Reference file

Screening rubric

screening-rubric.md

Screening rubric

The bar every candidate is measured against. Criteria 1 and 2 are credibility and are weighted first; 3 supports them; 4 is amplification; 5 is a tiebreaker only.

The five criteria

1. Field experience - credibility, weighted first

Meaningful, demonstrable experience in the company's specific field, not adjacent or general. "Ran demand generation at a B2B SaaS company" counts for a B2B demand-gen ambassador; "20 years in marketing" does not. Evidence: roles held, work output (talks, articles, case studies), and how specifically they speak about the field's real problems rather than its surface.

2. Independence - credibility, hard gate

  • Automatic out: current full-time employee of a direct competitor; current full-time employee of any company where publicly representing another brand would be a visible conflict; a paid advisor or board member of a direct competitor.
  • Downgrade, not out: full-time at a non-competitor (they can still advise, but public representation is a harder ask); "advisor to 15 startups" with no visible focus - the spread dilutes credibility, so treat their reach and content as worth less.
  • Ideal: consulting, fractional, freelancing, between roles, or running a practice focused on this field.

3. Seniority signal - a proxy for scale

Held a leader role or above at a company that is a notable brand at some point in their career. This is a proxy for having operated at real scale, not a requirement that they still hold the title. Sanity-check against the actual company revenue or size and role scope at the time - titles inflate at small companies and compress at large ones.

Step 3b - Verify the top candidates

For the top 3-5 candidates that pass criteria 1 and 2, pull their recent LinkedIn activity. Use it to:

  • check the red flags: dormant 12+ months, all promotional or sponsored, generic posts with no depth
  • upgrade "publishes content" to confirmed, or downgrade it
  • spot current-role signals that contradict the search results Do this for the top candidates only, not the full list.

4. Reach - amplification, screened after credibility

Platform floors, as starting defaults. Adjust to whichever platform is genuinely native to the field.

Platform Default floor Notes
LinkedIn 6,000 followers / connections The default for most fields
X / Twitter 3,000 followers Only if the field's conversation actually lives there
Substack / newsletter 2,000 subscribers An engaged list beats a larger cold one
YouTube 5,000 subscribers For fields where video is the native format
Industry forum / community "recognized name" Qualitative - do the regulars know them?

Reach below the floor does not automatically disqualify when field credibility is exceptional and the angle is advisory rather than public - but note it explicitly on the row so the user decides with eyes open.

5. Publishes original content - tiebreaker only

Actively publishes original material about the field: LinkedIn posts, articles, a blog, a newsletter, a podcast. Moves a borderline candidate up the shortlist. It never lifts a candidate who fails criteria 1 or 2, and search naturally over-favors it (visible people are easier to find), so weight it lightly.

Red flags - sink an otherwise-qualified candidate

  • Engagement that doesn't match the follower count (bought or inflated audience)
  • A feed that is all promotional or sponsored posts - audience trust is already spent
  • Dormant 12+ months on their main platform
  • Public controversy or stated positions that would create brand risk
  • Speaks only in generalities about the field - visibility without demonstrated depth

The evidence bar

  • Confirmed: direct evidence in what search actually surfaced - a bio, an article byline, a documented role history, a recorded talk.
  • Likely: a strong circumstantial signal - e.g. a personal site describing current consulting work implies independence but does not confirm the absence of a conflicting full-time role.
  • Unverified: no evidence either way. Not a negative mark - a to-do. Most reach and current-employment cells start here and are cleared by a manual check.
  • Freshness: for current role and independence, use the most recent dated source. If sources conflict, the newer one wins. Evidence older than 12 months caps the cell at "unverified", however confident the source sounds.

Tiering the shortlist

  • Approach now: field experience and independence both confirmed, LinkedIn activity checked with no red flags (dormant, all-promotional, generic), seniority signal at confirmed or likely. Reach may stay unverified; list the follower check as the one manual step before outreach.
  • Nurture / revisit: credibility confirmed, but reach is unverified or below floor, or independence is a downgrade rather than an out.
  • Pass: fails field experience, fails independence at the hard gate, or carries a red flag that outweighs the rest.