Resources

Lead Quality Before Volume

By Michael Santiago

Define lead quality through observable fit, intent, timing, reachability, and contact context, then test the rubric against real outcomes.

“We need better leads” sounds specific until two people try to define it. Marketing may mean the right company completed a high-intent form. Sales may mean a reachable decision maker has an active project and agreed to a meeting. Finance may mean the eventual opportunity can support the cost of acquisition. All three views can be reasonable, but they describe different points in the process.

A useful definition of lead quality does not force those points into one adjective. It breaks quality into observable dimensions, records what is known, and shows which questions remain open. That gives teams a shared way to decide what should happen next without treating every unknown as a failure.

The five-part rubric below works as a starting point for many B2B programs: fit, intent, timing, reachability, and contact context. It is a decision aid, not a universal model. Each company should rewrite the criteria around its actual offer, market, sales motion, and legal obligations.

1. Fit: can the company genuinely serve this account?

Fit describes the relationship between the prospect and the offer. It may include company type, location, employee range, technical environment, required capability, or minimum project scope. Good fit criteria are tied to real operating boundaries. “Enterprise” is too loose if no one agrees whether it means revenue, headcount, complexity, or contract size.

Start with hard exclusions. If the business is licensed in three states, an inquiry from a fourth state may be out of scope regardless of enthusiasm. If the product requires a particular platform, compatibility is a gating question. Then identify softer indicators that help prioritize but should not silently disqualify a record.

Record the evidence for each fit judgment. A value selected by the prospect, a company website, a verified account record, and a vendor inference are different sources. If they conflict, show the conflict rather than allowing the latest sync to overwrite it invisibly.

2. Intent: what action did the person actually take?

Intent should begin with the behavior you can describe without interpretation. A request for a pricing conversation is different from a webinar registration. A return visit to implementation documentation may be useful context, but it does not carry the same meaning as a stated project need.

Write intent levels as actions: requested consultation, asked a product question, downloaded a planning guide, attended an event, or viewed a page. Avoid labels such as “hot” unless the team can trace them back to defined behavior. That makes it possible to change the priority of an action later without rewriting history.

Do not make every activity cumulative by default. Ten low-commitment page views do not necessarily equal one explicit request. If a behavior is used to infer intent, label it as an inference and evaluate it separately from direct statements.

3. Timing: when could a decision reasonably happen?

Timing is often unknown at capture. That is not the same as “no project.” Use states such as active deadline, planned period, exploratory, deferred, and unknown. A prospect who asks for implementation next month should be handled differently from one researching next year's budget, but both may deserve a useful response.

Ask for timing only when the answer changes the next step. A required dropdown that offers arbitrary windows can create false precision. In some sales motions, the best timing signal arrives during the first conversation and belongs in the disposition rather than on the form.

Timing also decays. A six-month-old “this quarter” value cannot remain current indefinitely. Store the date and source with the answer, then define when a review is required.

4. Reachability: is there a working path to the person?

Reachability is narrower than willingness. A syntactically valid email may still bounce. A delivered email may never be read. A telephone number may be correctly formatted but unsuitable for the intended use. Keep technical validation, delivery, reply, and permission as separate facts.

A practical record could distinguish email format checked, domain accepts mail, message delivered, human reply received, telephone type known, and preferred channel stated. None of these should be stretched beyond what was actually verified.

The program also needs suppression handling. In the United States, the FTC's CAN-SPAM compliance guide explains that commercial email rules apply to business-to-business email too, including sender information and opt-out duties. Other channels and jurisdictions carry different requirements. The operational lesson is simple: contact status and objections need to be available at the moment of outreach, not stored in a disconnected system no one checks.

5. Contact context: why is this outreach appropriate?

Contact context answers questions that a score can hide. Did the person ask this company for a reply? Did a partner make an authorized introduction? Was the information collected for another purpose? Which organization is identified as the sender? Which channels are covered?

This dimension is partly about customer experience and partly about legal and privacy obligations. For organizations subject to the GDPR, the European Commission describes principles including transparency, purpose limitation, data minimization, accuracy, storage limitation, and accountability. Its overview of those principles is a useful design reference. Counsel should review the actual program because duties depend on facts such as jurisdiction, legal basis, channel, and data source.

A lead should not receive a stronger quality label merely because more personal data was appended. Collecting and enriching information creates responsibility. Include only data that serves a defined decision and can be maintained appropriately.

Use gates before weights

Teams often begin by assigning points to every field. A clearer first version uses gates. A lead cannot move into a particular sales queue unless required conditions are met. For example:

• Serviceable fit: yes
• Relevant request: yes
• Contact route: available
• Contact context: documented
• Timing: any state, including unknown

Once a record clears those gates, the team can use softer factors to order work. An explicit implementation deadline may move a lead ahead of an exploratory download. A strategic account may route to a named owner. The gate protects the definition; the priority supports capacity decisions.

If the team later builds a model, evaluate it with metrics that reflect the tradeoff. Google's explanation of precision and recall is helpful here. Precision asks what portion of predicted positives were actually positive, while recall asks what portion of all actual positives the model found. Raising a threshold can reduce false positives while increasing false negatives. There is no responsible threshold choice without deciding which mistakes matter in the real workflow.

A worked rubric

Consider a hypothetical company selling maintenance software to multi-site manufacturers. A plant operations manager requests a demo. The company operates six facilities in the supported countries, uses a compatible system, and states that it is reviewing tools for the next budget cycle. The business email passes delivery checks. The person asked for a demo from the seller named on the form.

The rubric could read:

Now change one detail. Suppose the same record came from a conference attendee list, and the organizer's notice did not say the details would be sent to this software company. Fit and possible interest might remain, but contact context is unresolved. The record should not inherit the same outreach path. Operations needs to review the source terms and applicable rules before use.

The rubric does not declare one person “good” and another “bad.” It identifies what the company knows and which action the evidence supports.

Test the rubric against outcomes

Drafting criteria in a meeting is only the first pass. Pull a small, recent cohort with enough time to reach meaningful dispositions. Hide the outcome, have marketing and sales apply the rubric independently, then compare:

1. Where did reviewers disagree about the same evidence?
2. Which fields were unavailable or stale?
3. Which accepted leads never received accountable follow-up?
4. Which rejected leads later became valid opportunities?
5. Which criteria correlate only because one source or segment dominates the sample?

The goal is not to prove the first rubric correct. It is to find definitions that fail under real examples. Keep version history so historical reports remain interpretable after a rule changes.

Put the rubric into the weekly operating rhythm

Assign an owner to each dimension and to the rubric as a whole. Marketing might own source and capture behavior. Revenue operations may own routing and data definitions. Sales managers may own disposition quality. Legal or privacy specialists should advise on contact context and data handling. Ownership should be explicit even when one person fills several roles.

Review a small set of edge cases each week. Choose records that were rerouted, rejected for unclear reasons, reopened, or corrected after contact. Those cases reveal more than a broad dashboard. Update one definition at a time, communicate the change, and monitor whether it solves the original disagreement.

Volume still matters for staffing, budgets, and forecasting. It simply belongs after the team can explain what entered the system and what happened next. A quality rubric gives that volume a denominator people can trust.

Build a clearer lead operation on Leadz.com

The Leadz.com domain and the starter-brand assets presented here are available, subject to final transaction terms. Inquire about acquiring Leadz.com and tell us what you plan to build.