GoodLeads — Data Dictionary v2.0

How to read this

Built for four readers at once.

Most data dictionaries serve one audience — usually a marketer skimming for value. This one serves four. The buyer evaluating fit, the data engineer scoping ingestion, the compliance reviewer sizing license risk, and the end-user piloting integration with a sample record in hand. Every section answers all four.

License classification, per field.

Every value carries one of three license tiers — at the field level, not the record level. Some fields stack both (a derivation built on top of an enriched source). Your compliance team can reason about residency, sharing, and resale on every cell.

Public Record

Sourced directly from Secretary of State filings and federal public data. No resale restrictions.

Enriched

Third-party-provider-licensed. Skip-trace, validation, and parcel data. Resale subject to your master agreement.

Derived

GoodLeads proprietary. Scores, classifications, and intelligence we compute. Licensed under your contract.

EnrichedDerived

Stacked. A derivation built on an enriched input. The parent license flows through — both apply.

Coverage vocabulary.

"Always" is ambiguous. We use a controlled set of four values so you can size your population for any field at a glance. All measured percentages are anchored to the sellable cohort (operating businesses worth selling) — not the full pulled set.

Always present 100% non-null on every record. You can rely on this column existing in every row.

State-dependent Present only when the source state files it. We document which states for each field.

Vendor-conditional Present when the enrichment channel returned a value. Otherwise null.

NN% Measured fill rate on the sellable cohort. Refreshed monthly from production.

Reading this with a delivered sample.

This dictionary describes the full product — 130+ fields across every state and enrichment path. Any single delivery ships a subset, and every delivery comes with a cover sheet that states exactly which columns it contains, the filter that produced it, the row count, and the coverage measured on those rows. A field that isn't in your file's columns isn't missing data — it's just not in that export. Because coverage depends on the filter applied, the percentages here are product-wide estimates on the sellable cohort; your sample's real numbers live in its cover sheet.

Row grain One row = one business × contact pairing. A business with two contacts is two rows.

Keys entity_id identifies the business, contact_ref the person. Use the pair as the row key — lead_ref repeats when a business has more than one contact.

Operational columns Some deliveries include helper columns — contact_id, has_phone, has_email, phone_type, gtm_segment_display — defined in the cover sheet.

Confidence on every derivation.

No value ships without a signal telling you how much to trust it. Three patterns recur throughout.

Pattern 01 Numeric scores 0–100, reported with raw score plus named tier — Reachability 73 → "Very Hot."

Pattern 02 Confidence tiers confirmed · likely · possible · unknown — used for classifications.

Pattern 03 Match status MATCHED · NEAREST · NO_MATCH — used for spatial and identity joins.

What you'll see in source attribution.

We've curated a supply chain you don't have to think about. Sources are described by role, not by brand — so your team isn't tempted to go around us, and so we can swap providers when one underperforms. The four roles you'll see:

SOS filing The state's own record. We pull within days of public availability.

Public federal data US Census Bureau (geocoding, county FIPS, census tract).

Curated enrichment channel Skip-trace, gap-fill, email validation, parcel data. Provider relationships managed by GoodLeads.

GoodLeads proprietary Our classifiers, scorers, and intelligence engines.

No matches

Nothing here. Try a different search.

Section 01 · Part I

Identifiers

The persistent keys you use to track a lead through your CRM, your campaigns, and your sales conversations. Two are ours, two are the state's. Ours stay stable across re-enrichment cycles; the state's stays stable for original-record lookup.

Identifier	Type	Description / Values	Source	License	Coverage	Use case
lead_refLead Reference	Text	Format `GL-{STATE}-{NNNNN}`. Example: `GL-CO-12345`. Stable for the life of the entity.	GoodLeads proprietary	Derived	Always present	Speak about a lead in plain language across your team: "GL-CO-12345 just hit Very Hot." Survives re-enrichment, re-scoring, and source-state schema changes.
contact_refContact Reference	Text	Format `GLC-{NNNNN}`. Allocated atomically — one person, one ref, even if they show up on multiple businesses.	GoodLeads proprietary	Derived	Always present	Track a person across multiple businesses they've formed. One ref = one human. Critical for de-duplication when a serial founder shows up on three LLCs.
state_entity_idState Entity ID	Text	The state's own assigned ID. Format varies by state: numeric (CO), alphanumeric (FL), document-number (NY).	SOS filing	Public	Always present	Audit back to the state filing. If a customer or compliance team asks "where did this come from?" this ID is the answer.
source_systemSource System	Categorical	Values: `CO_SOS`, `FL_Sunbiz`, `VA_SCC`, `NY_DOS`. Identifies which state's filing system produced this record.	SOS filing	Public	Always present	Filter, route, and segment by source state. Useful when your sales motion differs by jurisdiction.

Section 02 · Part I

Business Profile

The core firmographic record — what every B2B data product ships. The difference: we normalize 50 different state formats to one canonical schema, so your CRM ingests one shape no matter how many states you turn on.

Field	Type	Description / Values	Source	License	Coverage	Use case
entity_nameBusiness Name	Text	Legal name exactly as filed with the state. Case-preserved.	SOS filing	Public	Always present	The name you use in your opening line. Trigger-word analysis on the name itself feeds our industry classifier and "name in entity" signal.
entity_typeEntity Type	Categorical	Values: `LLC`, `Corporation`, `LP`, `LLP`, `PC`, `PLLC`, `Partnership`, `Other`.	SOS filing	Public	Always present	Legal structure. Drives banker/CPA/payment-processor qualification rules and changes the language you use in pitch ("LLC" vs "S-Corp" vs "PLLC").
formation_dateFormation Date	Date	ISO 8601 (`YYYY-MM-DD`). The date the entity was legally formed in the state.	SOS filing	Public	Always present	Days-since-formation is the single strongest "moment of formation" signal in B2B data. Day 1–8 buyers behave nothing like Day 30+ buyers. Filter aggressively.
statusFiling Status	Categorical	Values: `Active`, `Inactive`, `Dissolved`, `Withdrawn`. Normalized across state-specific raw values.	SOS filing	Public	Always present	Current standing with the state. Exclude non-Active on most outbound motions; include for compliance / dissolution-services pitches.
jurisdictionJurisdiction of Formation	Categorical	Pattern: `DOMESTIC_{ST}` or `FOREIGN_{ST}` where `{ST}` is the 2-letter state code (e.g., `DOMESTIC_CO`, `FOREIGN_DE`).	SOS filing	Public	Always present	Domestic vs. foreign authority. Foreign filings (especially DE-incorporated, operating in another state) skew sophisticated — different motion than a local LLC.
officer_countOfficer Count	Integer	Number of officers/members listed on the filing.	SOS filing	Public	State-dependent · FL, NY: reliable. CO, VA: RA-only filings often show 0.	Drives our solo-founder inference. 1 officer + self-rep RA + LLC = highest-confidence founder-led signal.
public_available_atPublic Available Date	Date	ISO 8601. The date the record first became publicly searchable on the state's site.	SOS filing	Public	Always present	The start of the speed-to-contact race. We measure our pull latency from this date, not from formation date — because formation date isn't when the world finds out.
standing_status_as_ofStanding Status As Of	Date	ISO 8601. The date the state last recomputed the standing status. CO surfaces this on every record.	SOS filing	Public	State-dependent · CO only	Detect newly-dissolved entities for win-back campaigns or compliance-services pitches.

Section 03 · Part I

Location & Geography

The state gives you a string. We give you a parsed, geocoded, census-attributed, MSA-tagged, area-code-derived location. Same record. Five additional layers of geographic context.

Field	Type	Description / Values	Source	License	Coverage	Use case
principal_address_rawPrincipal Address (Raw)	Text	Single line. Address string exactly as the state stored it. No parsing, no cleanup.	SOS filing	Public	Always present	Kept for audit and back-reference. If our normalization is wrong, this is what we worked from.
principal_address_normalizedPrincipal Address (Normalized)	JSON object	Shape: `{street, city, state, zip, country}`. USPS-style cleanup; uppercase normalization; secondary unit extraction.	GoodLeads proprietary	Derived	Always present	Drop straight into your address-sensitive workflows (CRM ingestion, dedup keys, mail-merge). No per-state cleanup logic on your end.
principal_cityPrincipal City	Text	City name, normalized to uppercase.	SOS filing	Public	Always present	City-level targeting. First-class filter column on every read path.
principal_statePrincipal State	Text	2-letter USPS state code.	SOS filing	Public	Always present	May differ from formation state on foreign filings — a DE-incorporated business operating in CO would show `DOMESTIC_CO` here and `FOREIGN_DE` in jurisdiction.
principal_zipPrincipal ZIP	Text	5-digit or ZIP+4 format depending on what the state captured.	SOS filing	Public	Always present	ZIP-level targeting and DMA rollup. First-class filter column.
mailing_address_rawMailing Address (Raw)	Text	Where the business receives mail. Often differs from principal address — frequently a personal home.	SOS filing	Public	State-dependent	When the principal address is a virtual office, the mailing address is usually the founder's home — the strongest contact-resolution signal we get.
mailing_address_normalizedMailing Address (Normalized)	JSON object	Same shape as principal_address_normalized.	GoodLeads proprietary	Derived	State-dependent	Parsed mailing address for CRM ingestion.
mailing_cityMailing City	Text	City of the mailing address.	SOS filing	Public	State-dependent	Use for residential-vs-commercial routing when mailing differs from principal.
mailing_stateMailing State	Text	2-letter USPS state code of the mailing address.	SOS filing	Public	State-dependent	Detect out-of-state founders operating in-state — common DE/NV asset-protection pattern.
mailing_zipMailing ZIP	Text	ZIP code of the mailing address.	SOS filing	Public	State-dependent	Use when you need a deliverable address and the principal is a commercial RA building.
latitudeLatitude	Numeric	Decimal degrees, WGS84. Geocoded from the principal address.	US Census Bureau	Public	~92%	Map placement, radius search, drive-time targeting, territory routing.
longitudeLongitude	Numeric	Decimal degrees, WGS84.	US Census Bureau	Public	~92%	Pair with latitude for spatial joins to your own service-area polygons.
geocode_statusGeocode Match Status	Categorical	Values: `match` (clean geocode), `no_match` (address didn't resolve), `partial` (ambiguous match).	US Census Bureau	Derived	Always present	Filter out ambiguous geocodes when location precision matters. `partial` rows have lat/lon but the underlying address was malformed.
county_fipsCounty FIPS Code	Text	5-digit federal county code (2-digit state FIPS + 3-digit county FIPS).	US Census Bureau	Public	~92%	Universal join key into BLS employment data, IRS migration data, FEMA disaster declarations, and any government dataset.
census_tractCensus Tract	Text	11-digit census tract ID (state + county + tract).	US Census Bureau	Public	~92%	Demographic and income-band rollups via ACS 5-year estimates. Highest-resolution public economic data available.
msa_codeMSA Code	Text	5-digit CBSA code from the OMB metropolitan-statistical-area definitions.	GoodLeads proprietary	Derived	Tied to geocode (~92%)	Roll territories up to metros. "All entities in greater Boulder" = one MSA filter, not a list of every Front Range ZIP.
msa_nameMSA Name	Text	Human-readable metro name. Example: `Miami-Fort Lauderdale-Pompano Beach, FL`.	GoodLeads proprietary	Derived	Tied to geocode (~92%)	Drop into reports and emails without a CBSA-to-name lookup table.
area_codeArea Code	Text	3-digit phone area code derived from the primary phone number.	GoodLeads proprietary	Derived	Matches phone coverage	Caller-ID matching for outbound dialer campaigns. Connect rates lift 30–60% when the area code matches the receiver.

Roadmap

Address Mailability — for direct-mail customers

If your motion is a postal piece, you need CASS certification (DPV, DPC, CheckDigit) so the post office honors your bulk-rate discount. We don't ship these fields today. They're on the roadmap; CASS certification adds real per-record cost we want to bake into the right pricing tier before we ship.

If you have a direct-mail motion gated on these fields, tell us — we'll prioritize.

dpv_code Planned Delivery Point Validation result (Y / N / S).
dpv_footnotes Planned Reason codes when DPV partial-match.
dpc Planned 2-digit Delivery Point Code (last two digits of barcode).
check_digit Planned POSTNET / IMb check digit.
cass_certified_at Planned Timestamp of CASS-cycle certification.

Section 04 · Part I

Industry Classification

We don't buy industry codes from a third party. We classify every entity ourselves at load time — rule-based fusion of suffix patterns, keyword matching, and entity-type rules. Confidence ships with every value. And we surface the result at three altitudes: our own GoodLeads taxonomy for sales segmentation, and standardized SIC + NAICS at every granularity for the CRM and BI tools that already speak those codes.

GoodLeads Taxonomy

Built for sales segmentation — the right altitude for outreach

Field	Type	Description / Values	Source	License	Coverage	Use case
industry_sectorIndustry Sector	Categorical	14 sectors. Examples: `Healthcare`, `Construction`, `Technology`, `Real Estate`, `Finance`, `Retail`, `Manufacturing`, `Hospitality`, `Transportation`.	GoodLeads proprietary	Derived	~96%	High-level segmentation for campaign routing. The right altitude for vertical-strategy decisions.
industry_nameIndustry	Categorical	113 industries within sectors. Examples: `Psychology`, `Landscaping`, `Software Development`, `Property Management`, `Accounting`, `Dental Practice`.	GoodLeads proprietary	Derived	~96%	Granular industry. The right altitude for vertical-specific outreach — "we work with dental practices" not "we work with healthcare."
industry_formation_codeIndustry Code	Text	Hierarchical code, sector-then-industry. Examples: `HC-DENT` (Healthcare / Dental), `CON-LAND` (Construction / Landscaping).	GoodLeads proprietary	Derived	~96%	Stable join key for CRM mappings and reporting. Designed to never change even when we expand the taxonomy.
industry_confidence_scoreIndustry Confidence	Numeric	0.0–1.0. Output of the rule-based fusion classifier.	GoodLeads proprietary	Derived	~96%	For teams that want raw control of the threshold. Most teams use the tier instead.
industry_confidence_tierIndustry Confidence Tier	Categorical	Values: `confirmed` (≥0.85), `likely` (≥0.60), `possible` (≥0.40), `unknown` (<0.40).	GoodLeads proprietary	Derived	Always present	The decision rule. Most teams target "confirmed + likely" for outbound; "possible" for nurture; "unknown" for exclusion or human review.
industry_signals_countClassifier Signals	Integer	Number of independent signals the classifier fused into the result.	GoodLeads proprietary	Derived	Always present	Higher = more robust. 4+ signals = bet your campaign on it; 1 signal = single-source, review before scaling.

SIC + NAICS

Standardized codes at every granularity — for the tools that speak them

Every granularity is its own column. We never derive shallower codes from deeper on read — your filter on naics_3_digit = "541" is a real indexed column, not a substring match.

Field	Type	Description / Values	Source	License	Coverage	Use case
sic_2_digitSIC (2-Digit)	Text	Standard Industrial Classification at the division level. Example: `80` (Health Services).	GoodLeads proprietary	Derived	~96%	Coarsest SIC for legacy CRMs that only carry 2-digit. Industry rollups in BI.
sic_3_digitSIC (3-Digit)	Text	SIC at the major-group level. Example: `802` (Offices & Clinics of Health Practitioners).	GoodLeads proprietary	Derived	~96%	Industry-group filtering — common in insurance underwriting and credit scoring.
sic_4_digitSIC (4-Digit)	Text	Most-granular SIC. Example: `8021` (Offices & Clinics of Dentists).	GoodLeads proprietary	Derived	~96%	Full-resolution SIC. Joins to Dun & Bradstreet and other SIC-keyed datasets.
naics_2_digitNAICS (2-Digit)	Text	NAICS sector. Example: `62` (Health Care & Social Assistance).	GoodLeads proprietary	Derived	~96%	Sector-level rollups for BLS / Census joins.
naics_3_digitNAICS (3-Digit)	Text	NAICS subsector. Example: `621` (Ambulatory Health Care Services).	GoodLeads proprietary	Derived	~96%	Subsector filtering for vertical-specific campaigns.
naics_4_digitNAICS (4-Digit)	Text	NAICS industry group. Example: `6212` (Offices of Dentists).	GoodLeads proprietary	Derived	~96%	Industry-group filtering — most common altitude for B2B targeting.
naics_5_digitNAICS (5-Digit)	Text	NAICS industry. Example: `62121` (Offices of Dentists).	GoodLeads proprietary	Derived	~96%	Industry-level filtering. Most CRMs default to 5-digit NAICS.
naics_6_digitNAICS (6-Digit)	Text	Full-resolution NAICS national industry. Example: `621210` (Offices of Dentists).	GoodLeads proprietary	Derived	~96%	Highest-resolution NAICS. Joins to Census Bureau industry-specific datasets.

Section 05 · Part I

Property Intelligence

Where most B2B data products stop, we begin. Every principal address spatial-joined against millions of county parcel boundaries. A residential principal address is a different sales motion than a commercial one — and a vacant-land principal address is usually a shell company you want to exclude. We surface all of that at load time.

Field	Type	Description / Values	Source	License	Coverage	Use case
property_classificationProperty Type	Categorical	Values: `RESIDENTIAL`, `COMMERCIAL`, `INDUSTRIAL`, `AGRICULTURAL`, `MULTI_FAMILY`, `MIXED_USE`, `VACANT`, `EXEMPT`, `UNKNOWN`. Our normalization layered on top of the county source data.	County parcel data → GoodLeads classification	EnrichedDerived	~83%	Distinguish home-based businesses from offices. Filter out shell companies registered at vacant lots. Note: stacked license — the underlying parcel source is enriched; the normalization is ours.
property_classification_rawProperty Type (Raw Code)	Text	The county's unmodified property-use code (varies per county).	County parcel data	Enriched	~83%	When our normalization disagrees with your downstream logic, this is what we worked from.
property_type_detailProperty Subtype	Text	Sub-type from county records. Examples: `Single Family`, `Condo`, `Office Building`, `Warehouse`, `Apartment`.	County parcel data	Enriched	~83%	Finer-grained property targeting — "office building" vs "single-family home" matters for many B2B motions.
property_is_vacantIs Vacant	Boolean	`true` when the parcel is classified as vacant land.	County parcel data → GoodLeads classification	EnrichedDerived	~83%	Strong shell-company signal. Vacant principal address = exclude on most outbound motions.
property_assessed_valueAssessed Value	Integer	USD. County-assessed total value.	County parcel data	Enriched	~83%	Proxy for owner net worth and account size. $3M+ assessed value on a residential property = high-net-worth founder signal.
property_lookup_confidenceProperty Match Confidence	Categorical	Values: `MATCHED` (point-in-polygon), `NEAREST` (best parcel within 50m), `NO_MATCH`.	GoodLeads proprietary	Derived	Always present	For teams that need spatial precision. Filter to MATCHED-only when assessed value drives a decision.
parcel_idParcel ID	Text	County parcel identifier. Format varies per county.	County parcel data	Enriched	~83%	Join key for deeper county-records lookups (ownership history, sales records).
data_vintageProperty Data Vintage	Date	ISO 8601. When the underlying parcel data was published by the county.	County parcel data	Enriched	~83%	How fresh is the assessed value? FL refreshes annually; some CO counties trail 18+ months.

Section 06 · Part II

Registered Agent Intelligence

Proprietary IP. The state gives you a name and address. We give you cross-state volume, market tier, formation-service brand attribution, RA portfolio shape, and a Go-to-Market segment — derived from a curated registry built from SEC filings, BBB records, and acquisition tracking. This is the section that turns "another LLC" into "a $99 ZenBusiness customer who's price-shopping" or "a CT Corporation premium client running a real legal team."

Field	Type	Description / Values	Source	License	Coverage	Use case
registered_agent_nameRegistered Agent Name	Text	RA's legal name as filed. Person or organization.	SOS filing	Public	Always present	The named RA. Anchor for all downstream RA intelligence.
registered_agent_address_rawRA Address (Raw)	Text	RA's address string as filed.	SOS filing	Public	Always present	Used for self-representation detection (RA address == principal address) and address-cluster signals.
registered_agent_address_normalizedRA Address (Normalized)	JSON object	Same shape as principal_address_normalized.	GoodLeads proprietary	Derived	Always present	Parsed RA address.
ra_typeRA Type	Categorical	Values: `P` (Person), `C` (Corporation).	SOS filing	Public	State-dependent · FL only	Person vs. commercial RA service. FL files this explicitly; other states require inference.
ra_organization_nameRA Organization Name	Text	Commercial RA organization name when the RA is a corporation.	SOS filing	Public	State-dependent · CO only	When CO files a commercial RA, the org name is captured separately from the contact name.
ra_self_representedSelf-Represented	Boolean	`true` when the principal address matches the RA address (the owner filed as their own agent).	GoodLeads proprietary	Derived	Always present	Strong "founder-led, no formation service" signal. Best single proxy for "the contact is the decision maker."
ra_is_commercialRA Is Commercial	Boolean	`true` when the RA is identified as a commercial service (any tier).	GoodLeads proprietary	Derived	Always present	First-level split: commercial RA pipeline vs. individual / self-rep pipeline.
ra_market_tierRA Market Tier	Categorical	Values: `premium` ($300+/yr), `midmarket` ($100–200/yr), `budget` ($0–50/yr), `local`, `individual`, `state_agency` (state-default agent — the owner named no RA, e.g. Connecticut’s Secretary of State), `suspicious`, `unknown`.	GoodLeads proprietary	Derived	Always present	Indicates the price tier the founder paid for formation help. Premium = mature, cost-insensitive buyer. Budget = price-shopping DIY founder. Two completely different sales motions.
ra_entity_volumeRA Entity Volume (Cross-State)	Integer	How many entities this RA serves across our entire live dataset.	GoodLeads proprietary	Derived	Always present	Cross-state book size. Premium and budget commercial RAs have huge volumes; an individual self-rep RA shows 1.
ra_address_cluster_countRA Address Cluster Size	Integer	Number of entities sharing this exact RA address.	GoodLeads proprietary	Derived	Always present	Flags shared commercial offices and mail-drops. 50+ = virtual office or mail-drop — don't trust the address for contact resolution.
ra_is_attorneyRA Is Attorney	Boolean	`true` when the RA is identified as an attorney or law firm.	GoodLeads proprietary	Derived	Always present	Premium legal-services formation path. Different motion than DIY founders — these are typically established operators with retained counsel.
is_serial_filerSerial Filer	Boolean	`true` when this person+address combination filed 3+ entities in the same month.	GoodLeads proprietary	Derived	Always present	Serial filer pattern — usually a holding-company or asset-protection structure. Often worth different treatment than first-time founders.
contact_name_in_entity_nameContact Name In Entity Name	Boolean	`true` when the contact's name appears in the business name (e.g., "Smith Consulting LLC").	GoodLeads proprietary	Derived	Always present	Solopreneur signal. Lifts both Contact Relevance and Contact Confidence scores.
gtm_segmentGo-to-Market Segment	Categorical	Values: `owner_operator`, `budget_formation`, `midmarket_formation`, `premium_established`, `suspicious_exclude`, `unclassified`.	GoodLeads proprietary	Derived	Always present	The single most useful field for campaign segmentation. Fuses RA tier + self-rep + jurisdiction + officer count into one buyer persona.

RA Portfolio Shape

What this RA's book of business actually looks like

Beyond volume and price tier — three percentages that describe the kind of clients an RA serves. A budget RA with 95% LLC + 90% single-officer clients is selling to one persona; a premium RA with 30% foreign-filed corporates is selling to a completely different one. Use this when you want to triangulate the buyer persona without waiting for individual contact-level signals.

Field	Type	Description / Values	Source	License	Coverage	Use case
domestic_pctDomestic Filing %	Numeric	0–100. Percentage of this RA's clients formed in the same state where they're registered.	GoodLeads proprietary	Derived	Always present	High = local RA serving local businesses; low = cross-border filing operation. A premium RA with 20% domestic is doing multi-state portfolio work.
llc_pctLLC Mix %	Numeric	0–100. Percentage of clients that are LLCs vs. corporations / LPs.	GoodLeads proprietary	Derived	Always present	Discount RAs skew heavily LLC; corporate-focused premium RAs less so. A signal for "type of buyer this RA attracts."
single_officer_pctSingle-Officer %	Numeric	0–100. Percentage of clients filing with exactly one officer/member.	GoodLeads proprietary	Derived	State-dependent · FL and NY have rich officer data; CO and VA may report 0 — sparse officer data, not a bug.	Proxy for kitchen-table small business vs. real operating entity. A premium RA with 80% single-officer is serving solopreneur professionals (lawyers, consultants).

Section 07 · Part II

Formation Service Attribution

Who did this founder buy their LLC from? LegalZoom? Bizee? TailorBrands? Stripe Atlas? We answer that with a curated registry of 20+ formation services and their RA subsidiaries — evidenced from SEC 10-Ks, acquisition disclosures, and trademark filings. Tells you which competing product the founder already trusts.

Field	Type	Description / Values	Source	License	Coverage	Use case
formation_serviceFormation Service	Categorical	20+ brands incl. `LegalZoom`, `ZenBusiness`, `Bizee`, `TailorBrands`, `Inc Authority`, `Northwest`, `IncFile`, `Stripe Atlas`, `Doola`, `None`.	GoodLeads proprietary	Derived	~30% identifiable	Identify which formation product the founder already trusts. Drives partnership conversations and competitive-displacement plays — "we integrate with LegalZoom" or "we replace ZenBusiness."
ra_providerRA Provider	Categorical	Same value space as `formation_service`. The brand that operates the RA subsidiary.	GoodLeads proprietary	Derived	Always present when RA is a known provider	May differ from `formation_service` when ownership is layered (e.g., Inc Authority → LegalZoom). Useful for understanding the actual operator.
formation_service_confidenceAttribution Confidence	Categorical	Values: `certain`, `very_high`, `high`, `medium`, `low`.	GoodLeads proprietary	Derived	Vendor-conditional · present when formation_service is set	Higher = brand name appears directly in the RA field. Lower = inferred from address co-location with known subsidiaries.
formation_service_attributionAttribution Method	Categorical	Values: `direct` (brand name in RA field), `ra_only` (subsidiary RA, brand inferred).	GoodLeads proprietary	Derived	Vendor-conditional · present when formation_service is set	How we made the call. Direct is bulletproof for outbound copy; ra_only requires confidence-tier filtering before you name the brand to the prospect.

Section 08 · Part III

The Contact

The person you actually call. We never ship info@ aliases or main switchboard numbers — every contact is name-matched to an officer, owner, or self-representing RA. Up to three phones and two emails per contact, ranked by reachability and recency.

Field	Type	Description / Values	Source	License	Coverage	Use case
contact_nameContact Name	Text	Full name. From the state filing when officers are listed; from skip-trace otherwise.	SOS filing · Skip-trace provider	PublicEnriched	Always present	The person you address. Anchors the Contact Relevance and Confidence scores.
role_titleRole / Title	Text	Examples: `CEO`, `President`, `Owner`, `Manager`, `Member`, `MGR`, `P/S`. State-specific abbreviations preserved.	SOS filing	Public	State-dependent	Job title. Drives decision-maker weighting in Contact Relevance. Officer > Owner > Manager > Member.
relationship_typeRelationship to Business	Categorical	Values: `OWNER`, `OFFICER`, `REGISTERED_AGENT`.	GoodLeads proprietary	Derived	Always present	Normalized across state-specific role labels. The cleaner cousin of `role_title`.
phone_primaryPrimary Phone	Text	Currently delivered as a 10-digit North American number (`3035551247`). E.164 standardization (`+13035551247`) is rolling out. Best phone the enrichment pipeline could resolve.	Curated enrichment channel	Enriched	Enrichment-dependent — measured per delivery (see the sample cover sheet)	The number you dial first. Mobile preferred over landline; landline preferred over VoIP. Phone type carried in `phone_type`.
secondary_phoneSecondary Phone	Text	10-digit North American number (E.164 standardization rolling out). Alternate channel discovered by skip-trace.	Curated enrichment channel	Enriched	Vendor-conditional	Second attempt when the primary doesn't connect.
tertiary_phoneTertiary Phone	Text	10-digit North American number (E.164 standardization rolling out). Third channel.	Curated enrichment channel	Enriched	Vendor-conditional	Long-tail attempts. Rarely the right answer but valuable when the first two fail.
email_primaryPrimary Email	Text	Best email address. Already passed through our validation engine — see Email Validation section.	Curated enrichment channel	Enriched	Enrichment-dependent — measured per delivery (see the sample cover sheet)	The address you send to first. Pair with the deliverability fields to gate which messages send.
secondary_emailSecondary Email	Text	Alternate email address.	Curated enrichment channel	Enriched	Vendor-conditional	Useful when the primary is a work address and the secondary is personal (or vice versa).

Section 09 · Part III

Email Validation

Every email runs through our validation engine. We report the deliverability result, the inbox-activity recency, the infrastructure provider, and whether the name on the inbox matches the person on the filing. Most providers ship "valid" or "invalid." We ship the full diagnostic so you can route catch-all, free-provider, and stale-inbox traffic differently than confirmed-active mailboxes.

Field	Type	Description / Values	Source	License	Coverage	Use case
zb_email_statusDeliverability Status	Categorical	Values: `valid`, `catch-all`, `abuse`, `do_not_mail`, `spamtrap`, `invalid`, `unknown`.	Email validation engine	Enriched	Vendor-conditional	Whether the email will deliver. Most teams gate on "valid" + "catch-all" for cold outbound.
zb_email_sub_statusDeliverability Detail	Categorical	Detail behind the status. Common values: `alternate`, `mailbox_not_found`, `greylisted`, `role_based`, `role_based_catch_all`, `possible_trap`, `global_suppression`, `antispam_system`, `alias_address`, `disposable`, `mailbox_quota_exceeded`. List is non-exhaustive; the engine returns ~20 sub-status values total.	Email validation engine	Enriched	Vendor-conditional	Detail behind the status. `role_based` = `info@`-style. `possible_trap` = honeypot; do not send. `alternate` = a better address exists.
email_is_free_providerIs Free Provider	Boolean	`true` for Gmail, Yahoo, Outlook, Apple, AOL, and similar free email providers.	Email validation engine	Enriched	Vendor-conditional	Personal vs. corporate signal. A founder with `name@company.com` behaves differently than one still on `name@gmail.com`.
email_smtp_providerEmail Infrastructure	Categorical	Examples: `google`, `microsoft`, `yahoo`, `apple`, `comcast`, `rackspace`, `zoho`.	Email validation engine	Enriched	Vendor-conditional	Who runs the inbox. Tier-1 providers (Google, Microsoft) deliver more reliably and shape your warm-up strategy.
email_mx_foundMX Records Present	Boolean	`true` when the domain has live MX records.	Email validation engine	Enriched	Vendor-conditional	Prerequisite for delivery. `false` = domain can't receive mail.
email_mx_recordMX Record	Text	Actual MX string (e.g., `aspmx.l.google.com`).	Email validation engine	Enriched	Vendor-conditional	For teams that want to derive infrastructure manually or detect specific patterns.
email_domain_age_daysDomain Age (Days)	Integer	Days since the email domain was first registered.	Email validation engine	Enriched	Vendor-conditional	New domains (<30 days) carry higher spam-trap risk. Older domains route through the standard cold-outbound playbook.
email_activity_foundInbox Activity Detected	Boolean	`true` when the validation engine observed recent inbox activity.	Email validation engine	Enriched	Vendor-conditional	The strongest "is this person reading mail" signal we have. Drives 20 points of the Reachability score.
email_active_in_daysDays Since Last Activity	Integer	Recency buckets: `60`, `90`, `180`, `365`, `365+`.	Email validation engine	Enriched	Vendor-conditional	Recency of inbox activity. 0–60 days = full Reachability points; 365+ = the inbox is effectively dormant.
zb_name_firstInbox Account First Name	Text	First name registered on the inbox account (returned by the validation engine).	Email validation engine	Enriched	Vendor-conditional	Compared against the filing's officer/owner name to compute name verification.
zb_name_lastInbox Account Last Name	Text	Last name registered on the inbox account.	Email validation engine	Enriched	Vendor-conditional	Same role as the first-name field.
email_name_verificationName vs. Filing Match	Categorical	Values: `FULL_MATCH`, `LAST_MATCH`, `FIRST_MATCH`, `NO_MATCH`, `NO_DATA`.	GoodLeads proprietary	Derived	Vendor-conditional	Does the name on the inbox match the name on the filing? Catches forwarded inboxes, family-shared addresses, and impersonation. Drives 10 points of Reachability and is a major Contact Confidence modifier.

Section 10 · Part III

Phone Validation

Every phone is classified by line type before we ship it. Mobile, landline, and VoIP have different connect rates, different TCPA implications, and different best-time-to-call patterns. You should treat them differently — we make that possible.

Field	Type	Description / Values	Source	License	Coverage	Use case
phone_validityPhone Validity	Categorical	Values: `MOBILE_VALID`, `LANDLINE_VALID`, `VOIP_VALID`, `INVALID`, `UNKNOWN`.	Curated enrichment channel	Enriched	Vendor-conditional	Line type and validity. Mobile has the highest connect rate; VoIP signals lower intent. Drives 20 points of Reachability.
fallback_phoneFallback Phone	Text	10-digit North American number (E.164 standardization rolling out). Phone discovered by the fallback enrichment channel when the primary skip-trace missed.	Fallback enrichment channel	Enriched	Vendor-conditional · fallback only	Recovers ~15% of records that primary skip-trace couldn't find.
phone_reported_in_servicePhone Reported In Service	Boolean	`true` when the data source reported this line in service at last observation — a source claim, not an independently verified property.	Fallback enrichment channel	Enriched	Vendor-conditional	Treat as a weak positive signal, not a liveness guarantee. Recency (Phone Last Seen) is the stronger signal.
phone_last_seenPhone Last Seen	Date	ISO 8601. Most recent date the number was reported active.	Fallback enrichment channel	Enriched	Vendor-conditional	Recency filter. Sub-90 days = high-confidence current; 12+ months = stale, may be reassigned.

Section 11 · Part IV

Reachability Score

0–100 composite answering one question: can we reach this person? Every component ships alongside the score in the parameters payload — see exactly how it was built and tune your own thresholds. Version: reachability_v1.0.

Field	Type	Description / Values	Source	License	Coverage	Use case
reachability_scoreReachability Score	Integer	0–100. Composite of 8 signals (7 core + 1 bonus). Capped at 100.	GoodLeads proprietary	Derived	Present for scored records	The headline score. Tunable threshold for cohort selection.
reachability_tierReachability Tier	Categorical	Values: `On Fire` (80–100), `Very Hot` (60–79), `Hot` (40–59), `Warm` (20–39), `Cold` (0–19), `Withheld` (score suppressed for a small set of records).	GoodLeads proprietary	Derived	Present for scored records; a small set is Withheld	Named tier — drop directly into a CRM lifecycle stage without further work. The tier is what your reps see.

Score Components

Shipped in the parameters payload

Component	Max points	What it measures
email_validity_pts	25	Deliverability — valid and catch-all score highest.
email_activity_pts	20	Inbox activity recency. 0–60 days = full points.
phone_reach_pts	20	Phone availability and type. Mobile > landline > VoIP.
geo_intel_pts	10	Geocode precision — lat/lon + county + census tract all present.
name_verify_pts	10	First/last name match between inbox account and filing.
identity_conf_pts	10	Skip-trace identity-confidence signal.
email_infra_pts	5	Top-tier email provider + MX records present.
industry_bonus_pts	5	Industry classification confidence bonus.

Hard caps — honest about what we can't promise

UNREACHABLE_CAP = 20. No phone AND no valid/catch-all email caps the score at 20 (Cold) — no amount of geocode or industry signal can rescue a record with no channel to reach the human.

PHONE_ONLY_CAP = 79. A great phone with no valid email maxes at Very Hot (never On Fire). On Fire requires multi-channel confirmation.

Never-Downgrade Protection

Once a record reaches a score, a subsequent reload cannot lower it. Paid enrichment is irreversible — the data is too. A record that hit Very Hot in February stays at Very Hot even if the score model re-evaluates lower on a re-run.

Section 12 · Part IV

Contact Relevance Score

0–100 composite answering an orthogonal question: is this the RIGHT person to sell to? A contact can be highly reachable but the wrong person — a commercial RA employee instead of the founder. This score separates them. Version: contact_relevance_v1.1.

Field	Type	Description / Values	Source	License	Coverage	Use case
contact_relevance_scoreContact Relevance Score	Integer	0–100. Composite of 7 weighted signals + 1 interaction penalty.	GoodLeads proprietary	Derived	Present for scored records	Decision-maker proxy. The headline score.
contact_relevance_tierContact Relevance Tier	Categorical	Values: `Decision Maker` (85–100), `Likely Decision Maker` (65–84), `Probable Contact` (45–64), `Uncertain Contact` (25–44), `Unlikely Decision Maker` (0–24).	GoodLeads proprietary	Derived	Present for scored records	Named tier. Most teams target Decision Maker + Likely for outbound; Probable for nurture; Uncertain and below for exclusion.

Score Components

Shipped in the parameters payload

Component	Max points	What it measures
self_representation_pts	25	Self-rep RA + individual RA = founder filing alone. Strongest single signal.
contact_name_source_pts	20	Officer (20) > Owner (15) > RA person (12) > Commercial RA (5) > Attorney RA (3).
location_uniqueness_pts	20	How many entities share this lat/lon. 1=20, 2=15, 3–5=10, 6–20=5, 21+=0.
contact_exclusivity_pts	15	How many entities this contact links to. 1=15, 2–3=10, 4–10=5, 11–50=2, 51+=0.
name_in_entity_pts	10	Person's name appears in the business name (e.g., "Smith Consulting LLC").
ra_tier_pts	5	Individual RA scores higher than commercial.
entity_type_pts	5	Domestic LLC + 1 officer = highest.
interaction_penalty	−10	Applied when contact appears on 4+ entities AND 6+ entities share the same location. Catches the commercial-RA-office pattern that would otherwise game the score.

Section 13 · Part IV

Contact Confidence Score

0–100 composite answering a third orthogonal question: are we confident this email or phone really belongs to this person? A reachable, relevant contact can still be the wrong identity — a forwarded inbox, a family-shared phone, a name collision. This score isolates that risk. Five weighted signals (100 points) plus three additive modifier blocks. Floor of 5; never zero. Version: contact_confidence_v2.1.

Field	Type	Description / Values	Source	License	Coverage	Use case
contact_confidence_scoreContact Confidence Score	Integer	5–100. Floor of 5 (every contact has some probability of being real). Capped at 100 after modifiers.	GoodLeads proprietary	Derived	Present when the contact has been enriched/validated (null otherwise)	Identity-confidence headline. Pair with Reachability + Relevance for a three-axis lead-quality picture.
contact_confidence_tierContact Confidence Tier	Categorical	Values: `Verified Contact` (75–100), `Likely Contact` (50–74), `Possible Contact` (25–49), `Uncertain Contact` (5–24).	GoodLeads proprietary	Derived	Present when the contact has been enriched/validated (null otherwise)	Named tier. Most teams gate outbound at Verified + Likely.

Core Signals

Five weighted signals · sum to 100

Component	Max points	What it measures
email_name_match	30	Does the email local-part contain name tokens that match the filing? Strongest single identity signal.
zb_quality	20	Validation status + recent inbox activity (someone is actually reading this mailbox).
phone_ra_alignment	20	Mobile phone + self-represented RA = the founder's personal line. Strongest paired identity signal.
property_context	15	Residential principal address raises confidence (vs. shared commercial / virtual-office address).
ra_self_rep	15	Self-rep RA = the contact almost certainly IS the founder, not an employee or RA service rep.

Modifier · Cross-Source Verification

Additive · range −8 to +13

Cross-checks the name on the inbox against the name on the filing, plus a bonus when the contact's name appears in the business name. Independent identity verification beyond the core signals.

Component	Points	When it applies
email_name_verification: FULL_MATCH	+8	First and last name on inbox match the filing.
email_name_verification: LAST_MATCH	+5	Last name matches; first name differs.
email_name_verification: FIRST_MATCH	+3	First name matches; last name differs.
email_name_verification: NO_MATCH	−8	Inbox name differs from filing on both axes — likely forwarded inbox or impersonation.
contact_name_in_entity_name	+5	Contact's name appears in the business name.

Modifier · Deliverability Detail

Additive · range −5 to +2

Penalizes risky deliverability sub-statuses; rewards confirmed alias addresses. Catches honeypots, role-based mailboxes, and global-suppression hits before they damage your sender reputation.

Sub-status	Points	What it indicates
possible_trap	−5	Honeypot / spam trap. Heaviest penalty.
role_based_catch_all	−5	Role-based mailbox at a catch-all domain.
global_suppression	−4	Address on the global suppression list.
role_based	−3	Role mailbox (info@, support@). Not a person.
antispam_system	−2	Aggressive filtering at the destination.
alias_address	+2	Confirmed forwarding alias to a verified inbox.

Modifier · Enrichment Strategy

Additive · 0 to +10

Rewards records joined by the strongest enrichment path. How the contact was matched is itself a confidence signal — name + address resolution is more trustworthy than reverse-address-only lookup.

Strategy	Points	What it means
name_and_address	+10	Strongest join: both name and address independently confirmed.
mailing_fallback	+6	Resolved via secondary mailing address.
address_only	+2	Reverse-address lookup only; no name match.
hold / unknown	0	No enrichment performed.

Three Orthogonal Scores

Reachability answers "can we reach them?" · Relevance answers "are they the right person?" · Confidence answers "is this contact really them?" A lead can be high on any two and low on the third. Most providers ship one score (or none). We ship all three with full component-level transparency.

Section 14 · Part V

Pre-Enrichment Strategy

Before we spend a single enrichment credit, our Contact Intelligence engine decides how to enrich each entity — and whether it's worth enriching at all. This section explains why some leads got the deep treatment, why others got the cheap path, and why some were held back entirely. Transparency on the routing IP that drives our fill rates.

Field	Type	Description / Values	Source	License	Coverage	Use case
enrichment_strategyEnrichment Strategy	Categorical	Values: `name_and_address` (strongest join), `address_only` (reverse lookup), `mailing_fallback` (when principal address is unusable), `hold` (don't enrich).	GoodLeads proprietary	Derived	Always present	The routing decision. Held records are not enriched — this is how we control quality at the entry, not at the output.
name_confidenceName Confidence	Categorical	Values: `high`, `medium`, `low`, `unknown`.	GoodLeads proprietary	Derived	Always present	How confident we are that the contact name is a real, unique person — not a placeholder like "Owner" or "Manager."
address_confidenceAddress Confidence	Categorical	Values: `high`, `medium`, `low`, `unknown`.	GoodLeads proprietary	Derived	Always present	How confident we are that the address is a real, single-occupant location — not a virtual office or commercial RA building.
name_densityName Density	Integer	Count of entities sharing the same contact name across the dataset.	GoodLeads proprietary	Derived	Always present	High = generic name (John Smith) — lower identity confidence; 1 = uniquely identifiable.
address_densityAddress Density	Integer	Count of entities at the same principal lat/lon.	GoodLeads proprietary	Derived	Always present	1 = real address; 50+ = mail-drop, virtual office, or commercial RA building. Drives the hold decision for non-operating archetypes.
cross_state_contact_countCross-State Contact Count	Integer	1–4. Number of live states this contact name appears in.	GoodLeads proprietary	Derived	Always present	Multi-state presence = strong serial-founder or professional-filer signal.
entity_archetypeEntity Archetype	Categorical	Values: `operating_business` (the default — sellable), `mail_drop_or_formation_suite`, `single_asset_property_llc`, `asset_protection_holding`. The last three short-circuit to `hold`.	GoodLeads proprietary	Derived	Always present	The headline classification. Operating businesses get enriched, scored, and sold. The three non-operating archetypes get held back — see Sellable Inventory next.
archetype_reasonArchetype Reason	Text	Plain-English explanation of why the archetype was assigned. Example: `Single-officer LLC, residential principal address, self-represented RA`.	GoodLeads proprietary	Derived	Always present	Human-readable rationale. If your team wants to override our hold decision on a specific cohort, this tells you what signals to argue against.

Section 15 · Part V

Sellable Inventory

Most data products ship everything they pulled. We don't. We split inventory into "sellable" (operating businesses worth selling to) and "non-operating" (mail-drops, single-asset-property LLCs, asset-protection holdings). You see exactly how many of each. You're never charged for held records.

This section is the transparency layer behind entity_archetype — it tells you how much of each state's volume actually rolls into your campaign-ready inventory, vs. how much we set aside because the buyer-quality wasn't there.

Field	Type	Description / Values	Source	License	Coverage	Use case
sellableSellable	Boolean	`true` when entity_archetype is `operating_business`. Available as a first-class filter on Browse and exports.	GoodLeads proprietary	Derived	Always present	The default filter on every workflow. Turn it off only when you want to look at what we held back and why.
non_operatingNon-Operating	Boolean	`true` when entity_archetype is one of the three held categories.	GoodLeads proprietary	Derived	Always present	Use when your motion is compliance, dissolution, or asset-protection services — populations we deliberately exclude from the main inventory.
sellabilitySellability Filter	Categorical	UI filter. Values: `sellable` (default), `non_operating`, `all`.	GoodLeads proprietary	Derived	Always present	Three-way switch on the Browse page and exports. Lets your analyst look at the held cohort to disagree with our routing.

Per-State Aggregates

How big is each state's actual sellable inventory?

Pre-computed totals per state. Surfaces on the dashboard and on every state-summary API call. The phone / email fill-rate denominators on the Coverage table are measured against the sellable cohort — anchoring fill rates against held records would inflate the numbers and mislead.

Field	Type	Description / Values	Source	License	Coverage	Use case
sellable_totalSellable Total	Integer	Count of operating-business records in the state.	GoodLeads proprietary	Derived	Always present	Your campaign's denominator. The number we'll bill against on volume tiers.
non_operating_totalNon-Operating Total	Integer	Count of held records (sum of the three non-operating archetypes).	GoodLeads proprietary	Derived	Always present	How much of the state's pull we held back. Auditable — you can request the held cohort and disagree.
sellable_with_phoneSellable With Phone	Integer	Count of sellable records with a primary phone present.	GoodLeads proprietary	Derived	Always present	Phone fill-rate numerator. This is the number anchored to our 45% phone SLA, not the count across all pulled records.
sellable_with_emailSellable With Email	Integer	Count of sellable records with a primary email present.	GoodLeads proprietary	Derived	Always present	Email fill-rate numerator, anchored to the 35% email SLA.

Honest disclosure

Our "looks like an address" heuristic flagging single-asset-property LLCs has a known ~24% false-positive rate on low-single-digit-prefix business names ("2 Pups Hauling", "4K's Trucking"). We're tightening it before we expand the rule. If your team flags a held record you believe should be sellable, the archetype_reason field will tell you exactly which signal caught it.

Section 16 · Part V

Fallback Enrichment Channel

When the primary skip-trace channel can't find a contact, the fallback channel runs second. These fields only appear on records where the fallback ran. Recovers roughly 15% of records the primary missed — the difference between a 45% and a 52% phone fill rate.

Field	Type	Description / Values	Source	License	Coverage	Use case
identity_match_confidenceIdentity Match Confidence	Integer	0–100. The fallback provider's own identity-match confidence.	Fallback enrichment channel	Enriched	Vendor-conditional · fallback only	Average ≥96 on matches — the provider is conservative about returning a match in the first place.
fallback_emailFallback Email	Text	Best email from the fallback channel when the primary missed.	Fallback enrichment channel	Enriched	Vendor-conditional · fallback only	Already passed through email validation before delivery.
estimated_ageEstimated Age	Integer	Reported age of the contact in years.	Fallback enrichment channel	Enriched	Vendor-conditional · fallback only	Useful for life-stage targeting (retirement-planning, first-business-owner programs, etc.).

Section 17 · Part VI

Provenance & Audit

Every value in this dictionary ships with five pieces of audit metadata at the field level. Source, license, timestamp, derivation method, and service version. Available on every record on every API call. If a customer or compliance officer ever asks "where did this come from and when did you learn it?" — every cell has an answer.

Field	Type	Description / Values	Source	License	Coverage	Use case
source_nameSource (per attribute)	Text	Identifies which channel produced this specific value. Generic descriptors (SOS filing, Email validation engine, GoodLeads proprietary), not vendor brands.	GoodLeads proprietary	Derived	Always present	Per-field, not per-record. Most vendors only timestamp the row. We timestamp every cell.
license_flagsLicense (per attribute)	Categorical	Values: `PUBLIC_RECORD`, `ENRICHED`, `COMMERCIAL`, `DERIVED`. Stacked values supported for derivations on enriched inputs.	GoodLeads proprietary	Derived	Always present	Drives compliance, resale, and residency rules. Your legal team filters by this when deciding what to redistribute.
observed_atObserved Timestamp (per attribute)	DateTime	ISO 8601 with timezone. When this specific value was captured or last refreshed.	GoodLeads proprietary	Derived	Always present	Field-level freshness. Filter exports by "values observed in the last 30 days" without re-pulling whole records.
methodDerivation Method	Text	Versioned method identifier. Examples: `geocode_batch_v1`, `address_comparison_v1`, `name_compare_v1`, `rule_based_fusion_v1`.	GoodLeads proprietary	Derived	Always present (derived fields)	How we computed this value. Lets your team reason about model drift — and lets us ship a v2 method on a specific field without invalidating the v1 history.
attr_versionAttribute Version	Text	Semver of the service that wrote this attribute. Example: `ra_intelligence_v1.2`.	GoodLeads proprietary	Derived	Always present (derived fields)	When we ship a v2 of a brain, you can identify which rows still carry v1 output and re-run if you need consistency.
score_versionScore Version	Text	Semver of the scoring algorithm. Current values: `reachability_v1.0`, `contact_relevance_v1.1`, `contact_confidence_v2.1`.	GoodLeads proprietary	Derived	Always present	The scoring algorithm version that produced this record's scores. Reason about model drift; re-score on demand.

Reference · Sample

Sample Record

One fully-enriched lead. The shape, sparsity, and value patterns your pipeline will actually see. Use this to size your ingestion code, mock your CRM mapping, and confirm the deliverable matches what your team needs before any contract conversation.

Download sample CSV Copy JSON to clipboard

{ "lead_ref": "GL-CO-04829", "contact_ref": "GLC-12873", "state_entity_id": "20261234567", "source_system": "CO_SOS", // — Business Profile — "entity_name": "MOUNTAIN VIEW DENTAL LLC", "entity_type": "LLC", "formation_date": "2026-04-15", "status": "Active", "jurisdiction": "DOMESTIC_CO", "officer_count": 1, "public_available_at": "2026-04-16", // — Location & Geography — "principal_address": { "raw": "1422 ELK PARK DR, BOULDER, CO 80303", "normalized": { "street": "1422 ELK PARK DR", "city": "BOULDER", "state": "CO", "zip": "80303" }, "latitude": 39.985, "longitude": -105.197, "county_fips": "08013", "census_tract": "08013012204", "msa_code": "14500", "msa_name": "Boulder, CO", "area_code": "303" }, // — Industry Classification — "industry": { "sector": "Healthcare", "name": "Dental Practice", "formation_code": "HC-DENT", "sic_4_digit": "8021", "naics_6_digit": "621210", "confidence_score": 0.94, "confidence_tier": "confirmed" }, // — Property Intelligence — "property": { "classification": "COMMERCIAL", "type_detail": "Medical Office", "is_vacant": false, "assessed_value": 1245000, "lookup_confidence": "MATCHED", "parcel_id": "146325104017" }, // — Registered Agent Intelligence — "registered_agent": { "name": "JANE SMITH", "self_represented": true, "is_commercial": false, "is_attorney": false, "market_tier": "individual", "entity_volume": 1, "address_cluster_count": 1, "gtm_segment": "owner_operator" }, // — Contact — "contact": { "name": "JANE SMITH", "role_title": "Manager", "relationship_type": "OWNER", "phone_primary": "3035551247", "secondary_phone": "3035554089", "tertiary_phone": null, "email_primary": "jane.smith@mountainviewdental.com", "secondary_email": "jsmith@gmail.com" }, // — Email Validation — "email_validation": { "deliverability_status": "valid", "is_free_provider": false, "smtp_provider": "google", "mx_found": true, "domain_age_days": 31, "activity_found": true, "active_in_days": 60, "name_verification": "FULL_MATCH" }, // — Phone Validation — "phone_validation": { "validity": "MOBILE_VALID" }, // — The Three Scores — "scores": { "reachability": { "score": 87, "tier": "On Fire", "version": "reachability_v1.0" }, "contact_relevance": { "score": 88, "tier": "Decision Maker", "version": "contact_relevance_v1.1" }, "contact_confidence": { "score": 89, "tier": "Verified Contact", "version": "contact_confidence_v2.1" } }, // — Pre-Enrichment Strategy & Inventory — "enrichment_strategy": "name_and_address", "entity_archetype": "operating_business", "sellable": true, // — Provenance — "provenance": { "freshest_observation": "2026-04-16T08:24:31Z", "stalest_observation": "2025-12-01T00:00:00Z" } }

Appendix A

Coverage by State

Observed fill rates from production runs (refreshed monthly), anchored to the sellable cohort — not all pulled records. Phone and email rates measured against operating businesses only, because those are the records we expect you to charge against. Forecasts for newly-onboarded states are dashed and prefixed with "~" so they're never mistaken for measurement.

Solid cell = data-status="observed" · production fill rate, sellable cohort Dashed cell = data-status="forecast" · projection · not yet measured

Stage / Source	CO	FL	VA	NY	GA next
SOS filing fields	100%	100%	100%	100%	100%
Sellable inventory (% of pull)	78%	82%	84%	71%	—
Geocoding	90%	92%	93%	92%	—
MSA tagging	89%	91%	92%	90%	—
Industry classification (GoodLeads + SIC + NAICS)	95%	96%	96%	96%	—
Property classification	76%	88%	84%	80%	—
RA intelligence	100%	100%	100%	100%	—
Phone fill (sellable cohort)	56%	64%	87%	51%	45%
Email fill (sellable cohort)	41%	53%	59%	38%	35%
Email validation	48%	48%	48%	48%	—
Reachability score	100%	100%	100%	100%	—
Contact Relevance score	100%	100%	100%	100%	—
Contact Confidence score	100%	100%	100%	100%	—

Appendix B

What's Different

A scraper ships you state data with a timestamp. We ship you that — plus seven things no other B2B data product delivers.

— 01

Three Orthogonal Scores

Reachability ("can we reach them?"), Contact Relevance ("are they the right person?"), Contact Confidence ("is this contact really them?"). Plus 15+ derived classifications — Industry, Property Type, RA Market Tier, GTM Segment, Formation Service, RA Portfolio Shape, Density signals, Entity Archetype, Enrichment Strategy. All computed by us.

— 02

Sellable vs. Held Inventory

We explicitly hold back records that won't convert — mail-drops, single-asset property LLCs, asset-protection holdings. You're never billed for held records. Competitors ship them all and let you eat the bounce.

— 03

Cross-State Entity Resolution

Same RA, same contact — identified across CO, FL, VA, NY at load time. Without you doing any joins.

— 04

Per-Field Provenance

Every value carries source, license, timestamp, method, and version. Audit any cell. Most vendors ship one timestamp per record. We ship five per field.

— 05

Per-State Semantic Mapping

"What does mailing address mean in NY vs. FL vs. CO?" We've solved that. Your CRM sees one canonical schema across 50 state formats.

— 06

Confidence on Every Derivation

No value ships without a tier or score telling you how much to trust it. No black boxes. Hard caps and never-downgrade rules documented in the score sections.

— 07

Field-Level License Classification

PUBLIC_RECORD, ENRICHED, DERIVED — plus stacked-license support for derivations on enriched inputs. Your compliance team can reason about residency, sharing, and resale at the cell level.