Custom lead qualification checks companies against your actual buying criteria. It can identify whether a business offers the right services, covers the right territory, or matches a particular operating model. It should not pretend website research can prove budget, purchase intent, or a decision maker's willingness to buy.
Start with an explicit rubric. A generic “good lead” score is difficult to inspect and easy to misunderstand.
Separate requirements from preferences
Write down the conditions a company must satisfy, the conditions that exclude it, and the signals that merely make it more interesting. Do not let several weak preferences override a failed mandatory requirement.
For an illustrative commercial equipment supplier, a rubric could require evidence of commercial maintenance services and coverage of the target county. It could exclude equipment-only retailers. Several branches might increase priority but would not be mandatory.
Treat firmographic assumptions carefully. The number of visible staff profiles is not a verified headcount, and a polished website is not proof of revenue. Name those fields as observable signals if you use them.
Give each check an evidence rule
Specify which pages or source types can establish a criterion. A dedicated services page may be stronger evidence than an old blog post mentioning the subject once. Distinguish the business's own claims from third-party directory classifications.
Use yes, no, and unknown for individual checks. “No” should mean the agreed exclusion or contrary evidence was found. “Unknown” means the available material did not establish the answer. Keeping these separate prevents silent rejection of businesses with sparse websites.
Retain supporting URLs and short research notes. The reviewer should be able to understand a decision without rerunning the entire investigation.
Test the rubric against real disagreements
Have your team label a small representative set manually. Include borderline cases, companies outside the target, and businesses with several service lines. Compare automated research with those labels.
When people disagree, ask whether the rule is ambiguous or the evidence is weak. Improve the rubric and retest. A useful acceptance review examines false inclusions and missed suitable companies, not just the total number marked qualified.
Use scores only after the underlying checks work
A score can help prioritize an already eligible list. Explain what contributes to it and keep the component checks available. A score without reasons makes it hard for sales staff to trust or correct the results.
Do not present a qualification score as a measured probability of a sale unless it has actually been calibrated against suitable outcome data. Start with transparent categories such as review first, eligible, and needs research.
Keep feedback and refresh dates
Let sales staff flag incorrect findings and record why they disagree. Version the rubric so you know which rules produced each result. Recheck time-sensitive criteria before relying on older research.
Our business research and lead data service can turn your criteria into a scoped research process. Share examples of suitable and unsuitable companies so the first sample can test the distinction that matters to your team.
