Skip to content

NewUnlimited plans are here. Fetch, search and enrich without counting credits.See plans

Reverse Contact

Our Data Quality Standard

No data over bad data.

We would rather return nothing than return something wrong. This page explains what that means in practice: how we collect, validate, refresh and deliver every profile, email and phone number.

Book a demoSee Documentation
verified work-email accuracy
99.3%
profile refreshes per month
200M
highly verified profiles, out of 1B+ collected
450M
charged when we return nothing
0 credits

Principles

Three commitments behind every response

  1. Freshness you can see

    Every response carries the date of the last full profile refresh, not a light check on a few fields. Read it for free with /check before you decide how to resolve.

  2. Quality by construction

    Validation runs at sourcing, at refresh and at delivery. Bots, fakes and unverified records are kept out of the database. They never reach your product.

  3. Matching precision

    We never return a guess. Identity resolution runs every check we have and returns a confidence-scored match, or null.

The data layer

We operate the whole pipeline, end to end

Since 2024 we no longer source data: we run the full infrastructure ourselves. That's what lets us guarantee what comes out of it.

  1. Live collection

    Publicly available professional data, sourced by us

  2. Data validation

    Bot & fake filtering, consistency checks, batch QA

  3. On-demand refresh

    Tiered schedule plus live re-sourcing

  4. Entity mapping

    People ↔ companies ↔ domains ↔ social IDs

  5. Mapping validation

    Confidence-scored identity resolution

  6. Distribution

    API, MCP, CLI, CSV: same schema everywhere

Validation layer

We don't resell data. We validate it.

Where we use external sources, we never trust their flags. Every result passes our own checks before it reaches you.

/email

Deliverability checks

  • MX records resolvedpass
  • SMTP handshakedeliverable
  • Catch-all detectionflagged
  • Cross-source name/domain consistencymatch

Validated in real time, before delivery. Waterfall providers re-benchmarked monthly against ground truth.

/phone

Ownership checks

  • Number existence & carrieractive
  • WhatsApp / iMessage presencefound
  • Partial-digit cross-checksconsistent
  • OSINT cross-validationowner ✓

The number exists and belongs to the person, not just one of the two.

profiles

Identity checks

  • Strict identity resolution0.96
  • Bot & fake-account filteringupstream
  • Consistency vs previous versionpass
  • Sampling & QA per sourcing batchpass

Below the confidence threshold, the answer is null, never a best guess.

Verified profiles only

Why our numbers look smaller in benchmarks

Most providers count every record they hold. We only return profiles that passed verification. Fake accounts and bots are blocked upstream, so our coverage figures are lower, and our match is right.

Example · search “John Doe, Microsoft”

  • John Doe · Microsoftbot
  • John Doe · Microsoftfake
  • John Doe · Microsoftverified
  • John Doe · Microsoftbot

Keep them all and every match is a coin flip. We remove them, so the match is the real one.

Refresh engine

Two mechanisms keep the database fresh

Freshness is the hardest problem in B2B data. We don't solve it with a “30-day refresh” claim. We solve it with usage and with scoring.

Live API traffic

Every live person or company call re-fetches the source and updates the profile in our database. Usage keeps the data current by design.

Opt-out available by contract if you don't want your calls used for refresh.

Scored, tiered refresh

Every profile is scored on demand signals, job title and completeness, then assigned a refresh tier.

Refresh tiers
Tier 1High demandevery 30 days
Tier 2Active profilesevery 90 days
Tier 3Long tailon demand via API

200Mprofile refreshes per month, with a last-updated timestamp on every response

Resolution modes

You decide how fresh is fresh enough

Freshness is a trade-off with latency and cost. We expose it instead of hiding it: check the date for free, then choose per request.

/check

Free

Send an ID or profile URL, get our last-updated date. Decide: cache, live, or skip.

"exists": true,
"experienceCount": 8,
"skillCount": 24,
"schoolCount": 3,
"lastUpdate": "2026-08-30T09:12:44.000Z"

cache

from our database

The profile as stored, with its timestamp. Best for volume and latency-sensitive features.

Mode
Synchronous
Latency
~500 ms P95

live

re-sourced now

Goes back to the source, re-validates, updates our database, returns the new version. For stale records and high-stakes decisions.

Mode
Async · webhook
Freshness
Real-time

How to benchmark us

Rank providers by precision, not by count

Most waterfalls rank providers by match rate because quality is hard to measure. But a high-count provider placed first pushes its false positives straight into your product. Put the most precise provider first, then let lower-confidence sources fill the gaps if you want more coverage.

Ranked by count

  1. Provider A92% · 71% right
  2. Provider B80%
  3. Provider C64%

Ranked by precision

  1. Provider C99% right
  2. Provider Bfills gaps
  3. Provider Alast resort

Compliance

Compliance-by-design. Built to be defensible.

Reverse Contact is designed to support GDPR/CCPA-aligned data practices. We maintain clear sourcing principles, respect data subject rights, and prioritize accuracy over guesswork.

  • GDPR
  • CCPA
Book a demo Check our Trust Center

Don't take our word for it. Benchmark us on your records.

500 free credits, 7 days, no card. Run a sample where you already know the truth and measure precision, not just fill rate.

Start for freeBook a demo
Profiles
1B+
Refreshes / month
200M
Response time (P95)
500ms