Talent Sourcer

Talent Sourcer

When to Use

• Intake a requisition and translate it into search strategy, channels, and boolean strings
• Run boolean, X-ray, and platform-native searches (LinkedIn, GitHub, communities)
• Build a talent map, market scan, or talent pool for a role or company
• Identify and qualify passive candidates with evidence-backed fit notes
• Draft recruiting outreach and multi-touch engagement sequences at scale
• Maintain pipeline hygiene—CRM stages, notes, tags, and duplicate handling
• Apply diversity sourcing considerations without tokenism or biased filters
• Analyze competitor and target-company talent pools for mapping and poachability
• Prepare handoff packages for recruiters and hiring managers with clear next steps

When NOT to Use

• Design interview loops, scorecards, or calibration → interview-system-designer, interview-prep
• Draft offer letters or total comp packages → draft-offer, comp-analysis
• HR policy, ER cases, workforce planning, or org design → hr-business-partner
• Generic B2B sales prospecting or account research → lead-researcher, account-research, enrich-lead (unless explicitly recruiting)
• Sales call prep for non-hiring conversations → call-prep
• Full recruiting program design, employer brand campaigns, or TA strategy → talent-acquisition, cmo-advisor
• Track reqs across hiring stages without sourcing work → recruiting-pipeline (coordination only)

Related skills

Core Workflows

1. Role intake and search strategy

1. Confirm must-haves vs nice-to-haves, level, location, work model, and comp band (directional only) 2. Define ideal candidate profile (ICP): titles, skills, companies, tenure patterns, exclusions 3. Choose primary channels and backup channels by persona (e.g., IC eng → GitHub + LinkedIn) 4. Set weekly sourcing targets and quality bar for handoff 5. Document search strings and platform-specific filters in a reusable playbook block

See `references/role_intake_and_search_strategy.md`.

2. Sourcing execution

1. Run boolean / X-ray searches per channel; log variants and result quality 2. Screen profiles against ICP; tag fit tier (A/B/C) with one-line rationale 3. De-duplicate against CRM and prior outreach; respect do-not-contact lists 4. Capture evidence (profile URL, snippet, project link) for each qualified lead 5. Batch-add to pipeline with consistent field mapping

See `references/sourcing_channels_and_boolean_search.md`.

3. Talent mapping and market research

1. Map target companies, teams, and alumni flows for the role family 2. Estimate pool size and competition (open roles, funding, layoffs) 3. Identify warm paths: referrals, communities, conference speakers, OSS maintainers 4. Produce a talent map artifact: segments, names (where appropriate), and gaps 5. Refresh map on cadence aligned to req urgency

See `references/talent_mapping_and_market_research.md`.

4. Outreach and engagement

1. Personalize first touch from public signals (talks, repos, posts)—no creepy inference 2. Sequence follow-ups with channel-appropriate spacing; stop on opt-out 3. A/B test subject lines and hooks where volume warrants it 4. Track reply rate, interested rate, and conversion to screen 5. Escalate warm leads to recruiter with context package

See `references/outreach_and_engagement_sequences.md`.

5. Pipeline hygiene and handoff

1. Enforce stage definitions: sourced → contacted → replied → interested → handed off 2. Write CRM notes that answer: why fit, risk flags, outreach history, next action 3. Hand off A-tier candidates with profile summary and suggested pitch to HM 4. Report sourcing metrics weekly; flag blockers (low pool, low reply, bad JD)

See `references/pipeline_handoff_and_sourcing_metrics.md`.

When to load references

• Scope and boundaries → references/talent_sourcer_scope.md
• Intake and strategy → references/role_intake_and_search_strategy.md
• Channels and boolean → references/sourcing_channels_and_boolean_search.md
• Mapping and market → references/talent_mapping_and_market_research.md
• Outreach → references/outreach_and_engagement_sequences.md
• Handoff and metrics → references/pipeline_handoff_and_sourcing_metrics.md

Outreach and engagement sequences

Table of contents

1. Principles 2. Personalization framework 3. First-touch templates 4. Multi-touch sequences 5. Channel-specific guidance 6. Compliance and opt-out 7. Measurement

Principles

• Relevant: Tie message to public work (talk, repo, post)—one specific detail
• Concise: Mobile-readable; respect senior candidates' time
• Honest: Real role, real company; no bait-and-switch
• Low friction: Clear CTA (15-min chat, not "apply to 12 steps")
• Persistent but polite: 3–4 touches max for cold; stop on decline

Personalization framework

Avoid: marital status, age cues, political views, health, unrelated personal data.

First-touch templates

Engineering (LinkedIn/InMail):

Hi {name} — I saw your work on {hook} and thought it lined up with
{skill_match}. We're hiring a {title} to {role_hook} at {company}.
Open to a 20-min intro? No pressure if timing's off.

Executive (shorter):

{name}, your track record in {domain} at {company} stood out. We're
building {company_pitch} and looking for a {title}. Worth a brief
conversation?

Referral intro:

{name}, {referrer} suggested I reach out regarding {title} at
{company}. {one_line_pitch}. Happy to share more if relevant.

Customize every send; templates are scaffolding only.

Multi-touch sequences

Example cold sequence (business days):

Rules:

• Vary subject lines; never "just following up" alone
• If open but no reply, try different channel once
• Move to nurture pool if interested but timing bad (6-month revisit)

Channel-specific guidance

Compliance and opt-out

• Include opt-out language where required by org policy
• Honor global unsubscribe and CRM suppression lists
• Do not re-add candidates who declined within cooling period
• Document consent basis per privacy team for EU/UK candidates

Measurement

Track per sequence variant:

Review weekly; pause variants with reply rate <10% after 50+ sends without testing new hooks.

Pipeline handoff and sourcing metrics

Table of contents

1. Pipeline stages 2. CRM hygiene 3. Handoff criteria 4. Handoff package 5. Sourcing metrics 6. Reporting cadence 7. Escalation triggers

Pipeline stages

Standard sourcer-owned stages:

Align stage names with ATS/CRM (recruiting-pipeline may define global req stages).

CRM hygiene

Required fields per candidate:

• Source (boolean v2, referral, GitHub, etc.)
• Fit tier (A/B/C)
• Req ID
• Last activity date
• Next action

Dedup rules:

• Match on LinkedIn URL, email, or ATS candidate ID
• Merge duplicates; preserve full activity history
• Never delete audit trail—mark inactive

Note standard (one paragraph):

Fit: A — staff-level backend, 8y Go/K8s, ex-{company}.
Evidence: {link}; led {project} per profile.
Outreach: 2026-05-01 LI; replied interested 2026-05-03.
Risks: needs visa; comp may be above band.
Next: recruiter screen — suggest HM intro on platform scale.

Handoff criteria

Handoff to recruiter when:

• Candidate is A-tier and interested, OR
• B-tier with HM pre-approval for scarce skill, OR
• Referral with intro made

Handoff to HM only when process allows sourcer→HM skip-level; default through recruiter.

Handoff package

Send recruiter:

1. Profile links (LinkedIn, GitHub, portfolio) 2. Fit summary (3 bullets) 3. Outreach history (dates, channels, responses) 4. Suggested pitch angle for screen 5. Open questions (visa, comp, level)

Attach candidate list export if batch handoff.

Sourcing metrics

Set targets with recruiter based on role difficulty.

Reporting cadence

Weekly sourcer report (5-minute read):

• Qualified adds vs target
• Reply and interest rates by sequence
• A-tier backlog awaiting handoff
• Blockers: JD clarity, comp, visa, channel fatigue
• Plan next week: booleans to test, companies to map

Escalation triggers

Escalate to recruiter/HM/TA when:

• <2× pool size after 1 week of search
• Reply rate <8% after 100+ sends with tested variants
• HM rejects >80% of A-tier handoffs (ICP drift)
• Legal flags on target company list
• Diversity slate gap vs plan with no path forward

Document escalation with data, not anecdotes.

Role intake and search strategy

Table of contents

1. Intake checklist 2. Ideal candidate profile 3. Search strategy canvas 4. Channel selection 5. Exclusions and knock-outs 6. Playbook template

Intake checklist

Gather from hiring manager, recruiter, and JD before sourcing:

Document ambiguities as assumptions until confirmed.

Ideal candidate profile

Structure ICP as searchable fields:

Titles:     [e.g., Staff SWE, Principal IC]
Seniority:  [years, scope: team vs org]
Skills:     [required keywords + synonyms]
Companies:  [tier 1 targets, acceptable, avoid]
Trajectory: [promotion pattern, scope growth]
Signals:    [OSS, talks, patents, certifications]
Anti-patterns: [job hop <12mo x3, wrong domain]

Map each ICP element to at least one channel + filter.

Search strategy canvas

Set weekly targets: profiles reviewed, qualified adds, outreach sent, replies.

Channel selection

Align channel effort with where ICP density is highest.

Exclusions and knock-outs

Document hard excludes to save review time:

• Wrong level or function
• Geographic ineligibility
• Competitors under NDA/no-poach
• Recent outbound by another req (cooling period)
• Former employee on do-not-rehire list (recruiter-owned)

Soft flags (review with recruiter): short tenure, title inflation, gap years.

Playbook template

# Sourcing playbook: [Role] — [Req ID]

## ICP summary
[3–5 bullets]

## Booleans (LinkedIn)
- Core: ...
- Variant A: ...
- Variant B: ...

## X-ray / web
- site:github.com ...
- site:linkedin.com/in ...

## GitHub
- language:... stars:>...
- org:...

## Weekly targets
| Metric | Target |
|--------|--------|
| Qualified adds | |
| Outreach sent | |
| Replies | |

## Handoff bar
A-tier: [criteria]
B-tier: [criteria]

Store playbook in CRM or shared doc; version when JD changes.

Sourcing channels and boolean search

Table of contents

1. Boolean fundamentals 2. LinkedIn sourcing 3. X-ray and web search 4. GitHub and technical sourcing 5. Communities and events 6. Boolean pattern library 7. Quality and iteration

Boolean fundamentals

Rules:

• Start broad, then narrow with NOT and filters
• Keep a variant log (v1, v2) with result counts
• Prefer title + skill + company over keyword soup
• Test synonyms: "kubernetes" OR "k8s"

LinkedIn sourcing

Typical filter stack:

1. Title keywords (current + past) 2. Skills (required vs optional) 3. Geography + open to remote 4. Company (current/past) lists 5. Seniority / years of experience 6. Activity signals where available (open to work — use ethically)

Recruiter-style strings (adapt to platform syntax):

("staff engineer" OR "principal engineer") AND (rust OR golang)
NOT (recruiter OR "talent acquisition")

Save searches; re-run on cadence for new profiles.

X-ray and web search

Find public profiles outside native recruiter UI:

site:linkedin.com/in ("distributed systems" OR "data platform")
("Acme Corp" OR "Beta Inc") -intitle:recruiter

site:github.com "location: San Francisco" language:Rust stars:>50

Respect robots.txt and terms of service; prefer official APIs where licensed.

GitHub and technical sourcing

Signals for engineering ICs:

Cross-link GitHub username to LinkedIn via public bio or search—do not assume identity without confirmation.

Communities and events

Engage in community norms—value-first, not spam.

Boolean pattern library

Backend IC (example):

(title:"software engineer" OR title:"backend engineer" OR title:"SWE")
AND (python OR java OR go)
AND (microservices OR "distributed systems")
NOT (intern OR student OR recruiter)

Product leader (example):

(title:"product manager" OR title:"group product manager")
AND (B2B OR SaaS OR "enterprise")
AND (roadmap OR "go-to-market")

Security (example):

(title:"security engineer" OR title:"application security")
AND (OWASP OR "threat modeling" OR AppSec)

Customize per ICP; store winners in playbook.

Quality and iteration

Review boolean performance weekly; retire strings with <5% qualify rate after sufficient volume.

Talent mapping and market research

Table of contents

1. When to build a talent map 2. Map structure 3. Company and competitor analysis 4. Market sizing 5. Warm path identification 6. Talent map artifact 7. Refresh cadence

When to build a talent map

Build a map when:

• New role family or geo with unknown pool
• Executive or scarce skill (low boolean yield)
• Build-vs-buy team stand-up (hire 5–20 in quarter)
• Competitive poach campaign with named target companies
• HM requests "who exists at X?"

Skip heavy mapping for high-volume evergreen reqs with proven booleans.

Map structure

Layers (bottom to top):

1. Market — segment, demand drivers, compensation pressure 2. Companies — tiers: primary targets, secondary, avoid 3. Teams — org units known to house ICP 4. Individuals — named profiles (where appropriate and lawful) 5. Relationships — referrals, alumni, community ties

Company and competitor analysis

For each target company capture:

Talent pool analysis = aggregate patterns, not dossiers on every employee.

Market sizing

Rough funnel for planning:

Addressable titles in geo
  × % with must-have skills (estimate from sample searches)
  × % likely reachable (not founder-only, etc.)
  = actionable pool

Compare to hiring target and timeline. If pool < 3× target, widen ICP or geo.

Warm path identification

Log warm path on CRM contact for higher priority outreach.

Talent map artifact

Template:

# Talent map: [Role] — [Date]

## Executive summary
- Pool estimate: N qualified in geo
- Top 5 target companies: ...
- Key risk: [competition / visa / remote policy]

## Company tiers
| Tier | Companies | Rationale |
|------|-----------|-----------|
| A | | |
| B | | |

## Named targets (A-tier sample)
| Name | Company | Title | Link | Notes |
|------|---------|-------|------|-------|

## Warm paths
- ...

## Gaps
- Underrepresented segments to expand
- Skills in short supply

Share map with recruiter/HM; redact if exporting externally.

Refresh cadence

Archive superseded maps; note date and req version.

Talent sourcer scope

Table of contents

1. Purpose 2. In scope 3. Out of scope 4. Deliverables 5. Quality bar 6. Ethics and compliance

Purpose

The talent sourcer role is proactive identification and engagement of candidates who may not be actively applying. Success is measured by qualified pipeline volume, reply and interest rates, diversity of slates, and clean handoffs—not by closing offers or designing interviews.

In scope

Out of scope

Deliverables

Typical artifacts the sourcer produces:

• Sourcing playbook — ICP, booleans, channels, weekly targets
• Candidate list — tiered with links and fit notes
• Talent map — companies, teams, segments, warm paths
• Outreach pack — templates + personalization tokens
• Handoff brief — per A-tier candidate for recruiter/HM
• Metrics snapshot — funnel counts and conversion rates

Quality bar

• Every qualified lead has a public evidence line (why this person, why now)
• No fabricated credentials; verify title/company from primary profile
• Dedupe before outreach; honor prior contact and do-not-contact
• Slates reflect diversity goals as defined by TA/legal—not quotas on protected classes in search filters where prohibited
• Handoff includes risks (visa, level mismatch, competing process) when known

Ethics and compliance

• Use only lawful data sources and platform terms of service
• Do not scrape private data or bypass paywalls/login walls without authorization
• Respect GDPR/CCPA where applicable: document lawful basis for processing candidate data per org policy
• Avoid discriminatory search filters (e.g., age, gender, ethnicity proxies)
• Stop outreach immediately on unsubscribe or explicit decline