Turn a Sales Navigator search into account research briefs

Start by creating an Airtable base or spreadsheet that defines what a qualified account looks like before anyone opens Sales Navigator. Add fields for company name, source search, ICP segment, company size range, region, industry, trigger signal, disqualifier, research status, account owner, and next action. Use Claude to turn your ICP notes into a short checklist, but review the checklist against 10 known good customers and 10 bad-fit accounts before using it. Mark disqualifiers clearly, such as wrong region, too small, agency, education-only, competitor, active customer, or no relevant buyer function. This gives the rest of the workflow a scoring standard instead of letting every public signal look like intent.

Build an account research checklist for {{product_or_service}}.

ICP summary:
{{icp_summary}}

Best-fit customers:
{{best_fit_customer_examples}}

Bad-fit or disqualified accounts:
{{bad_fit_examples}}

Output:
1. Positive fit signals
2. Disqualifiers
3. Trigger signals worth researching
4. Buyer roles to look for
5. Account scoring criteria
6. Fields to create in the research tracker
7. QA checks before outreach

Keep the checklist practical for a 25-account sales research sprint.

Create an account search first, then a lead search only after the account universe looks clean. Use filters such as geography, company headcount, industry, department headcount, seniority, function, job listings, buyer intent, recent leadership changes, and companies in CRM where those filters are available. Save the search with a name that includes audience, region, segment, and date so the team can rerun or edit it later. Manually review the first 50 results and tag at least 10 as good fit, questionable, or reject before exporting. Tighten filters until the questionable and reject examples have clear reasons rather than vague discomfort.

Use Wiza to export the approved Sales Navigator search or lead list when your plan, data policy, and local outreach rules allow it. For each exported record, capture company name, company domain, account LinkedIn URL, contact name, title, profile URL, location, email verification status, source search, export date, and enrichment status. If export is not available, manually stage 25 accounts and one to three likely buyer contacts per account in Airtable. Create a duplicate-check view that groups by company domain and contact profile URL so the same account does not appear multiple times with slightly different names. Before enrichment, reject rows with blank company names, missing current company, obviously wrong geography, or contact titles that do not match the buyer-role checklist.

Use Perplexity and public sources to add only context that could change outreach priority or message angle. Capture recent initiatives, funding, hiring clusters, product launches, expansion, partnerships, technology investments, executive changes, relevant job postings, or operational pressure. Add fields for recent signal, signal date, source URL, source type, business priority, evidence strength, and why it matters. Exclude trivia, old blog posts, generic company descriptions, and unsupported assumptions about internal strategy. QA the first 10 rows by opening every source link and confirming the signal is recent, relevant, and tied to the account rather than the broader industry.

Research {{company_name}} for account-based sales outreach.

Product or service sold:
{{product_or_service}}

ICP and buyer roles:
{{icp_and_buyer_roles}}

Return only source-backed findings from the last {{recency_window}} when possible.

For each useful signal, include:
1. Signal summary
2. Source URL
3. Source date
4. Why it may matter commercially
5. Likely buyer function affected
6. Evidence strength: strong, medium, or weak
7. Outreach implication

Do not include generic company descriptions or unsupported guesses.

Check HubSpot before giving any account to a rep. Add fields for CRM record link, lifecycle stage, account owner, open opportunity, last touch date, active sequence, closed-lost reason, customer status, do-not-contact status, and previous competitor mention. Use a status such as new, owned, active opportunity, customer, do not contact, or recycle so the list can be routed safely. Create a review view for records where the company exists in CRM but the owner or lifecycle stage is missing. Do not let any account move into the work-now queue until ownership and active-deal conflicts are resolved.

Use Sales Navigator, LinkedIn profiles, and the CRM record to identify the most likely buying roles for each account. Add fields for economic buyer, functional owner, technical evaluator, daily user, blocker, champion candidate, and recommended first contact. Do not invent a buying committee if the titles are not visible; mark the role as unknown and add a research task instead. Ask Claude to infer likely buyer involvement from titles and company context, then manually review the top accounts before assigning outreach. QA by checking that every recommended contact has a reason beyond seniority, such as owning the relevant department, initiative, tool, or pain area.

Map likely buying committee roles for these accounts.

Account context:
{{account_context}}

Available contacts and titles:
{{contact_list}}

Product or service sold:
{{product_or_service}}

Buyer-role definitions:
{{buyer_role_definitions}}

For each account, output:
1. Recommended first contact
2. Economic buyer if visible
3. Functional owner if visible
4. Technical evaluator if visible
5. Possible champion
6. Unknown roles to research
7. Reasoning based only on visible evidence
8. Confidence level

Do not invent contacts or responsibilities.

Feed Claude the cleaned account table, CRM context, source-backed research, and buyer-role map. For each account, create a short brief with fit score, trigger strength, relevant evidence, likely pain hypothesis, buyer roles, first-touch angle, risks, disqualifiers, and one thing not to say. Store the brief in Airtable or a linked Google Doc, not in a loose chat thread. Require Claude to separate evidence from assumptions so the rep does not repeat guesses as facts. Review 5 to 10 briefs and reject any that include unsupported claims, stale sources, vague pain statements, or generic outreach angles.

Create concise account research briefs from this cleaned research table.

Research table:
{{account_research_table}}

CRM context:
{{crm_context}}

Buyer-role map:
{{buyer_role_map}}

Scoring criteria:
{{scoring_criteria}}

For each account, include:
1. Fit score from 1 to 5
2. Trigger strength from 1 to 5
3. Evidence-backed signals with source references
4. Likely pain hypothesis
5. Recommended buyer role
6. First-touch angle
7. Risks or disqualifiers
8. One thing the rep should avoid saying
9. Unknowns to verify

Label assumptions clearly. Do not invent facts.

Score each account using fit, trigger strength, buyer accessibility, CRM status, evidence recency, and likely urgency. Create three Airtable views: work now, nurture, and skip. In the work-now view, require a named owner, recommended contact, next action, due date, and outreach channel before the record is considered ready. In the nurture view, include the reason for waiting and the signal that would make the account worth revisiting. In the skip view, preserve the rejection reason so the next search can avoid similar false positives.

Prioritize these accounts into Work Now, Nurture, and Skip.

Account briefs:
{{account_briefs}}

Scoring criteria:
{{scoring_criteria}}

Current sales capacity:
{{sales_capacity}}

For each account, output:
1. Priority group
2. Score rationale
3. Recommended owner or role
4. Next action
5. Due date recommendation
6. Nurture trigger if not ready
7. Skip reason if rejected

Be selective. Do not place weak accounts in Work Now.

Create one short email angle, one LinkedIn angle, and one discovery question for each work-now account. The message should connect a specific company signal to a plausible business issue without sounding like surveillance or fake familiarity. Package the handoff in Airtable or Google Docs with account priority, recommended contact, evidence summary, CRM notes, suggested angle, and next step. QA at least five messages by checking each one against the source link and removing anything that sounds too certain, too personal, or too generic. A rep should be able to scan the handoff, understand the reason for outreach, and act without opening 12 browser tabs.

Write first-touch sales angles for these Work Now accounts.

Account briefs:
{{account_briefs}}

Approved tone:
{{approved_tone}}

Outreach channel:
{{outreach_channel}}

For each account, create:
1. Email opening angle under 55 words
2. LinkedIn connection or DM angle under 300 characters
3. Discovery question
4. Source signal used
5. Risky wording to avoid

Rules:
- Do not mention personal profile activity in a creepy way
- Do not overstate the signal
- Connect the signal to a business problem
- End with a soft reason to talk.

After the accounts are worked, update HubSpot and Airtable with reply status, meeting booked, disqualification reason, no-response reason if known, signal quality, message quality, and rep feedback. Review the results by source search, trigger type, buyer role, and disqualifier so the team learns which filters created real conversations. Ask Claude to summarize what predicted replies and what created false positives, then update the research lens before the next Sales Navigator search. Keep a simple monthly change log for filters added, filters removed, and signals promoted or downgraded. The goal is not just one researched list; it is a search and briefing system that gets sharper with every sprint.

Analyze this account research sprint.

Original search criteria:
{{search_criteria}}

Account briefs:
{{account_briefs}}

Outreach outcomes:
{{outreach_outcomes}}

Rep feedback:
{{rep_feedback}}

Output:
1. Signals that predicted replies or meetings
2. Signals that created false positives
3. Buyer roles that performed best
4. Filters to tighten
5. Filters to relax
6. Scoring rule changes
7. Outreach angle changes
8. Next sprint recommendations.