---
name: whitespace-prospecting
description: Build a scored TAM list for a new segment, diffed against the CRM so it's true whitespace. Use this skill when: New segment, then monthly — One paragraph naming the segment kicks off the build.
---

# Build a scored whitespace target list

Goal: build a scored whitespace list for a new segment and keep it current.

Connected tools this skill uses: hubspot, googlesheets, slack.

## When to use

- New segment, then monthly — One paragraph naming the segment kicks off the build.
- Or when the user asks directly to build a scored whitespace list for a new segment and keep it current.

## Method

Build the whitespace list for {{segment}}.

1. Read the ICP definition for {{segment}} from memory; if this is the first
   run for this segment, ask for it — industry, size, signals — and save it
   before continuing.
2. Build a candidate company list for {{segment}} matching the ICP, sourced
   from the web.
3. Pull every account in {{crm}} and remove candidates already owned, in an
   active deal, or closed-lost within the last {{cooldown_months}} months.
4. Score the remaining list 0-100 against the ICP criteria and bucket into
   act-now, nurture, or long-term.
5. Re-check the top 20 act-now accounts against {{crm}} a second time before
   publishing, to confirm none became an active deal mid-run.
6. Update the whitespace-prospecting view with the scored list and save a
   snapshot to `reports/whitespace/{{segment}}-{{date}}.md`.
7. Post the top 20 to #sales-ops, with a one-line reason each.

Do not add any account to an outbound sequence directly — hand the list to
the team for review and assignment first.

## ICP fit scoring

Score every candidate company 0-100 against the stored ICP definition for the segment.

## Method
1. Read the segment's ICP definition from memory: industry, size, signals.
2. For each candidate company remaining after the CRM diff, check it against each ICP criterion.
3. Score the company 0-100 based on how many criteria it matches and how strongly.
4. Bucket each scored company into act-now, nurture, or long-term based on its score.
5. Write a one-line reason for the score, citing the specific criteria matched.

## Rules
- The ICP definition (industry, size, signals) is set on first run per segment, then stored in memory and reused every refresh.
- A score without a stated reason never publishes — every company needs its one-line justification.
- The score thresholds for act-now / nurture / long-term buckets are set on first run, then stored in memory.

## Output
A scored list: company, 0-100 score, bucket, one-line reason — ready for CRM diff's final safety pass.

## CRM diff

Remove any candidate company already owned, in an active deal, or inside the closed-lost cooldown window.

## Method
1. Pull every account currently in the CRM.
2. Remove candidates that match an owned account, an account in an active deal, or an account closed-lost within the cooldown window.
3. Re-run this check a second time on the top 20 act-now accounts immediately before publishing.
4. If any account changed status mid-run (became active), remove it at this second pass even if it scored well.

## Rules
- The cooldown period after closed-lost, before an account is eligible again, is set on first run, then stored in memory.
- The top 20 act-now accounts always get a second, fresh CRM check right before publishing — no exceptions.
- Never publish a candidate still inside its cooldown window, regardless of ICP score.

## Output
The candidate list with owned, active, and cooldown accounts removed — true whitespace only — flowing into ICP fit scoring.

## TAM sourcing

Build the initial candidate company list for a segment from public sources.

## Method
1. Read the segment's ICP definition (industry, size, signals) from memory.
2. Search public sources — company directories, news, funding databases — for companies matching those criteria.
3. Record each candidate with the specific criteria it appears to match.
4. Deduplicate the list before handing it to CRM diff.
5. Note the source for each candidate's inclusion, so a later reviewer can trace why it's on the list.

## Rules
- Only include a company if it matches at least one concrete ICP signal, not a vague industry guess.
- Every candidate carries a note on why it was included, tied to a specific ICP criterion.
- Never source from a paywalled or unverifiable list without recording where it came from.

## Output
A deduplicated candidate company list with per-company inclusion notes, handed to CRM diff before scoring.

## Recurring jobs

This skill pairs with scheduled jobs. Offer to set each one up; only create the ones the user approves:

- Monthly whitespace refresh (1st business day of the month at 9:00 AM) — https://eluu.ai/cookbook/whitespace-prospecting/jobs-setup/monthly-whitespace-refresh.md
- New segment brief detection (Every day at 9:00 AM) — https://eluu.ai/cookbook/whitespace-prospecting/jobs-setup/new-segment-brief-detection.md