# The Signal Room — PRD Vision
**Product:** CE Internal Sales Intelligence Tool  
**Status:** Pre-build concept — PRD draft for Assaf review  
**Meeting context:** Cropster / Christina Jakob — tomorrow

---

## Vision Statement

CE builds brands by listening better than anyone else in the room.  
The Signal Room is the proof.

It's an internal CE tool that researches any target interlocutor — CMO, brand manager, creative director — and automatically calibrates how CE speaks to them: what we lead with, what vocabulary we use, what service or attitude lands hardest.

It turns "we do great work" into "we already understand you."

---

## The Problem We're Solving (For CE)

CE's pitch is world-class. The tone-of-voice work, the visual research, the pitch pages — all excellent.  
But the sales motion is currently hand-crafted for each prospect. It doesn't scale.

More importantly: CE's core value proposition is **listening at scale and acting on signals**.  
We don't demonstrate that during the sales process. We tell people we do it. We should show it.

The Signal Room closes that gap.

---

## What It Does (Core Loop)

```
Input: a person (name + company, or LinkedIn URL)
         ↓
Step 1: RESEARCH — pull their public signal
        → LinkedIn posts + comments (last 30-90 days)
        → Interviews, talks, podcasts
        → Company news, job postings, announcements
        → Industry conversations they're part of
         ↓
Step 2: SYNTHESIS — build a Profile Card
        → What they care about (stated + implied)
        → The vocabulary they use naturally
        → Their current pressures and ambitions
        → How they think about agencies / partners / creative
        → What "mastery" looks like to them
         ↓
Step 3: TOV CALIBRATION — how CE speaks to THIS person
        → Which CE attitude to lead with (strategic, executional, cultural, technical)
        → Tone register (peer-to-peer, challenger, advisor, craftsperson)
        → What to avoid (what would read as generic / off-putting to them)
         ↓
Step 4: MESSAGING OUTPUT — ready to use
        → Cold email / first-touch LinkedIn message
        → First meeting framing (how Lukas opens the conversation)
        → Pitch page headline variant for this person
        → 3 conversation hooks (based on their actual recent signal)
```

---

## User Stories

**Lukas (primary user):**
> "Before I reach out to a new contact, I drop their LinkedIn into the Signal Room. In 3 minutes I have a profile card and a calibrated first message. I know what they care about. The outreach doesn't feel like a pitch — it feels like I've been paying attention."

**Assaf (strategic user):**
> "Before a first meeting, I run the contact through the tool. I walk in knowing their vocabulary, their current pressures, and what angle will land. I'm not guessing."

**The demo value (what Christina sees tomorrow):**
> "Here's what we found out about you, from your public signal, in the last 30 days. Here's how we adjusted our messaging for you specifically. And here's how that same process runs for every brand in your category."

---

## The Four Interlocutor Profiles CE Targets

Each gets a distinct CE tone calibration:

| Profile | Their Worldview | CE Leads With |
|---|---|---|
| **CMO** (Series B/C SaaS) | Growth mandate, attribution anxiety, team scaling | CE as leverage — agency output without headcount |
| **Brand Manager** (established brand) | Protecting the brand, proving ROI, managing up | CE as the system that runs while they direct |
| **Creative Director** (in-house) | Quality standards, vendor frustration, taste | CE as a peer — our aesthetic judgment, not a brief-follower |
| **Founder/CEO** (SMB) | Everything, all at once, limited bandwidth | CE as the missing creative department |

The tool knows which profile a target most resembles — and calibrates accordingly.

---

## Core Features (MVP)

### 1. Profile Ingestion
- Input: name + company OR LinkedIn URL
- Optional: add context (how we met, what they're working on, their role)

### 2. Signal Research Engine
- LinkedIn posts (last 30-90 days)
- Company announcements / press
- Podcast/interview mentions
- Industry forum participation
- Job postings (what they're hiring reveals strategy)
- Source: web search + Last30 skill + LinkedIn scrape

### 3. Profile Card (Output)
- **What they care about** (top 5 themes from signal)
- **Their vocabulary** (words/phrases they use)
- **Current pressures** (implied from signal)
- **How they see agencies/creative partners** (do they mention vendors? frustrations? results?)
- **Tone register** (formal/casual, tactical/strategic, data-first/intuition-first)
- **Risk signals** (anything that might make CE a hard sell)

### 4. TOV Calibration
- Which CE attitude to lead with
- Tone register recommendation
- What to avoid
- Which CE case studies/examples are most relevant
- Which CE deliverable to feature first

### 5. Messaging Suite
- **Cold email** (subject line + body, 150 words max)
- **LinkedIn first touch** (90 words max)
- **First meeting opener** (how Lukas frames the conversation — 3 sentences)
- **3 conversation hooks** (pulled from their actual recent signal)
- **Pitch page headline variant** (for a personalized landing page)

### 6. Profile Library
- Save + revisit profiles
- Compare messaging across multiple targets at same company
- Tag by segment (CMO / brand manager / CD / founder)
- Track: sent / replied / meeting booked / client

---

## What This Proves in the Room

1. **CE eats its own dog food** — the tool is live, Lukas uses it before every outreach
2. **Signal listening at scale** — we demonstrate the research system that underpins all CE work
3. **Tone of voice is not a document** — it's a dynamic calibration per audience
4. **CE mastery is systematic** — not vibes, not guesswork, not "we'll figure it out after onboarding"
5. **This is what CE builds FOR clients** — and we built it for ourselves first

---

## The Tomorrow Demo (Christina)

**The arc:**
1. Open the tool (even as a Figma prototype or a working alpha)
2. Show Christina's profile card — built from her public signal
3. Show how CE's messaging was calibrated for her specifically
4. Contrast: "Here's how we'd message the Cropster CMO at a consumer brand vs you"
5. Explain: "This is the same system we'd build for Cropster's four audience segments — Origin, Roast, Cafe, Commerce"

**The close:**
> "This is what it looks like when the tone of voice work we did yesterday becomes a living system — not a PDF on a shelf."

---

## Technical Architecture (Phase 1)

```
Frontend: Simple web UI (CE design system — Larken + Inter, CE red)
Backend: Node.js / Python
Research: web_search + LinkedIn scrape + Last30 skill
Synthesis: Claude (sonnet) with CE TOV system prompt
Storage: JSON profile library (local, no DB needed for MVP)
Hosting: staging.curiousendeavor.com/signal-room/
```

**Phase 2:** Connect to Notion CRM. Auto-run on new prospect added. Generate personalized pitch page URL automatically.

---

## Anton's Critique (Pre-emptive)

**Anton says:**

*"The concept is strong but there are three traps:*

*1. This reads as 'AI personalization tool' — which is exactly what every B2B SaaS vendor is selling right now. The differentiation has to be in the output quality, not the concept. If the emails it generates read like GPT cold outreach, the demo will hurt you.*

*2. The demo only works if Christina feels genuinely seen — not just profiled. Lead with 2-3 specific observations from her actual signal (a post she wrote, a talk she gave, a decision she made) before you show the tool. The tool is the explanation for how you knew. Not the lead.*

*3. 'Lukas uses it daily' must be literally true before you say it. A demo prototype is not daily use. Build the habit loop first — even if it's just Lukas running it manually on 3 prospects before tomorrow. Then you can say it honestly."*

**My response to Anton:**

Point 1: Valid — the TOV framework from the Cropster work needs to be deeply embedded in the synthesis layer, not just a system prompt. The output quality has to be unmistakably CE.

Point 2: Correct — the opening to Christina should be "here's what we noticed about you" not "here's our tool." The tool is revealed as the system behind the observation.

Point 3: Fair — we scope the claim correctly. "Lukas ran this before reaching out to you" is true. "Daily use" is the near-term aspiration. We can say that honestly.

---

## Open Questions for Assaf

1. **Scope of tomorrow:** Do we show a working alpha or a designed concept? (I can build a basic working version today — research + profile card output. No full UI.)
2. **The library play:** Do we want to show Christina a pre-built library of 5-6 Cropster-adjacent interlocutor profiles? (Head of Marketing at a coffee brand, CMO at B2B food SaaS, etc.) This makes the contrast compelling.
3. **Name:** "The Signal Room" is my suggestion. It positions the tool as a listening system, not an AI writer. Open to other directions.
4. **Scope post-Cropster:** Is this a CE-internal tool only, or is this eventually a CE product we sell? (Matters for how we build it.)

---

## Recommendation

**Don't lead with the tool tomorrow. Lead with Christina.**

Open with: "We ran your signal through our research system before this meeting. Here's what we found."  
Show her the profile card.  
Then reveal: "This is The Signal Room — the same system we'd build for Cropster's four audiences."

The tool becomes proof that CE does what it claims. Not a feature demo — a demonstration of methodology.

---

*Drafted: 2026-03-31 | Kitt + Anton*