● LIVE · EPISODE 03 · MAY 2026 The Trust Market · interactive systems journalism
03 · BUILD A COMPOSITE · WATCH THE SCORES SHIFT

Tap a life event.
Watch the dossier fill in. Watch the scores follow.

Each tap fires a life event — graduation, a card, a marriage, a medical collection. The composite sections ratchet. Six scoring systems update. The trust-tax meter reacts. The point isn't the number — it's how information you give in one place flows to systems you didn't know were watching.

LIFE EVENTS · TAP TO FIRE

Your life so far · what you tell the system without realizing it

⚠ FOR PARENTS · WHAT KIDS DON'T KNOW THEY'RE GIVING

Your kid's pre-adult composite is ~40% built by age 13.

Before they ever apply for a job or a loan or an apartment, the system already has their video-game spending patterns from Roblox/Fortnite, their YouTube watch history back to age 7, their school-district medical records, their school-bus GPS pattern, their TikTok engagement profile, and their browser-tracking history. At 18, the composite is waiting for them. They just inherit it.

~40%
composite filled by age 13
~700
tags on the average teen
11 yr
avg span of self-served history before age 18
COMPOSITE · LIVE FILL

The copy of you takes shape · one event at a time

Identity + IDsLexisNexis · LiveRamp
Behavioral signalsAcxiom · Oracle
Credit + cash flowExperian · TransUnion · Plaid
Employment + payrollWorkday · Equifax Work Number
Housing + tenantCoreLogic · RealPage
Desirability + datingMatch Group
Health + medicalOptum · Athenahealth
Genetic + family23andMe · AncestryDNA
Location + movementSafeGraph · X-Mode
Court + public recordsThomson Reuters CLEAR
SCORING OUTPUTS · LIVE

Six systems · updating right now

INSURANCE1.00×
JOB CALLBACK71%
MORTGAGE RATE+0bps
DATING TIER
RENTAL FLAGgreen
BACKGROUND0.08
TRUST-TAX · ANNUAL · ILLUSTRATIVE MIDPOINT
$0
Conservative midpoint estimate of annual extra cost for this composite vs. the best-decile composite, same ability and intent. Lifetime range (40 yr): ~$120K–$220K in published estimates · we lead with the midpoint, not the high end.
Try a sequence. Fire first credit card, then moved cities, then medical collection. Watch how an honest life — one bad medical bill, one work move — collapses scores in five different systems. None of these decisions are explained to you in real life. They just happen.
02 · WHO RANKS YOU · WHO OWNS YOU · WHAT THEY SEE

The architecture of the quiet ranking.

Before we play with the composite, see the gatekeepers. Six sectors already score you. Four companies own most of the picture. Different gatekeepers see different slices — together they see everything.

If your news, ads, prices, and social proof are all different from the person next to you — your opportunities are already diverging. The next layer takes those signals and turns them into scores attached to your name. You don't get told. You just stop getting the call back.

HIRING
Resume-ranking AI in 99% of Fortune 500
75%
of resumes filtered before a human reads
Applicant-tracking systems score keyword fit, prior-employer prestige, and social-graph proxies. Some cross-check broker-supplied attributes silently.
WHO RANKS YOU · Workday · iCIMS · LinkedIn Talent Insights · HireVue · Eightfold.
INSURANCE
Underwriting by composite
700+
attributes per quote
Credit-based insurance scoring is legal in 47 states. Auto + home + life + health insurers buy from the same broker layer that prices your ads. Same data, different decision.
WHO RANKS YOU · LexisNexis Risk Solutions · TransUnion · Verisk · Milliman.
CREDIT + LENDING
FICO + alternative scoring
~1,200
alt-score inputs per applicant
Traditional FICO is just the front door. Behind it: utility payments, rental history, BNPL patterns, social-graph stability, even smartphone model. Sub-prime lenders score most aggressively.
WHO RANKS YOU · FICO · VantageScore · Experian Boost · ZestAI · Upstart.
DATING
Desirability scoring is real
internal
"elo" rankings per user
Major apps assign each user an internal attractiveness/engagement score that decides who sees you and who you're shown. They don't publish it. They do match against it.
WHO RANKS YOU · Match Group (Tinder, Hinge, OkCupid) · Bumble · Grindr.
DNA + GENEALOGY
Your and your family's genome
30M+
US samples in commercial databases
23andMe, AncestryDNA, MyHeritage. Once one cousin tests, ~60% of US adults of European descent can be identified by triangulation — whether they opted in or not.
WHO RANKS YOU · 23andMe · AncestryDNA · MyHeritage · GEDmatch (law-enforcement accessible).
BACKGROUND + RENTAL
Eviction scoring · tenant blacklists
90%
of landlords use a scoring service
Tenant scoring layers eviction records, credit pulls, and predictive models that estimate "future eviction risk." Errors are common, appeals are rare, and the score follows you across cities.
WHO RANKS YOU · CoreLogic · TransUnion SmartMove · RealPage · Yardi.
WHO OWNS THE MOST COMPLETE COMPOSITE OF YOU

Not the company you signed up with.
The four that bought the rest.

Block size = approximate share of total US adult digital-personhood data under each operator's roof. Tier-1 owns the substrate (platforms + ad networks + cloud). Tier-2 brokers fuse cross-source. Tier-3 sector specialists own deep verticals. Tier-4 is fragmented.

META / GOOGLE / APPLE / AMAZON
Substrate + ads + cloud~62% of total signal
MICROSOFT
Cloud + LinkedIn + ads~12% of total signal
ACXIOM / LIVERAMP
Consumer data broker~6%
EXPERIAN
Credit + marketing~4%
LEXISNEXIS
Risk + insurance~3%
CORELOGIC
Housing~2%
23ANDME
Genomics~1.5%
PALANTIR
Gov + enterprise~1.5%
EVERYONE ELSE
Long-tail brokers + sector tools~8%
SCOPE · WHAT EACH TIER CAN SEE

Different gatekeepers see different slices. Together they see everything.

Platforms
Brokers
Insurers
Banks
DNA / Health
Browsing + clicks
FULL
PARTIAL
Location history
FULL
FULL
PARTIAL
PARTIAL
Purchase + transactions
PARTIAL
FULL
PARTIAL
FULL
Social graph + DMs
FULL
Medical claims + Rx
PARTIAL
FULL
FULL
Income + assets
PARTIAL
FULL
PARTIAL
FULL
Genetic + family history
PARTIAL
FULL
Photos + face IDs
FULL
PARTIAL
PARTIAL
WHERE THIS IS HEADING · THE END GOAL

The next decade isn't about more data.
It's about fewer humans in the decision.

A single composite score
Hiring + insurance + lending + housing + dating fuse into one cross-sector reputation token. Privately held, opaque, mostly accurate, occasionally devastating.
Autonomous decision agents
Bidders, recommenders, underwriters, recruiters — all AI agents transacting on your composite without any human in the chain until it's already locked in.
The quiet denial
You won't be told no. The interview just doesn't come. The rate is slightly higher. The match never shows up. The credit is approved but at a different number than your neighbor.
04 · THE PEOPLE WHO PAY · TWO TRUE-SHAPE STORIES

The composite is not abstract.
It costs particular people particular money.

Two composite case studies. A second cousin's DNA. A spouse's old credit file. Each one shows how a single signal you didn't authorize can re-shape a life.

CASE 01 · THE COUSIN'S DNA

She never spit in the tube.
Her premium went up anyway.

M
MARA · age 41
DESIGNER · NEVER TESTED · MAINE
"Mara is one of thirteen extended family members the model can identify with confidence — without her ever submitting a sample. Press play to watch the schedule unfold, or click any relative to toggle them tested / untested and see how the model's confidence + her premium re-compute live."
RELATIVES TESTED
0
MODEL CONFIDENCE
0%
PREMIUM MULT.
1.00×
CONSEQUENCE · 30 YR
+$48,600
Estimated additional life + LTC + disability premium from data she never authorized.
Sweeney et al. 2018
GINA §201–§215
42 USC §2000ff
FAMILY · LIVE YEAR 2021
1.00×
CASE 02 · THE SPOUSE COMPOSITE MERGE

She kept her credit clean.
Her wedding cost her 0.3%.

L
LENA · age 29
NURSE · NEWLY MARRIED · CHICAGO
"Lena had a 790 FICO. Theo had a 651. They didn't merge bank accounts. The system merged them anyway within 90 days — and Theo's 2018 late-payment markers transferred onto Lena's household composite. Press play."
MARRIAGE STATUS
SOLO
EXTRA PREMIUM
$0/mo
GHOST MARKERS
0
CONSEQUENCE · 5 YR
+$8,940
Extra in insurance + lending from a spouse's history she never agreed to share.
CFPB 2023 household-scoring
LexisNexis CLUE
Markup 2024
COMPOSITE COLLISION · LIVE YEAR 2024
HOUSEHOLD COMPOSITE · ACTIVE LENA FICO 790 · CLEAN THEO FICO 651 · 3 LATES (2018)
Lena marker Theo flag Household merge
05 · THE PROTECTION LAYER · MINIMAL VIABLE DEFENSE

Two hours of work · one bad week never reaches your future.

The asymmetry is real — the system is set up to extract more than you can ever clean up. But there's a defensible floor most people never reach. Every action below is rated by Time · Cost · Effectiveness (★) · Discomfort so you can pick the ones that fit your hour. Three of them, done today, cover ~70% of the avoidable damage.

CREDIT + FINANCE
Freeze credit at 3 bureaus + Innovis
15 min · $0
★★★★★
Pull annual reports + dispute errors
1 hr · $0
★★★★
Lock Equifax Work Number
10 min · $0
★★★★
INSURANCE
Request CLUE report · LexisNexis
30 min · $0
★★★★
Decline telematics opt-in
2 min · $0
★★
HOUSING
Pull tenant-screening files
30 min · ~$25
★★★★
Dispute eviction records
2 hrs · $0–200
★★★★
HEALTH + DNA
Don't submit DNA. Ask family not to.
0 min · $0 · this one is hard
★★★★★
Period tracking → paper / Drip
15 min · $0
★★★★★
23andMe: request federal deletion
20 min · $0
★★★
EMPLOYMENT
LinkedIn: signals off
5 min · $0
★★
Request HireVue / Pymetrics accommodation
10 min · $0
★★
FOR KIDS · FAMILY
Freeze children's credit at birth · all 3 bureaus
30 min · $0
★★★★★
Opt out of school-district data-sharing (FERPA)
20 min · $0
★★★★
Roblox / Fortnite / TikTok privacy settings
20 min · $0
★★★
EPISODE 04 · UP NEXT · DRAFTING

The Influence War.

If Episode 03 is what other systems do with your composite, Episode 04 is what political + persuasion systems do with it. Microtargeting · synthetic constituents · campaign trust-market integration. The system tries to make you become someone specific.

The strongest counter-read · the speculative-fiction charge click to expand

A reasonable critic would say: "this is not journalism, it is Black Mirror dressed in data viz." The Trust Market episode describes a system that is partly documented (the broker stack) and partly speculative (programmable trust as tradable infrastructure). Mixing them at the same visual confidence is the strongest charge against this entire publication.

We accept the diagnosis. The provenance layer (lime/amber/pink badges) and the press-E Evidence Mode are our answers — every load-bearing claim on this page is now visibly tagged with its category. Documented harms are not extrapolation. Extrapolation is not prediction. Read the badges.

The fragmented-regulation rebuttal is also worth taking seriously: existing CCPA, GDPR, the FTC's broker actions, and state insurance regulation may prevent the worst Trust-Market scenarios from materializing as drawn. That outcome is on the table. We hope it lands.

Press E for evidence mode
Evidence mode · on
Documented Modeled Speculative