnaround.
AThe four instruments, already built. Each with its own panel size and turnaround — N 800 and one week for concept test, N 600 for pricing, N 10 for quali. And on its own screen, without anyone having to ask: «the panels are simulated and still not calibrated against fieldwork in Mexico; they rank hypotheses, they don't measure anyone».
The brief for a concept test: brand, category, the dropdown to anchor on a market source, competitors, panel size and seed; below, the creative routes described in the client's own language.
BYou write the brief in your own language; the system builds the experimental design. Each route is described the way you'd describe it in a meeting —what it's about, what tone, what mechanic, who's saying it— and the engine translates that into attributes and levels. What matters is the dropdown at the top: anchor on a market source or run without anchoring, in which case the assumptions get marked JUDGMENT. There are 16 categories to anchor on, with their index cut in plain view.
Result of a concept test: choice index per route with its confidence interval, the weight of each lever in the decision, and above, the warning that the category moves the result by only 1.3 points.
CThe result, and above it, the warning that bounds it. Two routes at 78.6 and 77.2 with their confidence interval, and the breakdown of what drives it: territory 30.9%, tone 28.0%, mechanics 19.6%, voice 18.9% and the name just 2.6%. But before any of that, in red: «the category moves the result by only 1.3 points; the conclusion is almost entirely the team's judgment, not market data», with the list of what was declared and never reached the model —age profile, price sensitivity, brand share— because the source doesn't carry them.
Audit of six tests with its verdict: T1, T2 and T3 pass, T4 is borderline, T5 and T6 fail; below, the provenance of the assumptions and the block «what this study won't sign off on».
DAnd this is where this product breaks away from everything else. Every study goes through six named tests and publishes the result of all six. In this real run: T1, T2 and T3 pass; T4 is borderline; T5 and T6 FAIL. The system itself writes the verdict: «Fails T5, T6. Do not deliver as conclusive without correcting.» Below, the provenance —MEASURED 0 · DERIVED 4 · JUDGMENT 6— and the block «what this study won't sign off on»: without external anchors, T6 fails by design; that's not a blind spot, it's a lack of verification. A vendor looking to get paid doesn't print this.

What it looks like when it runs: Caribe Cooler

RTD · Mexico · August 2026 · seed 20260808

The brand had three campaign routes and had to decide which one to pay for: Bájale, Nos la llevamos cool and Soy tu Bias. Against New Mix, Smirnoff Ice and Topo Chico Hard Seltzer.

The three Caribe Cooler campaigns audited, each with its photography and its branded-reach score: Bájale 94.3, Soy tu Bias 21.7, Nos la llevamos cool 76.8
BAll three, each with its number. The real assets for each route, measured one by one: Bájale 94.3 in branded reach, the current campaign Nos la llevamos cool 76.8, and Soy tu Bias 21.7, because outside of young people only 26.5% understood the name. None of the three is bad; one simply had no gap left to fix.
The assets for the winning route Bájale: billboard, family group chat, December drama and icon
CThe route that came out ahead in simulation, piece by piece. Mexican vernacular lettering and a complement that adapts: to the family group chat, to the December drama. A campaign that only has executions runs out; one that has a grammar doesn't. HUM△NS prioritizes creative territories for research: no route is declared the winner without human validation or a market test.

What was fielded

Panel
1,500 profiles across five weighted cohorts aged 18 to 40, with three modeled traits —youth, fandom, nostalgia— and their deviation by cohort
Conjoint
five attributes (territory, name, tone, mechanics, voice), 10 tasks, 3 alternatives, 400 bootstrap repetitions and 10 audit seeds
Group
800 people, 166 interventions across seven segments: opening, warm-up, body, stimulus, laddering, projective and close
Control
one fixed comparison cell, so the winner is measured against something, not against itself

The finding that changed the decision

Name comprehension, measured cohort by cohort:

Bájaleeveryday Spanish imperative
100%
Nos la llevamos coolunderstood across every age group
96.6–98.8%
Soy tu BiasKorean fandom slang
41.2% → 14.8%

«Soy tu Bias» is understood by 41% of 18–25-year-olds and only by 14.8% of those over 26. Five out of six people in the category's largest group wouldn't know what it's talking about. That's a national campaign falling apart before it's even bought — and it took one week to find out.

What they said, in their own words

«Michelada, mezcal with something mixed in, those flavored vodka cans that are trendy now, a club with really ugly lights, and prices that go up at the supermarket for no reason.»

18–21 · warm-up

«Honestly, they should stop messing with something that's already good. If a brand works for me, I stick with it — I don't need them changing it every five minutes with some new angle.»

26–30 · close

And here the product corrects itself

The group was checked against what people actually search for online about the brand. Six topics turned out to be blind spots, and six came up on their own:

What the group never touched, and people do ask
  • What it's made of50 searches
  • How strong it hits4 questions
  • How it's drunk3
  • Who makes it3
  • Effect on the body3
  • What it tastes like2
What the group brought up on its own
  • Price64 mentions
  • Get-together occasion58
  • Belonging / neighborhood44
  • «Sounds fake»18
  • Doesn't understand the name5
  • Smoothness5

It doesn't say the group is wrong. It says the guide never asked.

That's the line the system writes on its own — and it writes the consequence too: those six topics go into the guide before any more qualitative work is fielded in the category. A tool that only confirms what you ask it isn't useful for deciding anything.

How far the model goes, and where the data takes over

Comprehension of «bias» has a measured ceiling of 55.8% — what people search for doesn't support more than that. The model places it at 26.5%, which fits under that ceiling. The ceiling is data; the position within the ceiling is model, and the two are labeled separately.

With «bájale» it's the opposite: nobody searches for what it means. The ceiling sits at 100% and restricts nothing — and that absence of searches is behavioral evidence that it's understood. It's the cheapest one there is.

USD 0.60cost of the whole group, 800 people and 166 interventions
9 / 10internal consistency, verified cohort by cohort
1 weekversus the six a field survey takes

It's in beta, and the product says so — not the fine print

Without an external data point to check against, the report gets marked «not calibrated». No number comes from a language model: every one comes from each profile's utility model, with a fixed seed — the same question gives the same result.

And if you already have something ready

Send me the campaign you're about to pay for.

Before the market tells you whether it worked —once it's already charged you for the mistake— we tell you. And if we have no way to measure your category, we tell you that too, no quote required.