Turning lived experience into market information

The market prices the company. Who prices the customer’s lost afternoon?

AntiVZN is a research and bounty experiment around a Solana token paired, conceptually, with Verizon equity. The token coordinates evidence collection. It does not prove that Verizon falls, or that the token rises.

Working thesis

Hidden customer costs may be economically legible before they are visible in company reporting.

Support trees, billing repairs, cancellation loops, restoration delays, and repeated explanations transfer labor to the customer. Structured stories can test whether that labor is associated with churn, refunds, complaints, switching, or other measurable exposure.

One angry customer is an anecdote. Ten thousand coded experiences are still not truth without a sampling frame, a denominator, and contradictory cases.

The boundary

Not a short recommendation.

The artifact measures an information gap. It does not direct trades, promise a hedge, or manufacture a lower share price.

Virtual pair

Energy in → evidence out → uncertainty reduced

r_AV(t) + β · r_VZ(t)

That is a descriptive return comparison, not an inverse ETF. Beta, windows, liquidity filters, missing observations, and market clocks must be preregistered before interpretation.

Current receipt: ANTIVZN observed at approximately 0.00000288 USD with 343.93 USD 24h volume; VZ observed at 49.45 USD. The token market is too thin for comfort. Thin markets are not oracles, merely small rooms with loud echoes.

Falsifier

If the customer cost does not predict anything, publish that.

The thesis weakens if the pattern disappears against baselines, cannot be independently coded, or has no relationship to a measurable outcome.

Submit a story

Reward evidence, not outrage.

1. Redact

Remove names, phone numbers, addresses, account numbers, passwords, codes, payment details, and private correspondence.

2. Structure

Give the approximate date, service category, mechanism, time-cost range, resolution, and what you can support.

3. Validate

Independent review assigns an evidence tier. Points are provisional until the review receipt exists.

Bounty schedule

10 / 25 / 50 points

T1 structured firsthand incident: 10. T2 receipt-supported incident: 25. T3 replicated pattern: 50. Review labor: 5. Reproducible public-safe analysis: 40.

100-point contributor cap per 30-day season. Points are a shadow ledger unit, not a payout promise. Any live settlement requires a separately approved budget and readback receipts.

What Verizon says it measures

The CEO already named the scorecard.

Dan Schulman's public transformation narrative links customer friction to churn, acquisition and retention cost, customer lifetime value, and cash generation. AntiVZN is not inventing a complaint category. It is trying to measure the customer-side cost Verizon's own scorecard leaves aggregated.

Q1 2026: Consumer postpaid phone churn 0.90%; March below 0.85%; acquisition and retention cost approximately 35% below Q4; postpaid phone net adds 55,000; free cash flow approximately $3.8B; $5B 2026 OpEx savings target.

Q2 2026: 184,000 postpaid phone net adds; 348,000 broadband net adds; $23.4B mobility and broadband service revenue, up 2.8%; first-half operating cash flow $18.4B versus $16.8B.

Proposed research budget

5,000,000 ANTIVZN maximum shadow allocation.

This is a proposed, capped research pool for validated customer interviews and evidence review. It is not currently escrowed, transferred, or payable. No token payout is wired to this page.

The intended flow is: consented story → structured transcript → redaction → independent consistency review → cross-story mechanism map → human approval → bounded settlement receipt.

A story does not earn because it is angry. It earns only if its evidence is coherent, its uncertainty is visible, and its mechanism survives comparison with other stories. Funding, eligibility, identity, tax treatment, and settlement terms require a separate season approval.

Marvin architecture

Agentic review, not agentic custody.

Marvin may organize claims, detect contradictions, map recurring mechanisms, and recommend a tier. It must not decide final eligibility or send tokens autonomously.

Human review remains the payout gate. The agent is a clerk with a microscope, not a treasurer with a pulse.

Campaign scoreboard

Separate market weather from Verizon information.

r_VZ-specific = r_VZ − 0.23 · r_SPY

VZ is unusually insensitive to the broad market, so the useful question is not simply whether the stock fell. It is whether Verizon underperformed the market after a fixed beta adjustment.

Current five-year receipt: 1,254 matched daily observations; VZ +22.40%; SPY +83.18%; cumulative residual +5.02%; worst daily residual −7.58%.

This is a descriptive baseline, not evidence that AntiVZN caused any move. Event windows need timestamps, placebo dates, beta sensitivity, and competing-news controls.

Five success layers

Evidence before theater.

Instrument integrity → evidence production → independent replication → Verizon-specific response → company action.

The strongest receipt is a named owner, requested evidence packet, corrected mechanism, changed process, or leadership meeting with a written readback.