For personal injury media buyers

A county isn’t a media strategy.

A broad Google Ads location tells you where to advertise, not which local crash patterns deserve your attention. See how cleaner records and more complete maps can sharpen the research behind your next campaign.

Compare the maps Discuss a local research brief ↗

1–45–1920+crashes per cell

1 located records fall outside this county view. They remain in the totals.

Orange County · Serious injury / fatal · January–September 2026
Reported and inferred locations, grouped into approximately 2 km cells. Not exact crash sites.

Start with the map

Change the question.
See a different pattern.

A county-wide campaign can cover very different local contexts. Compare all reported crashes with serious-injury/fatal crashes or commercial and large-vehicle involvement before choosing the next content test.

Compare a county

January 1–September 30, 2026 · Same records and county definition in both panels · Snapshot captured October 29, 2026 UTC

21,467 all reported crashesOrange County · 100.0% of all reported crashes in this county

Before location enrichment

Usable coordinates from the source report only

1–45–1920+crashes per cell

18 located records fall outside this county view. They remain in the totals.

15,682 can be placed · 73.1% coverage
5,785 have no usable source coordinates.

With location enrichment

Source coordinates plus inferred locations

1–45–1920+crashes per cell

18 located records fall outside this county view. They remain in the totals.

18,885 can be placed · 88.0% coverage
2,582 still cannot be placed.

3,203 more records have a map location.

3,203 of the locations in the enriched view are inferred from road context, not recorded GPS positions. Both panels contain the same 21,467 crashes in their totals. Cleaning did not create additional crashes.

This reconstructs a source-only view from the current snapshot, not a screenshot of an earlier website release. Source coordinates pass geographic bounds checks, not an independent accuracy audit. Maps use the same approximately 2 km grid and fixed color scale. A shaded cell is not an exact crash site.

Look inside the county.

These are the five reported city labels with the most matching crashes in this selection. A crash location is not necessarily where the involved person lives or searches for help.

Open this filter in the interactive map ↗

Orange: all reported crashes
Reported city	Crashes
Anaheim	3,802
Santa Ana	3,621
Irvine	1,518
Garden Grove	1,412
Orange	1,241

Map sources and counting rules

Crash records: California Highway Patrol CCRS. County outlines: U.S. Census Bureau, 2025 TIGER/Line. Road context used for inferred locations: © OpenStreetMap contributors.

Each crash is counted once within each view. The serious-injury/fatal view uses reported fatalities or the source’s serious/fatal injury classification. The commercial / large-vehicle view uses reported vehicle-type descriptions. These filters can overlap. County membership follows the reported county, with the warehouse county used when that field is missing; it does not move between the before and after views.

City rows are reported labels, not verified jurisdiction assignments. Records without usable coordinates remain in the counts. This fixed research snapshot will not necessarily match the live map after source revisions or further reconciliation.

Why the distinction matters

Orange County changes position when the question changes.

Among these five counties, Orange County is second by all reported crashes, but fifth by serious-injury/fatal crashes. That is a reason to refine the research, not a verdict that Orange County is a bad advertising market.

All reported crashes

Los Angeles73,128
Orange21,467
Riverside20,127
San Bernardino18,938
San Diego18,106

Count of distinct reported crashes. Each chart uses its own labeled scale.

Serious injury / fatal

Los Angeles3,067
San Diego793
San Bernardino720
Riverside699
Orange611

Count of distinct reported crashes. Each chart uses its own labeled scale.

January–September 2026. Counts are not risk rates, available legal cases, audience sizes or forecasts of campaign performance.

Don’t let a broad location setting do all the thinking.

Paying for reach across an entire county can be frustrating when your message serves a much narrower need. Use crash patterns to choose more relevant local topics and markets to test, then use your own search, conversion and cost data to decide where budget belongs.

The map improves the research behind the decision. It does not show which clicks were wasted, where prospective clients live, or how much a campaign will save.

Behind a more useful search

Same crashes.
Less guesswork finding them.

Spacing, abbreviations and spelling differences can split a road’s records across different search labels. We bring reviewed variations together without rewriting the original reports.

Before · Fragmented search

136 separate search labels

340 different reported labels fed these groups.
Each square represents one old search group, not a crash.

After · Consistent search

One route name: I-405

I-405

Travel direction remains separate
Original labels stay available

24,696 distinct crashes in this reviewed example. The same records, organized more consistently.

Not additional crashes. Not proof that every record is correct.

Los Angeles city · Main road recorded in the report · All dates in the fixed October 27, 2026 review. These are reviewed examples, not complete highway totals. Checked October 28, 2026.

See all three examples and how we counted

Before and after within the same reviewed records
Route	Reported labels	Old search groups	Route names after	Distinct crashes
I-405	340	136	1	24,696
US-101	413	122	1	29,527
I-5	335	111	1	23,050

We compared the search labels before and after cleaning on the same set of reviewed records. We counted each crash once within each example. Different search groups do not mean different physical roads. Travel direction is a separate filter; ramps must not be treated as the main highway.

Better evidence. Clear boundaries.

What this can help you do

Build more specific local educational content.
Compare consistently defined groups of reported crashes.
Explain your sources, dates and missing information.
Choose a research question to test against campaign results.

What these records cannot promise

Case value, fault or a person’s legal eligibility.
More conversions or lower acquisition costs.
A “most dangerous” ranking from counts alone.
Permission to identify or contact individual crash victims.

What is ready today, and what still needs work?

The road-label improvements shown above are published. Precise road matching that separates two same-name roads within one city is still being verified and is not yet a statewide public feature.

Our quality review also found repeated 2025 vehicle/party records. The maps here use 2026 data and count distinct crashes, not vehicle or party rows. The commercial / large-vehicle view is a reported vehicle-type signal, not verified commercial use. Historical vehicle totals still need a corrected baseline.

Road-name cleaning and location enrichment are separate improvements. Neither proves every source report or coordinate is correct, and neither establishes that this dataset is better than every other source.

Where does the evidence come from?

California Highway Patrol public crash records, published through the California Crash Reporting System dataset. Evidence for Zero’s fixed October 27, 2026 road-label review covered 1,456,144 distinct crashes. The three examples above are subsets of that review, rechecked October 28.

Original road labels were preserved. Standardization changes how records are organized and found, not what the original report said. The figures on this page are a dated comparison, not a live dashboard.

Read our data methodology

Start with one local question

Bring a market.
Let’s build a clearer brief.

Tell us the geography, audience and decision you’re working on. We can discuss an aggregate research brief with clear sources and limits.

Discuss a research brief ↗Browse public road profiles

Public records stay public. Sponsorship does not buy control over findings.

Change the question.See a different pattern.

Before location enrichment

With location enrichment

3,203 more records have a map location.

Look inside the county.

Orange County changes position when the question changes.

Don’t let a broad location setting do all the thinking.

Same crashes.Less guesswork finding them.

136 separate search labels

One route name: I-405

Better evidence. Clear boundaries.

What this can help you do

What these records cannot promise

Bring a market.Let’s build a clearer brief.

Change the question.
See a different pattern.

Same crashes.
Less guesswork finding them.

Bring a market.
Let’s build a clearer brief.