Account research guide

How to research HVAC contractor accounts

Start with a defined company-location universe, then add only the public business facts needed for qualification. Preserve sources, dates, uncertainty, and exclusion reasons so another reviewer can reproduce the decision.

Prepared August 28, 2026 · 8 minute read

Account research should narrow a business question—not turn a directory row into an assumed prospect. A company name, location, or website can support further research, but it does not establish current operation, commercial fit, buying intent, consent, deliverability, or permission to contact.

1Define the target account and location unit

Write the operational HVAC inclusion rule before collecting rows. Decide whether one account means a legal entity, brand, headquarters, branch, or service location, and whether multi-trade companies, franchises, distributors, or adjacent trades belong in scope.

Specify geography at the level the decision needs: state, county, metro, ZIP list, or another explicit boundary. Keep company-level and location-level conclusions separate so a valid brand does not automatically validate every location.

2Choose qualification facts before researching

List the minimum public business facts needed to answer the use case. These might include official website, stated services, operating geography, location evidence, business category, or a buyer-defined characteristic visible on a first-party source.

Define each field, acceptable evidence, blank handling, and review date. Exclude sensitive personal data and avoid collecting personal contacts by default. Do not infer company size, revenue, technology use, or buying readiness without an identified, supportable source.

3Build a candidate universe without calling it qualified

Use business name, normalized industry, and geography to form a candidate set. Treat website, phone, and address as comparison or research-starting fields, not proof that a business is active, relevant, reachable, or appropriate for outreach.

Record the exact segment rule and extraction date. Preserve source values before normalization, assign a stable row identifier, and mark duplicates or parent-location relationships for review rather than silently collapsing them.

4Prefer first-party evidence for company claims

When practical, verify material facts against the company’s own current website or another authoritative public source. Save the page URL, the fact it supports, and the date reviewed. A search-result snippet alone may be stale or truncated.

If sources disagree, keep the conflict visible. Use explicit states such as supported, not supported, conflicting, and unresolved. Absence from a page is not proof that a service, location, or characteristic does not exist.

5Separate observation from interpretation

Store the observed fact and source separately from any buyer-specific classification. For example, a site may state that a company serves three named counties; whether that passes a territory rule is a separate decision based on the written rule.

Do not transform directory firmographics into claims about demand, capacity, account value, purchase timing, decision-makers, or intent. Those require buyer-supplied criteria or distinct evidence and should retain their own provenance.

6Create an exception and review queue

Queue ambiguous HVAC scope, inaccessible sites, conflicting locations, shared domains, franchise relationships, and suspected duplicates. Record the reason, evidence reviewed, reviewer decision, and review date.

Keep unresolved candidates out of “qualified” totals unless the buyer’s written rule explicitly says otherwise. Report every denominator and exclusion so a completed count cannot hide skipped or unverifiable rows.

7Test research quality on a reproducible sample

Select a reproducible sample from the exact candidate universe. Have an authorized reviewer apply the same definitions and evidence rules, then compare field-level and classification outcomes. Report supported, incorrect, conflicting, unresolved, and blank results separately.

Predefine acceptance thresholds and remedies in writing. A result for one dated segment does not establish universal catalog accuracy or future availability. The blank HVAC pilot scorecard can structure the test.

8Control delivery and downstream use

Deliver only the agreed business fields and evidence metadata through an appropriate secure method. Define access, retention, correction, and deletion responsibilities, and retain enough lineage to revise or remove a disputed finding.

Qualification for account research is not outreach authorization. Any later contact activity needs its own lawful-use, suppression, channel, identity, and approval checks.

Define the evidence before the research begins

Request a brief for the exact HVAC scope.

Submit the geography, inclusion rule, required business fields, approximate volume, evaluation method, intended use, timing, and budget status. Availability and field presence must be computed for the exact request.

Request an HVAC coverage brief