How to Prioritize B2B Prospects Before Outreach
When our prospect list is larger than our sending, research or reply-handling capacity, “contact everyone” is not a strategy. We need a transparent way to decide which accounts and people deserve attention first. That decision should be reproducible, editable and based on evidence we actually possess—not a black-box label that implies purchase intent.
In this guide, we will build a 100-point prioritization model across four dimensions: fit, timing, reachability and evidence confidence. We will add negative scores and hard exclusions, score a 15-prospect sample, and convert the result into Tier 1, Tier 2 and Tier 3 operating plans inside our prospecting and campaign workflow.
Prioritization is not inbound lead scoring
Inbound lead scoring often combines behaviors such as form submissions, website activity or product usage. A cold-outreach prospect may have performed none of those actions. We should not borrow an inbound “engagement” score and pretend we know the same thing.
Our outbound prioritization answers:
Given the accounts and contacts we have already selected, which records should receive our limited research and outreach capacity first?
The score organizes work. It does not prove:
- that the account is currently buying;
- that the contact has budget;
- that the person will respond;
- that an email will reach the inbox;
- that a meeting will convert.
We keep every rule visible so sales and operations can challenge it.
Prepare the source data
Prioritization starts after account and contact identity are stable. If we still need to define the target market, use the Ideal Customer Profile Builder. If we need to create the list, use the B2B prospect list guide. If the accounts are known but contacts are missing, use the company-list-to-contacts workflow.
The Getlead native B2B export contains:
First Name
Last Name
Job Title
Email
Company
Website
LinkedIn URL
Industry
Employee Count
Verification Status
We add scoring fields separately:
Fit Score
Timing Score
Reachability Score
Evidence Score
Negative Score
Priority Score
Priority Tier
Do Not Contact Reason
Evidence URL
Evidence Date
Score Version
Owner
Next Action
These are workflow fields, not native export columns.
The 100-point model
Positive score:
Fit: 40
Timing: 20
Reachability: 20
Evidence confidence: 20
Total positive: 100
Final score:
Priority Score =
Fit + Timing + Reachability + Evidence - Negative Score
Hard exclusions override the number. A suppressed contact cannot become eligible because the account scores 95.
Dimension 1: Fit — 40 points
Fit measures alignment with our documented ICP and buying-role hypothesis.
Account fit — 25
| Rule | Points | |---|---:| | Industry matches must-have ICP | 8 | | Geography matches operating scope | 5 | | Employee-count band matches approved ICP | 5 | | Business model/use case matches | 4 | | Account exclusions are clear | 3 |
Employee Count is used from the export for post-search qualification. We do not present it as an active Getlead search filter unless the product later supports that behavior.
Contact fit — 15
| Rule | Points | |---|---:| | Function matches the problem owner | 5 | | Seniority matches buying-role hypothesis | 4 | | Job title is a direct/accepted variant | 4 | | Buying role is documented | 2 |
The decision-maker workflow helps us distinguish economic buyer, champion, user and evaluator. A CEO does not automatically receive full points.
Fit example
Our approved example is sales leaders at US B2B SaaS companies. A US software company may receive industry and geography points. A VP Sales may receive function, seniority and title points. A Customer Support Manager at the same company does not become a fit merely because the account matches.
Dimension 2: Timing — 20 points
Timing records current, attributable context. It must not be called purchase intent unless we have data that genuinely supports that claim.
| Evidence | Points | |---|---:| | Recent, directly relevant operational trigger | 10 | | Trigger is moderately related or older | 5 | | No documented trigger | 0 | | Trigger affects the target function | 5 | | Evidence date is within team policy | 3 | | Evidence source is primary/reliable | 2 |
Examples of possible triggers:
- public SDR hiring;
- sales leadership appointment;
- documented new-market expansion;
- product launch affecting outbound demand;
- public operations or tooling project.
We store URL, date and a factual note. “Growing fast” without a source is not a trigger. Funding may provide context but does not prove that a sales leader wants our solution.
Timing points expire according to a documented policy. A job post from last year should not retain a “recent hiring” score indefinitely.
Dimension 3: Reachability — 20 points
Reachability combines the approved verification status and contactability context:
| Rule | Points | |---|---:| | Approved Verified Emails Only condition | 12 | | Current role and company identity confirmed | 4 | | LinkedIn URL supports identity review | 2 | | Verification timestamp is within policy | 2 |
In Getlead, Verified Emails Only retains contacts that passed the real-time SMTP check at verification time and excludes catch-all and risky addresses. This reduces hard-bounce risk. It does not guarantee future delivery.
Catch-all, risky, invalid and unknown records do not receive the 12 verified points under this model. An invalid or suppressed record can also be a hard exclusion.
Reachability does not measure permission, relevance or likely response. It is one operating dimension.
Dimension 4: Evidence confidence — 20 points
Evidence confidence asks whether another reviewer can reproduce our score.
| Rule | Points | |---|---:| | Company and website match confirmed | 5 | | Title/function evidence is current | 4 | | Data source and date recorded | 4 | | Trigger has attributable evidence | 4 | | No unresolved field contradiction | 3 |
A score based on a company page, current profile and dated source is more trustworthy than one built from an unlabeled spreadsheet. Evidence points do not make the account “better”; they make our decision more auditable.
Contradictions include:
- contact listed at two companies;
- Website does not match Company;
- title and function disagree;
- trigger applies to a similarly named company;
- verification timestamp is missing while status claims recency.
Unresolved contradictions move the record to review.
Negative scores and hard exclusions
Negative rules prevent positive evidence from hiding risk:
| Condition | Adjustment | |---|---:| | Title is an indirect/weak variant | -5 | | Trigger is stale | -5 | | Company/domain relationship uncertain | -10 | | Duplicate active campaign | -20 | | Unresolved role mismatch | -15 |
Hard exclusions:
- suppression or opt-out;
- invalid email under campaign policy;
- explicit out-of-ICP account;
- prohibited geography or policy exclusion;
- unresolved identity conflict;
- active overlapping campaign where duplication is not allowed.
Hard exclusions produce:
Priority Tier = DO_NOT_CONTACT_YET
Do Not Contact Reason = controlled reason code
We preserve the record and reason rather than deleting it.
Tier definitions
Default tiers:
| Tier | Score | Operating treatment | |---|---:|---| | Tier 1 | 80–100 | Account research and tailored evidence | | Tier 2 | 60–79 | Segment-level campaign with targeted QA | | Tier 3 | 40–59 | Hold, enrich or lower-cost research | | Review | 0–39 | Resolve missing/contradictory data | | Do Not Contact Yet | Hard exclusion | No outreach until reason is resolved, if resolvable |
Thresholds are starting policy. Teams can change them, but each score version must record the weights and thresholds used.
Tier 1
Tier 1 receives:
- current account research;
- buying-group review;
- factual trigger validation;
- human-reviewed personalization;
- a tailored proof asset;
- named owner;
- explicit next action.
High score does not justify fabricated intimacy or unsupported claims.
Tier 2
Tier 2 uses:
- a strong segment contract;
- verified variables;
- relevant proof;
- controlled campaign pilot;
- sampled human QA.
Tier 3
Tier 3 is not “send anyway.” We may:
- enrich missing title/evidence;
- wait for a relevant trigger;
- use a lower-cost research queue;
- exclude if the potential value does not justify more work.
Score 15 sample prospects
The table below uses synthetic companies and people. Scores demonstrate the method, not market performance.
| # | Synthetic prospect | Fit | Timing | Reach | Evidence | Negative | Final | Tier/reason | |---:|---|---:|---:|---:|---:|---:|---:|---| | 1 | VP Sales, Northstar Cloud | 39 | 17 | 20 | 19 | 0 | 95 | Tier 1 | | 2 | RevOps Director, Signal Harbor | 38 | 15 | 20 | 18 | 0 | 91 | Tier 1 | | 3 | Head of Sales, MetricForge | 37 | 10 | 20 | 18 | 0 | 85 | Tier 1 | | 4 | SDR Director, OrbitLedger | 35 | 8 | 20 | 17 | 0 | 80 | Tier 1 | | 5 | VP Revenue, PineStack | 38 | 5 | 18 | 17 | 0 | 78 | Tier 2 | | 6 | Sales Ops Lead, AtlasDesk | 33 | 9 | 18 | 16 | 0 | 76 | Tier 2 | | 7 | CRO, VectorPilot | 36 | 0 | 20 | 18 | 0 | 74 | Tier 2 | | 8 | Head of Growth, RelayWorks | 29 | 8 | 18 | 16 | 0 | 71 | Tier 2 | | 9 | Sales Director, NimbusLayer | 32 | 0 | 16 | 17 | 0 | 65 | Tier 2 | | 10 | SDR Manager, BrightOps | 27 | 5 | 18 | 14 | 0 | 64 | Tier 2 | | 11 | Revenue Analyst, CoveGrid | 22 | 0 | 18 | 14 | 0 | 54 | Tier 3 | | 12 | VP Sales, Northstar Systems | 35 | 10 | 12 | 8 | 10 | 55 | Tier 3: domain uncertainty | | 13 | Customer Support Manager, SwiftBase | 12 | 5 | 20 | 16 | 15 | 38 | Review: role mismatch | | 14 | Former VP Sales, EmberFlow | 28 | 0 | 0 | 7 | 15 | 20 | Review: stale role | | 15 | Suppressed CRO, LatticeFox | 40 | 20 | 20 | 20 | 0 | — | Do Not Contact Yet |
Prospect 15 illustrates the override. A perfect positive score cannot cancel suppression.
Build the score in a spreadsheet or CRM
Each rule should use controlled values, not a hidden formula maintained by one person.
Example:
fit_industry = 0 or 8
fit_geography = 0 or 5
fit_company_size = 0 or 5
fit_business_model = 0 or 4
fit_exclusions_clear = 0 or 3
fit_function = 0 or 5
fit_seniority = 0 or 4
fit_title = 0 or 4
fit_buying_role = 0 or 2
Then:
fit_score = sum(fit fields)
The same applies to timing, reachability and evidence. The model stores component fields so a reviewer can see why a total changed.
We include:
Score Version = PRIORITY-V1
Scored At
Scored By
When weights change, existing scores retain their version or are deliberately recalculated.
Assign accounts before people
If we score only people, five contacts at the same excellent account can crowd out account diversity. We use two layers:
- account score;
- contact/buying-role score.
We first prioritize accounts, then choose the relevant people inside each account. JTBD-02's buying-group coverage model helps us avoid selecting only the CEO.
A simple account allocation rule:
Tier 1 account: up to 2–4 relevant buying roles
Tier 2 account: 1–2 primary roles
Tier 3 account: hold/enrich
These are operating caps, not product limits or universal benchmarks.
Convert priority into campaign operations
The workflow is:
ICP
→ Account/contact list
→ Clean and verify
→ Score
→ Apply exclusions
→ Assign tier
→ Allocate research
→ Segment
→ Capacity check
→ Pilot launch
Tier becomes a campaign input, not a replacement for segment. Two Tier 1 contacts may belong to different pain hypotheses and should not receive the same message automatically.
The segmentation guide defines pain, proof and CTA. The CSV launch guide handles mapping, rendering and pilot.
Re-score at meaningful events
We do not recalculate continuously without reason. Re-score when:
- contact changes role/company;
- account enters/leaves the ICP;
- new attributable trigger appears;
- evidence expires;
- verification policy requires refresh;
- suppression status changes;
- segment or offer materially changes.
Score decay should be explicit. For example, timing points can expire after a team-defined window. Evidence confidence can fall when the source becomes stale. We do not silently change historical scores.
Measure whether prioritization helps
We compare tiers on:
- research time;
- eligible records;
- data corrections;
- attempted sends;
- qualified replies;
- meetings/next steps;
- opt-outs and complaints;
- bounce classification;
- disqualification reasons;
- sales follow-up time.
We do not claim the score “caused revenue” from a small sample. We look for whether higher tiers contain more relevant, actionable records and whether the operating cost is justified.
Calibration review:
- Which high-score records were disqualified, and which rule failed?
- Which low-score records produced qualified outcomes, and what evidence was missing?
- Are weights duplicating the same signal?
- Are sales teams overriding tiers? Why?
- Do exclusions protect the workflow?
We change one model version at a time and document the reason.
Governance and fairness
Prioritization should use business-relevant account, role and evidence criteria. We do not use protected personal characteristics or unsupported personal inferences. Geography is used only where it reflects legitimate operating scope, language, service or legal constraints—not as a proxy for personal traits.
Access to scoring data follows least privilege. Evidence notes remain factual and professional. The system must allow correction when identity or role data is wrong.
Common prioritization mistakes
Calling a fit score “intent”
Fit indicates alignment with our ICP. It does not show current buying behavior.
Hiding weights in an AI label
Operators cannot audit or improve the decision. Keep rules visible.
Scoring before cleaning
Duplicates and stale roles distort capacity allocation.
Letting points override suppression
Hard exclusions remain absolute.
Scoring people without account allocation
One account can consume the entire outreach queue.
Using verification as the whole score
Reachability does not prove relevance.
Never expiring timing evidence
Old triggers create false urgency.
Sending Tier 3 by default
Tier 3 is a work queue, not leftover campaign volume.
Prioritization QA checklist
- [ ] ICP and exclusions are documented.
- [ ] Native export and workflow fields are separate.
- [ ] Fit weights total 40.
- [ ] Timing weights total 20.
- [ ] Reachability weights total 20.
- [ ] Evidence weights total 20.
- [ ] Negative scores are visible.
- [ ] Hard exclusions override total score.
- [ ] Trigger source and date are stored.
- [ ] Verification language is time-bound.
- [ ] Score Version and reviewer are recorded.
- [ ] Account allocation precedes contact volume.
- [ ] Tier treatment changes research effort.
- [ ] Campaign segment remains separate from tier.
- [ ] Re-score events and decay are documented.
- [ ] Analytics are reviewed by tier and segment.
Continue through the Getlead hub
If fit rules are weak, return to the ICP Builder. Build new contacts with the B2B prospect list guide, decision-maker workflow or company-list enrichment guide. Clean data with JTBD-04. Then use JTBD-05 segmentation, FT-04 readiness and JTBD-07 launch.
Frequently asked questions
What is a good B2B prospect score?
A score is meaningful only inside the documented model and version. In this starting policy, 80–100 is Tier 1, but teams should calibrate against their own operational evidence.
Is prospect scoring the same as purchase intent?
No. The model ranks fit, timing context, reachability and evidence. It does not prove an active buying process.
Can a suppressed prospect have a high score?
The positive dimensions may be high, but suppression creates a hard exclusion and no outreach.
Should we prioritize accounts or contacts?
Prioritize accounts first, then choose relevant buying roles. This prevents several people at one account from consuming the queue.
How often should scores change?
At meaningful events such as role changes, new evidence, evidence expiry, verification refresh or an ICP change—not continuously without a documented reason.
Can Getlead's Employee Count be used?
It is part of the confirmed export and can support post-export qualification and scoring. We do not describe it as an active search filter in the current interface.
Does a verified email receive the maximum priority?
It can receive reachability points, but fit, timing, evidence, suppression and role relevance remain separate.
What if two prospects have the same score?
We use a documented tie-breaker rather than an arbitrary sort. First prefer the account with stronger evidence confidence, then the contact who fills the missing buying role, then the older waiting record. If research or reply capacity is constrained, an owner can make a manual decision, but the override reason should be recorded. The score organizes judgment; it does not remove accountable human review.
Should scoring weights be different for every campaign?
The core model should remain stable long enough to evaluate. A materially different offer or ICP may justify a new version, but we should not alter weights merely to push favored accounts into Tier 1. Version the model, explain the business reason and preserve historical scores for comparison.