For premises-liability media buyers
Give every condition the right context.
Separate reported sidewalk defects, location history and municipal payments. Build more specific campaign briefs while keeping the injury, notice and ownership questions visible.
Build the brief
Change the condition. Change the brief.
A blocked sidewalk, a displaced slab and a private-property spill call for different research and qualification. Keep the public evidence attached to the question it actually supports.
8,872 recent defect requests.
The selected SF window includes 8,872 sidewalk-defect requests, separate from 17,338 blocked street/sidewalk reports. Accepted exact-address groups include 3,478 with repeat requests across selected categories.
Fixed October 29, 2025–October 28, 2026 window. Repeated requests do not prove the same unrepaired condition.
Test a condition-specific explanation.
Compare a sidewalk-defect evidence checklist with a general premises page for comparable eligible traffic. Use exact place, condition and timing for subsequent research.
Qualify the inquiry
- Exact location, condition and incident date
- Injury, causation and responsible public/private party
- Notice evidence and governmental claim dates
Measure: Attorney acceptance, municipal/private routing and signed matters after lag.
No accepted same-condition incident-period outcome join was established. A 311 request is not a legal notice finding, and administrative closure is not verified repair.
The evidence register
Preserve the grain. Keep the missingness.
Keep the record grain, date window and missing information beside every claim. These are source-backed research observations.
| Observation | What was counted | Boundary |
|---|---|---|
| Full LA liability feed13,029 payment rows | All source categories; 64 amounts unknown and 2,267 rows without case/claim identifiers. LA liability payment feed | Not a complete census of filed, rejected, unpaid or successful claims. |
| Explicit fall categories$182.2M known amounts | 1,969 source payment rows; 16 amounts unknown. Source case/claim IDs group payments where present. LA liability payment feed | Not 1,969 distinct cases, a success probability or an average recovery benchmark. |
| Recent SF sidewalk defects8,872 requests | The fixed recent window includes selected Sidewalk/Curb and Street Defect subtype paths. SF service requests | Requests are not unique hazards, injuries or legal notice findings. |
| Repeat address context3,478 groups | Accepted exact-address groups with repeat requests; 10,497 requests in those groups across selected categories. SF service requests | Same address does not establish the same condition, parcel or responsible party. |
Why the LA feed cannot support a clean payment trend
The full feed has 4,127 missing payment dates and 167 ambiguous day/month dates left null. Another 1,632 unambiguous dates conflict with raw fiscal labels. Fiscal labels extend through FY2025/26, but populated unambiguous dates end December 6, 2024. Preserve these conflicts instead of presenting a current calendar trend. A name-only outcome candidate was also left unresolved.
Put the research to work
One clear question.
One testable change.
Choose an evidence-backed topic your counsel accepts. Keep the campaign comparison narrow enough to explain the result.
Define the evidence context
Pin the source, date window, record grain and exact entity. Preserve contradictory facts and missing information.
Change one part of the campaign
Compare a specific information page or qualification flow with a concurrent control. Keep geography, intake coverage and the outcome definition comparable.
Follow the matter through intake
Deduplicate inquiries, record attorney acceptance and allow time for signing. Reallocate only when mature economics, uncertainty and capacity support the decision.
The buying decision
Work back from
a signed matter.
Use your firm’s economics to calculate an acquisition ceiling. A public settlement amount cannot supply these inputs.
Inputs stay in this page’s memory. They are not sent to us or an ad platform.
Add all five inputs to calculate a ceiling.
Calculation and assumptions
Target CPL = signed-per-lead × expected contribution × (1 − retained share) − intake cost per lead. Target CPC = lead-per-click × target CPL. Negative ceilings display as zero. When signed-per-lead is zero, cost per signed matter is undefined.
Contribution incorporates expected noncollection, case costs and timing. Fixed overhead, uncertain conversion rates and additional acquisition costs need separate review. This calculator does not estimate legal merit, recovery or incremental return.
Sources and scope
Open the evidence.
SF provides selected sidewalk, blockage and street-defect requests. LA service data are separate historical lighting/water context ending July 2025. The new liability feed describes municipal payments but lacks consistent incident locations/dates. Neither source measures private-premises claim incidence.
Download the research brief (JSON) ↓Primary sources and counting methods
- SF service requests
Retained selected categories, creation cutoff and stable request-ID paging. Source updates are not transactionally frozen; latest recent opening is October 27, 2026.
- LA Controller data hub
Official hub links the liability dashboard; its public loader identifies the underlying full feed.
- LA liability payment feed
Full feed captured October 29 UTC. API pathname is stale; fiscal labels and inconsistent dates remain independent. Amount totals were not reconciled against the City ledger or bank completion.
- California GOV 835
Public-property liability has specific condition, causation and creation/notice elements.
- California GOV 835.2
Actual/constructive notice requires facts; a service request does not automatically establish legal notice.
- California GOV 911.2
Governmental claim-presentation procedures require prompt, fact-specific legal review.
- California CIV 1714
Private-premises negligence requires its own analysis; municipal totals do not determine its value.
The download includes the public claims, sources and four campaign hypotheses. Raw narratives, personal information and private research archives are excluded. Retrieved documents and current snapshots cannot establish what was publicly known before a historical incident.
Legal qualification and advertising boundaries
A particular case needs injury, causation, a dangerous condition, responsible ownership/control and the applicable notice analysis. Public-property claims have specific statutory elements and presentation procedures. GOV 911.2 generally has a six-month personal-injury/death presentation period, subject to fact-specific exceptions and other procedures; route promptly to counsel rather than promising or automatically rejecting timeliness.
Health and abuse-related promotions have platform targeting and data-use restrictions. Use eligible query/context campaigns and review each creative, landing page and measurement method. Sensitive intake belongs in restricted first-party systems. Public records are context, not a person-level audience list.
Google personalized advertising policy ↗ · California advertising requirements ↗
What would establish acquisition performance?
A dated keyword/auction export, mature deduplicated intake cohorts, attorney acceptance and signed outcomes, actual firm contribution inputs, and a prespecified concurrent comparison. Track pending matters and conversion lag. Public counts supply none of those ad-performance inputs.
No campaign was launched for this research. CPC, acquisition cost and incremental return remain unmeasured. Qualified and signed-matter details must not be exported to an ad platform without a method-specific policy and privacy review.
Start with a buying decision
Bring a place. Build a clearer brief.
Bring the geography, topic and decision you want to test. We’ll discuss an aggregate brief with dated sources and clear limits.
Public records stay public. A research brief does not promise legal eligibility or campaign results.