Our data

Data that shows its work.

FetchSet is business-identity infrastructure, not a mystery field. Every useful record is designed to carry freshness, validation risk, match confidence, provenance, and exact cost.

inspect a record review availability

business-email-firstfreshness is a dateconfidence stays separatemisses cost zero

The difference

Specific evidence beats a universal claim.

“100% verified” sounds simple, but it hides several different questions. FetchSet keeps those questions visible so a person or agent can apply the right policy.

01 / Validation

Verification belongs to a field

An email receives a deliverability classification and validation date. That does not automatically make every other field correct.

02 / Identity

Confidence is not deliverability

A strong person match and a valid mailbox are separate signals. The response preserves both instead of blending them.

03 / Freshness

Freshness is inspectable

Records carry when a value was observed or validated. Stale values are rechecked or withheld according to policy.

04 / Cost

The charge travels with the result

Cache and external resolution paths remain distinguishable. Empty, suppressed, and not-found results report zero tokens.

The record contract

A value, plus what an agent needs to decide.

The planned response contract keeps identity, contact quality, lineage, and billing evidence together. Callers can accept, review, retry, or stop without consulting another dashboard.

read why the contract matters →

response.jsonillustrative example

{
  "person": {
    "name": "Avery Lin",
    "linkedin_url": "linkedin.com/in/avery-lin",
    "company": "Northstar",
    "title": "VP Operations"
  },
  "best_email": "avery@northstar.dev",
  "risk": "valid",
  "validated_at": "2026-07-09",
  "identity_confidence": 0.984,
  "source_class": "licensed_cache",
  "tokens_charged": 0
}

Identity

Name, role, company, professional URL

Contact quality

Email type, risk, and validation time

Lineage

Source class, observation time, and policy

Economics

Exact charge and resolution path

Availability

One roadmap. Three honest states.

Availability follows data rights, validation quality, economics, and compliance. It does not follow the desire to make a longer feature list.

01 / Launch wedge Business-email-first

Core business identity

LinkedIn URL → business email
Name + company → business identity
Email ↔ professional profile resolution
Core person and company firmographics
Validation, freshness, confidence, provenance

US-first and conditional on signed field-level resale rights.

02 / Guarded expansion Phase 2

Higher-risk or broader coverage

Personal email
Carrier-validated mobile numbers
People and company search
IP address → company
Batch and higher-throughput workflows

Released only after licensing, validation, cost, and compliance gates close.

03 / Permanent no-go Disabled

Identity-graph advertising data

Mobile advertising IDs
AAID or IDFA resolution
Hashed-email identity graphs
Ad-audience append products
Special-category personal data

These remain outside FetchSet’s strategy and return 403 feature_disabled.

Sources and freshness

Licensed first. Freshness made visible.

FetchSet’s searchable cache depends on a signed redistribution license. Finder output is not silently converted into a resold database. External providers may resolve a miss, but the response still records how the value was produced.

01

Licensed seed

Searchable business records enter the cache only when contracts permit redistribution for the relevant fields.
02

Fresh cache

A cached value is served only while it meets the applicable freshness and suppression policy.
03

Verified external hit

A missing or stale value can enter a bounded provider waterfall that stops at the first acceptable hit.
04

Structured miss

Not found and suppressed are legitimate outcomes. They remain empty, explainable, and zero-charge.

Coverage

We will measure it before we market it.

Coverage changes with geography, role, company size, and the identifier you start with. FetchSet will not publish a universal match rate or a provider count before the representative vendor bake-off is complete.

read the data FAQ

Planned launch bake-off1,000 representative records

Dimension	What we will measure
Match rate	By endpoint
Validity	By risk class
Freshness	Stale-rate
Economics	Cost per hit
Performance	p50 / p95
Rights	Field by field

No customer quote or “20+ provider” claim appears here until there is evidence and permission to publish it.

The right question is whether the data works for your workflow.

join founding beta read the docs