B2B Leads QA: Fit, Role, Freshness, Reachability, and Risk
A B2B lead is a record of a person or organization that may be relevant to an offering and may be eligible for a defined follow-up. Lead QA verifies separate questions: account fit, the person’s role, evidence date, contact-point reachability, identity and duplicate risk, and whether the intended use is permitted. None of those fields proves buying intent.
There is no universal lead-quality score. A weighted total can hide a hard failure: the wrong person, a duplicate identity, an invalid address, an objection, or a use the team is not permitted to make. Verify hard stops first. Prioritize only the records that survive.
A lead is not an account, contact, opportunity, marketing-qualified lead, or sales-qualified lead. Microsoft Dynamics’ qualification documentation illustrates the distinction by allowing qualification to create or match separate account, contact, and opportunity records. Product behavior varies, but the data-model lesson travels: person, organization, buying process, and acquisition record are different entities.
The six-part QA contract
| Dimension | Question | Minimum evidence | Hard-stop examples |
|---|---|---|---|
| Fit | Does the organization match the declared market hypothesis? | Source, observed attribute, access date, inclusion and exclusion rule | Explicit exclusion, incompatible geography or use case |
| Role | What is this person’s relationship to the problem or decision? | Current title or first-party statement plus source and date | Wrong identity, no plausible relationship, role asserted only from stereotype |
| Freshness | When was each volatile field last verified? | Field-level source and verification timestamp | Required volatile field has no evidence date |
| Reachability | Does the contact point exist and work for the intended channel? | Validation result, delivery evidence, or direct confirmation | Invalid address, persistent delivery failure, wrong number |
| Identity | Is the record the intended person and organization, and is it already represented? | Stable identifiers and duplicate review | Conflicting identities, unresolved duplicate, shared inbox treated as one person |
| Risk and permission | May the organization use this contact point for this purpose and channel? | Source, notice, applicable basis, preference and suppression state | Opt-out, objection, suppression, prohibited source or use |
Fit should be rule-based rather than intuitive. Record the attribute, its source, and why it matters to the offering. “Looks enterprise” is not evidence. A verified company size can support a declared segmentation rule, but it still does not reveal the person’s authority or current project.
Role needs the same discipline. A title can suggest a function, not prove decision power. Distinguish user, evaluator, technical reviewer, budget holder, executive sponsor, procurement, and unknown only when the evidence supports that relationship. Unknown is a valid state; invented certainty is not.
Freshness belongs to fields, not files
A record does not become fresh because one enrichment job ran today. Company domain, employment, title, phone, email, consent, account status, and fit evidence change at different rates. Store the source and verified-at timestamp with each volatile field or evidence bundle.
Do not invent one cross-industry expiry window. A team should define review triggers based on volatility, consequence, and use. A role used for one manual research call carries a different risk from a role used to launch a large automated sequence. When a required field is beyond its approved window, route it to verification instead of silently treating stale as false or current.
“Unknown” and “invalid” are different states. Unknown means the team lacks enough evidence. Invalid means the available evidence contradicts the value or the contact point failed a declared test.
Reachability is not permission
An address can accept mail and still be inappropriate for a particular message. Permission and lawful use depend on jurisdiction, recipient type, source, purpose, channel, notice, preferences, and other facts. The UK ICO’s B2B marketing guidance shows how even one jurisdiction distinguishes corporate subscribers, sole traders, electronic mail, calls, consent, and legitimate interests.
That guidance is not a global permission slip. It supports a more general operating boundary: store the intended purpose and channel, retain the data source, provide required privacy information, evaluate the applicable basis, and honor objection or opt-out. Obtain qualified legal review for the actual regions and campaign.
Suppression must be durable. Microsoft documents contact-point consent controls that can share preference updates across integrated applications. The product implementation is optional; the system requirement is not. A downstream tool must not reactivate a suppressed address because its local copy was stale.
Resolve identity before qualification
Duplicate rules create candidates, not truth. Dynamics documents signals including the same email, same phone, similar lead and company names, and similar name plus email domain. Those signals can catch obvious duplicates, but shared mailboxes, recycled numbers, name collisions, subsidiaries, consultants, and job changes still need review.
Before merging, compare source provenance, timestamps, activity history, organization relationships, consent and suppression, ownership, and open processes. Preserve the more restrictive communication state when policy requires it. Never let an automatic merge erase the evidence needed to understand why a record was suppressed or qualified.
Run QA in dependency order
Declare the intended use
Record the campaign or sales purpose, channel, population, required fields, jurisdictional review, and approved hard stops before evaluating records.
Preserve source provenance
Store where each field came from, when it was obtained or verified, and whether it was observed, declared, inferred, or unresolved under the organization’s data contract.
Resolve entity and duplicate risk
Match the person to the correct organization, search for existing records, and route conflicts to human review before creating downstream activity.
Evaluate fit and role separately
Apply explicit account inclusion and exclusion rules, then assess the person’s current relationship to the problem. Do not let engagement overwrite poor fit.
Check contact point and permitted use
Validate reachability, applicable notice or basis, preference, objection, suppression, and channel restrictions. Stop the record when a hard condition fails.
Release with a receipt
Pass only approved fields and purposes downstream. Record the recipient system, action, owner, version, and result so corrections and suppressions can propagate.
Prioritization comes after validity
Once a record passes the hard gates, a team can prioritize based on declared fit, role, first-party behavior, timing, or sales capacity. Keep the component evidence visible. A score should accelerate review, not erase the reasons, unknowns, or disqualifying conditions underneath it.
Sources
- Microsoft Learn, “Customize lead qualification experience in Dynamics 365 Sales”
- Microsoft Learn, “Enable the detection of duplicate leads”
- Information Commissioner's Office, “Business-to-business marketing”
- Information Commissioner's Office, “Direct marketing guidance”
- Microsoft Learn, “Stay compliant with privacy regulations”
Continue the evidence path
Related reading
Read first
What Is CRM for a Lean SaaS Team? A Shared Data Contract, Not a Contact Database
Define people, companies, opportunities, and activity records before importing leads.
Next step
What Is a Sales Pipeline? Stages, Deal Evidence, and Qualification
Move verified evidence into opportunity stages without treating a lead as pipeline.
Related
What Is a Sales Funnel? Stages, Conversion, and CRM Handoffs
Keep aggregate stage conversion separate from the quality of individual records.