The No-Guessed-Contacts Review Framework

A free, vendor-neutral way to tell whether a B2B contact record is verified, inferred, stale, or guessed — five checks, a scoring system, and a worked example.

The problem

Most "verified" contact data isn't what buyers assume it means. A contact passing an email-deliverability check (the address is validly formatted and the receiving mail server accepts it) is a different claim from a contact being identity-confirmed (a real, current, public source says this person holds this role at this company, as of a specific date). Vendors routinely sell the first as if it were the second. The gap between the two is where stale titles, departed employees, and pattern-guessed emails (first.last@company.com generated from a formula, not confirmed) end up in a list you paid for.

This framework gives a business buyer a repeatable way to tell which one they actually have, for any single contact record, from any source — a purchased list, a database export, internal research, or a delivered custom research batch.

Definitions

This framework classifies a contact record's supporting evidence into one of six categories:

• Verified: A specific, dated, public source confirms this person currently holds this role at this company. The source is checkable by the buyer, not just asserted by the vendor.
• Inferred, clearly labeled: The record states a reasonable inference and says so — not presented as confirmed fact.
• Stale: The supporting source exists but is dated; the role, company, or contact detail may no longer be current.
• Incomplete: Some fields are confirmed; others are missing rather than filled in with a guess.
• Contradictory: Two or more public sources disagree on a material fact and the disagreement hasn't been resolved or flagged.
• Guessed: No public source supports the specific detail. Most commonly: an email address generated from a name-pattern formula with no confirmation it's real or in use.

Review sequence

1. Identify the specific claim being made about this contact.
2. Locate the source cited for that claim.
3. Check whether the source is public and independently checkable.
4. Check the source's date.
5. Check for a second source on the same fact; note agreement or contradiction.
6. Classify the record using the definitions above.
7. Decide whether the record is usable as-is, usable with a caveat, or should be excluded.

Source-quality hierarchy

This framework ranks source types from highest to lowest confidence:

1. The company's own current website, staff page, or press release naming the person in the role.
2. A recent, credible news article or interview naming the person in the role.
3. A professional profile the person appears to control, checked for recent activity.
4. A directory or database listing with no visible original source.
5. An inferred pattern with no source at all.

For this framework, Levels 1–2 support classifying a record as "verified." Level 3 alone supports "verified with lower confidence." Levels 4–5 do not support "verified" on their own.

Freshness/date-check standard

A source older than 12 months should be treated as stale until re-checked, regardless of how confident it looked when first found. Roles and employers change in ways a single snapshot source can't keep up with — a 2-year-old staff page is a starting point, not current proof.

Handling conflicting public information

• Prefer the more recent source, but don't discard the older one silently — note both.
• Prefer a primary source over a secondary one (a directory scraping other directories).
• If the conflict can't be resolved, the record should be labeled contradictory, not silently resolved in whichever direction is more convenient.

Contact-channel verification guidance

• A published contact channel is a materially stronger claim than a pattern-generated one.
• A pattern-generated email is not verified just because the format is common — it is, at best, a reasonable inference, and should be labeled as such.
• Neither this framework nor Be My Tech's own process claims that a verified contact will respond, or that a message to it will be delivered — only that the identity and role claim is supported by a checkable public source as of the date stated.

Scoring / classification

For this framework, a single contact record is scored on five yes/no items:

1. Is there a dated public source for this person's current role?
2. Is the source independently checkable by the buyer, not just asserted?
3. Is the source less than 12 months old, or has an older source been re-confirmed?
4. If more than one source exists, do they agree?
5. Is any listed direct contact detail published rather than pattern-guessed?

This framework classifies the result as follows:

• 5/5 (or 4/5 with #5 not-applicable): Source-confirmed.
• 3–4/5: Partially verified — usable with the specific gap noted.
• 0–2/5: Likely guessed or unsupported — should not be treated as verified.

Red-flag checklist

• No source cited at all, only a database name.
• Source is another directory that itself cites no original source (circular sourcing).
• Email format matches a common pattern with no independent confirmation it's real.
• Title or company is more than 12 months old with no re-check.
• The same record appears with different titles across sources and neither is dated more recently.
• Bulk records where every contact at a company shares an identical, formulaic source citation.

Buyer-side acceptance checklist

Before accepting a delivered batch of contacts (from any vendor, including Be My Tech), a buyer can reasonably ask:

• Can I see the source for each record, not just a confidence score?
• Is each source dated?
• Does the vendor distinguish published contact details from inferred ones?
• Is there a stated process for what happens when sources conflict?
• Can I get a replacement or explanation for a record that fails this framework?

Worked example (fictional — not a real vendor, client, or individual)

Claim: "Jordan Ellis, VP of Operations at Northfield Instruments (a fictional company), jordan.ellis@example.com" — example.com is the domain reserved for illustrative use by internet standards (RFC 2606) and does not resolve to a real mailbox.

• Step 1: Claim is role + direct email.
• Step 2: Source cited is a March 2026 press release naming Jordan Ellis as VP of Operations.
• Step 3: Press release is public and independently checkable. ✓
• Step 4: Dated 4 months before this review — within the 12-month standard. ✓
• Step 5: No second source found; no contradiction to resolve.
• Step 6 (email): No source confirms this specific email address — it follows a pattern inferred from the company's known format, not confirmed anywhere.

Classification: Role is source-confirmed (score 4/5, treating email verification as not-applicable-but-noted). Email is guessed, not verified, and should be labeled as such if delivered.

Limitations and appropriate use

• This framework judges what can be checked from public information. It cannot confirm a contact will read, respond to, or convert on an outreach message.
• It does not guarantee that public information is itself accurate — only that a claim traceable to a public, dated, independently-checkable source is stronger evidence than one that isn't.
• It is not a legal, privacy-compliance, or data-protection audit. Handling of any contact data should separately follow applicable regulations for the buyer's jurisdiction and use case.
• Scores are a classification aid, not a guarantee. A 5/5 record can still be wrong if the underlying public source itself was inaccurate or has since changed.

This framework focuses on a single contact record. For judging an entire delivered list at once, see What Makes a Prospect List Trustworthy?

Methodology

Version 1.0. Be My Tech's own internal research process is built on this same framework; every delivered prospect record is reviewed against it before delivery.

Get 3 free source-cited samples

If you want to see this framework applied to real prospect records instead of doing the review yourself, Be My Tech provides source-cited, date-checked, Founder-reviewed prospect research — start with 3 free samples.