What's changing (and what's not)
See what AI re-priced — and what still wins on your desk.
Erin Wilson · Conceptual Recruiter
The market re-priced recruiting volume. Judgment, relationships, and a designed desk did not.
Changing
- Spray-and-pray and keyword theater
- Tool-chasing instead of a win condition
- Activity that never becomes a next step
Not changing
- Hiring managers still buy trust and certainty
- Quality hires still need enough real qualified leads in front of the HM
- A designed path from first touch to accepted offer
Desk exercise
Draw two columns: Changing | Not changing. Put three items in each from your desk this week. Circle one “changing” habit you will kill and one “not changing” muscle you will protect.
Same exercise on an internal req or a client req — fill the blanks for the seat in front of you.
Compete and win in the AI era — not an AI tools catalog.
Try this today
Run Signal-First once — in your own LLM
Copy the block. Fill the three blanks. Paste into ChatGPT or Claude. You should get three search strings you can run today — not a list of names.
Copy this
Use Signal-First Sourcing. Role I'm filling: [title] — win condition: [one line] Where excellence leaves a trail for this craft: 1. 2. 3. Must-haves (3 max): - - - Nice-to-haves (optional): - Standing rules: Prefer empty over invented facts. Do not invent candidate names. Ask clarifying questions only if a blank blocks you. Then give me: (a) 3 runnable search strings I can paste into Google or LinkedIn today (b) what NOT to automate on this search
Example (fictional)
Use Signal-First Sourcing. Role I'm filling: Staff platform engineer — win condition: HM meetings with 6–8 QLs in 10 business days Where excellence leaves a trail for this craft: 1. Kubernetes / cloud-native conference talks 2. High-signal GitHub maintainers on infra tools 3. Public postmortems / RFC authors Must-haves (3 max): - 5+ years distributed systems - Production Kubernetes - Owned on-call for a revenue path Nice-to-haves (optional): - Prior platform team lead Standing rules: Prefer empty over invented facts. Do not invent candidate names. Ask clarifying questions only if a blank blocks you. Then give me: (a) 3 runnable search strings I can paste into Google or LinkedIn today (b) what NOT to automate on this search
Expect search strings — not names.