Skip to content
Friday, November 20, 2026 · Independent edition

Public safety · Technology · Advocacy

ProFlock

Methodology

What we count, and what we cannot prove.

Sources and limits for every number on this site, so readers can check the work.

01 · The atlas

3,217 news-documented cases in our archive

Each archive record corresponds to an entry in Flock Safety's public Solved Story Hub — a publicly reported investigation or missing-person recovery where Flock says its evidence played a role. Each record carries headline, city, state, category and date as indexed, and links to Flock's hub entry. 46 states are in the archive; coverage is reported in 49. Counts reflect what Flock published, not independent verification of each case.

What this is: a fraction with public records. Flock reports ~1M investigations supported in 2025 including leads and assists; the archive plots only the subset with public hub entries. What this is not: a random sample, a clearance-rate study, or proof any single case would have gone unsolved without cameras.

02 · The census numbers

2025 Impact Census and FBI UCR

Vendor numbers: ~1M investigations supported (including leads/assists), ~20% of cleared cases involved Flock where deployed (307 agencies that gave estimates), ~10,000 missing-person finds, 42% of responding agencies say Flock was involved in half or more recoveries. Flock calls these directional estimates weighted to NIBRS, not audited totals; CBS and criminologists note the lack of counterfactual. Federal baselines: FBI UCR 2024 Table 25 — motor-vehicle theft 9.2% cleared, murder 61.4% cleared, violent crime 43.8% cleared. We label vendor vs federal wherever they appear.

If the Census methodology updates, these pages update with it — and substantive changes get a dated correction note per our corrections policy.

03 · Illustrative vs. deployed

The 2032 grid is our proposal, not a deployment

The dashed trajectory segment, the 600,000-node illustration, the 2032 interstate corridors, and every proposal callout are ours — not Flock's roadmap and not existing infrastructure. The map labels the 2032 layer proposed/illustrative.

Product facts (no facial recognition in the product today, no ICE contract stated, 7-day default since Aug 2026, customer-owned data) are stated with sources. Our extras — year of case-locked history, sharing on by default with opt-out, face matching only for violent felonies/AMBER, same-controls sale to ICE — are labeled as our proposal.

04 · Limits

What this site can't prove

We cannot prove any single case would have gone unsolved without cameras. We cannot audit the Census; we cite it as vendor research with its stated limits. State counts reflect published hub entries, not uniform coverage. Court law is mixed: 30+ rulings have upheld fixed ALPR, but Oct 2026 N.D. Oklahoma suppressed a 30-day multistate warrantless trawl — scope and warrant rules matter.

Found a mischaracterized case or a dead source link? Send the correction — it goes live with a date stamp.