segment-scout
/segment-scout
Ocenia segmenty rynku według dopasowania i dostępności z cytowanymi dowodami, a potem wskazuje od 5 do 10 firm.
- Autor
- Leuenaar ↗
- Repozytorium
- async-contractor-agents ↗
- Licencja
- MIT
- Ostatnia zmiana repozytorium
- 11.11.2026
Ocena i bezpieczeństwo
- Pliki w paczce
- 4
- Skrypty
- 0
- Komentarze HTML
- 0
- Adresy URL w SKILL.md
- 0
Brak sygnałów. Repozytorium zmienione 11.11.2026, bez odwołań do wycofanych modeli i API.
MIT, otwarta: można używać i zmieniać, z zachowaniem informacji o autorze.
Co robi i kiedy
Ocenia segmenty rynku według dopasowania i dostępności z cytowanymi dowodami, a potem wskazuje od 5 do 10 firm.
Oryginalny opis (EN)
Scores market segments before picking companies to approach, using weighted fit and reachability criteria with cited evidence, then lists 5 to 10 companies inside the winning segment. Use when choosing a niche, an outreach target list, or deciding where to look for clients next.
Jak działa
- Input
- Steps
- Weights
- Negative signals (subtract)
- Rules
- Done when
- Output
- Not this skill's job
Podgląd SKILL.md
---
name: segment-scout
description: Scores market segments before picking companies to approach, using weighted fit and reachability criteria with cited evidence, then lists 5 to 10 companies inside the winning segment. Use when choosing a niche, an outreach target list, or deciding where to look for clients next.
license: MIT
metadata:
author: Harva Heru
version: "0.1.0"
shared: "evidence-rule,anti-slop"
---
# Segment scout
**Core rule:** pick the segment first, the company second. A perfect company inside a segment that cannot pay or cannot be reached is wasted research.
## Input
- A market idea or a list of candidate segments (for example "plumbing businesses in Singapore", "small n8n agencies").
- The owner's current proof: which demos, test reports and case studies exist. Read them, do not assume.
- Constraints: async only, no live calls, timezone UTC+7.
If the proof list is missing, ask once.
## Steps
1. List 3 to 6 candidate segments. A segment is a type of buyer plus a place, never a single company.
2. Score each segment 1 to 5 on the six factors below. Every score cites a source opened in this session, or is marked "unverified" and capped at 2.
3. Multiply by the weights and rank.
4. Apply the negative signals as subtractions after the weighted total.
5. For the top segment only, find 5 to 10 real companies and note one specific detail per company (a slow reply, a broken form, an after-hours gap, a job post). That detail is what an outreach email will lead with.
6. Write a one-line prediction ("expect 1 reply from 10 emails") and save it, so the result can be compared later.
## Weights
| Factor | Weight | 5 looks like | 1 looks like |
| --- | --- | --- | --- |
| Ability to pay | 25% | Spends on software or contractors already | Price is the only decision factor |
| Strength of the owner's proof for this need | 25% | A tested demo of this exact job exists | Nothing built that resembles it |
| Pain visibility | 15% | The leak can be seen from outside (test enquiry, public reviews) | Pain has to be guessed |
| Reachability, async | 15% | Public email or form, decision maker reads it | Only phone or in-person |
| Cost to serve | 10% | Same build reused with small changes | Custom work every time |
| Competition | 10% | Few sellers show proof for this | Crowded with cheap gigs |
## Negative signals (subtract)
- Buyers expect calls or WhatsApp voice notes as the main channel: minus 1.0.
- Segment already contacted in the last 30 days with zero replies: minus 0.5.
- Legal or regulated data (health, finance) the owner cannot handle yet: minus 1.0.
## Rules
- Follow [evidence-rule](references/evidence-rule.md): a score without a source is "unverified".
- Follow [anti-slop](references/anti-slop.md) in the written summary.
- Stop at the target count. More companies than needed is wasted context.
- Never suggest warm-network or referral routes unless the owner asks; this repo assumes none.
## Done when
- [ ] Every factor score has a link or is marked unverified.
- [ ] The top segment beats the second by a stated margin, or the tie is called out.
- [ ] Each listed company has one specific, checked detail.
- [ ] A prediction is written down with a date.
## Output
```
| Segment | Pay | Proof | Pain | Reach | Cost | Comp | Weighted | Neg | Final |
| SG plumbing 1-20 staff | 4 | 5 | 4 | 4 | 4 | 3 | 4.15 | -0.5 | 3.65 |
Top segment: ...
Companies: name, site, detail, source link
Prediction (2026-11-10): ...
```
## Not this skill's job
Writing the emails (use `outreach-drafter`). Screening job listings (use `job-screener`).
Research behind this skill: [references/research.md](references/research.md).
Pliki w paczce (4)
- SKILL.md 4 KB
- references/anti-slop.md 3 KB
- references/evidence-rule.md 2 KB
- references/research.md 1 KB
Wersje
W bazie nie ma innych skilli o nazwie segment-scout.
Migawka bazy z 13.11.2026 · wersja paczki 2bb829a7fa23 (skrót SHA-256 całej paczki)
Pobrania w ostatnich 30 dniach: 0 · ranking
Zgłoś problem z tym skillem: szkodliwa treść, błędny opis albo prośba autora o usunięcie.
Podobne skille
- 90segment-company-search
Szuka firm kandydatów w podanej niszy B2B, weryfikuje je pod kątem dopasowania segmentu i sygnałów ICP, a potem zapisuje priorytetowy arkusz. - 100gtm-segmentation
Definiuje i priorytetyzuje segmenty rynku, szacując je tylko z danych użytkownika, domyślnie z podejściem SMB-first. - 90customer-segments
Tworzy dokumenty segmentów klientów i użytkowników, w tym segmenty cyklu życia i podział odbiorców. - 100company-lists
Zapisuje firmy i porządkuje je w listach Hunter: zapisywanie firm, dodawanie i usuwanie z list, tworzenie list i folderów oraz masowe kopiowanie, przenoszenie i usuwanie. - 100score-my-list
Ocenia i szereguje listę firm lub kontaktów według dopasowania do ICP i sygnałów zakupowych, dzieląc ją na tiery A, B i C. - 100segmentation
Segmentuje rynek lub bazę użytkowników według potrzeb i zachowań, szacuje wielkość i ocenia segmenty oraz wybiera segment przyczółkowy z uzasadnieniem, w tym TAM/SAM na segment.