Harmony HR
01 1. Sourcing as a System02 2. The Client Brief Before Search03 3. Candidate Portrait04 4. Market Map05 5. Boolean, X-Ray06 6. LinkedIn07 7. Beyond LinkedIn08 8. GitHub and Technical Traces09 9. First Contact10 10. Candidate Evaluation11 11. Working with Clients12 12. Metrics13 13. Tools and AI14 14. Operational Model15 15. Practical Exercises
A Materials NavigatorB Mini GlossaryC RegulationsD Template LibraryE End-to-End CasesF Data, AI, PrivacyG AI Prompt Library
Home/Chapter 06

Chapter 06. LinkedIn as a Working Environment

pp. 32–37
LinkedIn as a work environment
LinkedIn as a work environmentHow LinkedIn fits into a disciplined sourcing workflow.
LinkedIn routes
LinkedIn routesThe main paths a sourcer uses inside LinkedIn to move from search to contact.

Evidence note: LinkedIn Future of Recruiting 2025 shows growing roles for AI, hiring quality,

and skills-based hiring. Tools accelerate the process, but sourcing quality still depends on

human judgement, structured criteria, and candidate trust.

How to Use This Chapter

This chapter is longer than the others because LinkedIn often becomes the sourcer's primary

working environment. Do not read it linearly if you have a narrow task: choose the search

strategy, profile reading, first contact, or team process section.

Remember in One Phrase

LinkedIn is useful not as an endless feed of profiles but as a managed loop: hypothesis, fact

evaluation, contact, status, and feedback.

If Your Task IsGo to SectionWhat You Get
Set up search by role6.1 Search StrategyProject, filters, and selection rules
Understand who to write to6.2 Reading ProfilesAn evaluation map before first contact
Improve messages6.3 First ContactA formula for a respectful first message
Remove team chaos6.4 Team ProcessStatus, duplicate, and handover rules

Quick Chapter Map

1. Start with a project and a hypothesis.

2. Read profiles by facts, not by impression.

3. Write only to people for whom you have a clear reason to contact.

4. Record the status, the owner, and the next step.

If time is short, follow the route you need: search in 6.1, profile reading in 6.2, messages in

6.3, team discipline in 6.4.

Minimal Start

In 45 minutes, pick one hypothesis, review 20 profiles, score the top 10 against the map,

and prepare 3 personalised messages. Do not launch a mass outreach until you have verified

the pool's relevance and the message angle.

6.1 Search Strategy

LinkedIn should be used as a research environment, not an endless feed.

Route Purpose How to Use

People search Direct candidate search Job titles, keywords, company,

location, level

Company search Donor company map Company page → people → current

and former employees

Alumni search Clusters of graduates from companies or Use as an additional route, not

programmes a quality criterion

Similar profiles Expansion from a strong profile Collect job titles, keywords,

companies

"Also viewed" Adjacent candidates Watch for drift and duplicates

Content search Speakers and practitioners Posts, articles, comments, events

Project Setup Checklist

One project or tracking view per role.

Inside: the search brief and mandatory traits.

Tags: direct match, adjacent profile, calibration, contacted, responded, irrelevant.

Source route recorded for each candidate.

Notes: facts, ambiguity, first-message angle.

Ownership check before messaging.

Duplicate check before messaging.

Daily LinkedIn Workflow

1. Open the search brief.

2. Pick one hypothesis.

3. Run one search route.

4. Review 25–40 profiles.

5. Save only profiles with facts or useful ambiguity.

6. Add a note before moving to the next profile.

7. Send a small batch of relevant messages.

8. Log responses and rejection patterns.

9. End with one conclusion: keyword, job title, company, or message angle.

6.2 Reading Profiles

Read facts, not "impressions":

Current role and scope of responsibility.

Previous roles and career trajectory.

Domain relevance.

Company environment and scale.

Tools, projects, product or market context.

Achievements with outcomes.

Level signals: task ownership, ambiguity, impact, complexity.

Motivation signals: tenure at current employer, growth ceiling, public interest in a topic,

relocation signals.

Red flags and ambiguity.

What cannot be inferred from the profile.

Profile Evaluation Map

Criterion 0 1 2

Mandatory traits No Possibly Clearly present

Relevant environment No Adjacent Direct

Level and scale Does not match Unclear Matches

Transferability of experience Weak Possible Strong

First-message angle None Generic Specific

Risk/ambiguity High Manageable Low

Decision Rule: 9–12 — contact; 6–8 — save or calibrate; 3–5 — reject unless the market is

extremely narrow; 0–2 — reject.

6.3 First Contact

Five-part formula:

1. A relevant observation.

2. Why the candidate's experience connects to the role.

3. An opportunity framed from the candidate's perspective.

4. A low-pressure next step.

5. A respectful exit without pressure.

Example for a Technical Candidate:

Hi Maya,

I noticed that your backend experience includes event-driven services and payment system

reliability.

We are looking for a senior engineer for a role involving high-load transactional flows where

fault tolerance and observability matter.

This may be relevant if you are interested in a deeper systems ownership scope.

Would it be appropriate to send a short brief? If the timing is not right, that is fine.

Poor → Better:

Poor: You have an impressive profile. We are growing fast. Interested?

Better: I noticed your experience with distributed systems in a marketplace environment.

The role focuses on high-load order/payment flows, so your reliability experience looks

relevant.

I can send a short brief if this is potentially interesting.

6.4 Team Process

One contact owner per candidate.

ATS/CRM check before messaging.

Source and date recorded.

Notes contain facts, ambiguity, and message angle.

Tags are used consistently.

No duplicate messages from different people.

Handover to the recruiter includes facts, message history, motivation, salary expectations

(if known), and open questions.

The weekly review covers searches, share of relevant profiles, response rate, and client

feedback.

Self-Check

Can another team member open a candidate card and understand: contact, wait, reject, or hand

over?

AI Tips for This Chapter

The full prompt for personalising the first message is in Appendix G.5. First hand-write 2–3

facts from the profile, then ask AI to generate variants. Send only after manually verifying

facts, tone, and disclosure constraints.

Key Takeaways / Where to Go Next

Take away the LinkedIn working rhythm: hypothesis, profile evaluation, small message batch,

status logging, and feedback. If a single LinkedIn source narrows the market, move to

Chapter 7 to add alternative sources.

Page 32 Page 33 Page 34 Page 35 Page 36 Page 37