The lead list is sorted by score when a sales rep asks, “What is the difference between 72 and 68?” The team can explain the points assigned to completion rate, time spent, and CTA clicks. It may still be unclear who should contact whom, and why.
Document-based lead scoring is not a contest to produce a more precise number. It is a policy that lets a team explain which observations permit which follow-up actions—and what to review when the judgment is wrong.
Define the action the score may change
Before writing rules, decide what the score is allowed to influence. Prioritizing outreach, preparing additional material, and asking an existing owner to verify context require different evidence.
One score should not replace every decision. A new contact who received a first document and an active opportunity with a known response deadline do not need the same policy.
Write a concrete subject and action. “High scores are hot leads” is less useful than “For owned opportunities, review unresolved questions with a near response deadline first.”
Observations, context, and responses have different weight
Document access, page views, revisits, and link clicks are recorded behavior. They can inform a follow-up question, but they do not establish budget approval, purchase intent, decision authority, or buying stage.
Separate inputs into three groups and give each a different policy role.
| Evidence | Example | Permitted role in the policy |
|---|---|---|
| Observation | document opened, page revisited | flag a topic for review |
| Business context | prior meeting, requested material, response deadline | determine urgency and owner |
| Customer response | reply, question, agreed schedule | record a next step the customer expressed |
An item with observations alone remains a human review task, not an automatic conclusion. Even when context or a response is available, the score should not speak for the customer.
Build a scoring rule sheet, not a universal scorecard
A fixed rule such as “80% read = 50 points” looks simple but is difficult to justify. A team can begin with a scoring rule sheet instead of a points table.
- Trigger: which observation and context must appear together?
- Permitted action: what is the furthest follow-up the rule allows?
- Owner: who checks the original record?
- Review cycle: when are the rule and its cases reviewed?
- Prohibited interpretation: what must the rule not conclude?
If a requested pricing document is reopened, the owner may receive a review task. The record should not be labeled “price concern”; ask which comparison or condition needs clarification. If someone opens a security appendix and sends a question, assign the appropriate respondent without recording “security approved.”
Points can be added later to help with sorting. The evidence, permitted action, and prohibited interpretation on the rule sheet still need to remain visible.
Do not treat missing and unverified as the same zero
No record can mean no action, but it can also mean the activity was not measured or happened in another channel. Check whether the document was sent, whether link identification was preserved, or whether an offline file was shared.
A missing click does not establish a lack of interest. The reader may have found the answer in the document, or the CTA may not fit the current situation.
The scoring policy needs separate states for negative evidence and unverified evidence. Keep data-poor cases in a verification queue instead of hiding them under a low score.
Revise rules using both false positives and omissions
If an outreach based on a high score has the wrong context, do more than reduce the points. Record which input was overweighted and what additional information the owner needed.
If a low-scored case advances, look for missing context. A customer-provided schedule or explicit request may not have been represented in the rule.
Review the scoring rule sheet after enough cases accumulate or when the sales workflow changes. Check whether permitted actions were consistent and owners could return to the original evidence—not whether the distribution looked tidy. Record the reason and effective date for each rule change.
FeatPaper records are inputs, not intent judgments
FeatPaper can surface document access, page views and time, revisits, and link or CTA clicks as observation inputs. This article does not assume that those records determine purchase intent or priority, or that a CRM or marketing automation system assigns scores automatically.
The team must define the purpose of the score, the permitted action, the owner, and the review rule while keeping the original observation separate from its interpretation.
Lead scoring quality shows up in explainability and revision. An owner should be able to say why an action was chosen and return to the same evidence to improve the scoring rule sheet when the judgment misses.
