tracksolid_timescale_grafana_prod/docs/superpowers/specs/2026-04-18-fleet-intelligence-pitch-design.md

# Fleet Intelligence — Partner Pitch Design
**Date:** 2026-04-18  
**Author:** David Kiania / Fireside Communications  
**Target:** Regional GPS tracking reseller (East Africa, mixed SME/enterprise client base)  
**Relationship:** Existing — structured meeting already booked  
**Outcome:** Exploratory partnership (white-label / revenue share / licence / joint GTM)  
**Format:** Slide deck sent ahead; one-pager conversation anchor at meeting

---

## 1. Pitch Philosophy — The Elicitation Method

Do not open with features. Open with questions the prospect cannot answer.

Every question surfaces a pain their clients are already feeling — silently, expensively. The prospect feels that pain on their clients' behalf. By the time the intelligence layer is revealed, the prospect is not evaluating a product. They are relieving a pressure they just felt.

**Structure:** 6 unanswered questions → dramatic reveal → feature answers → business case → flexible partnership close.

---

## 2. The 6 Elicitation Questions

Each question gets its own slide. Large type. One implication line. No answer yet.

---

### Q1 — The Fuel Leak
> *"Which of your clients' drivers is costing them the most money in fuel — right now, this month?"*

**Implication:** Every fleet has a 20% fuel variance between its best and worst driver. Without this answer, that money disappears quietly every month.

**Intelligence feature revealed:** Driver Behaviour Score — trip-by-trip fuel cost, speeding events, harsh braking, idle time per driver, ranked fleet-wide.

---

### Q2 — The Ticking Clock
> *"Which vehicle across your entire client base is overdue for a service today?"*

**Implication:** A $300 service missed becomes a $3,000 breakdown. And when it happens, your client blames the tracker company, not the maintenance manager.

**Intelligence feature revealed:** Service Interval Forecaster — odometer-based service alerts, service history log, upcoming service calendar per vehicle, configurable thresholds by vehicle type.

---

### Q3 — The Dead Hours
> *"Which of your clients had a vehicle parked for 3+ hours during working hours yesterday — and why?"*

**Implication:** Idle time is the silent tax on every fleet. Operations managers know it exists. They cannot prove it, attribute it, or act on it.

**Intelligence feature revealed:** Fleet Utilisation Dashboard — idle time per vehicle per day, utilisation rate, ghost shifts, vehicle productivity score, city-level cohort comparison.

---

### Q4 — The Dispute
> *"If a driver disputes an overtime claim tomorrow, can your client pull the proof in under 60 seconds?"*

**Implication:** Without a timestamped trip audit trail, every dispute becomes the driver's word against the manager's. HR decisions get made on gut, not data.

**Intelligence feature revealed:** Trip & Stop Audit Trail — full GPS replay, stop duration and address, driver timeline per day, exportable to PDF or CSV for HR/legal.

---

### Q5 — The SLA Gap
> *"Do your clients know if their field teams are actually reaching customers within their promised response window?"*

**Implication:** ISPs, telcos, and field service companies are contractually measured on SLA. Without dispatch intelligence, they are reporting to their own clients on guesswork.

**Intelligence feature revealed:** Dispatch & SLA Intelligence — first-response time, job completion rate, field team coverage maps, SLA breach alerts, trend by city and team.

---

### Q6 — The Unit Economics
> *"Do your clients know what it actually costs them to serve one customer — per trip, per kilometre, per job completed?"*

**Implication:** A school bus operator who drove 450 km this week cannot tell a school board whether that was efficient or catastrophic without knowing routes completed, cost per child per day, and km per route. The same gap exists in every vertical: logistics, courier, field service, NGO.

**Intelligence feature revealed:** Unit Economics Engine — cost per km, cost per job/route/delivery, output ratio (km driven vs customers/deliveries/routes served), configurable per client vertical.

---

## 3. Full Deck Structure

### ACT 1 — THE QUESTIONS (Slides 1–8)

| Slide | Content |
|---|---|
| 1 | Cover: *"6 Questions Your Clients Cannot Answer Today"* — Fireside Communications, date |
| 2 | Context setter: *"You give your clients a dot on a map. That solved visibility. This is about what comes next."* |
| 3 | Q1 — The Fuel Leak |
| 4 | Q2 — The Ticking Clock |
| 5 | Q3 — The Dead Hours |
| 6 | Q4 — The Dispute |
| 7 | Q5 — The SLA Gap |
| 8 | Q6 — The Unit Economics |

### ACT 2 — THE REVEAL (Slides 9–15)

| Slide | Content |
|---|---|
| 9 | Pivot: *"We built the answers."* — one line, full bleed |
| 10 | The Intelligence Layer — what it is, what it sits on, what it connects to |
| 11 | Answer to Q1 — Driver Behaviour Score |
| 12 | Answer to Q2 — Service Interval Forecaster |
| 13 | Answer to Q3 — Fleet Utilisation Dashboard |
| 14 | Answer to Q4 — Trip & Stop Audit Trail |
| 15 | Answer to Q5 — Dispatch & SLA Intelligence |
| 16 | Answer to Q6 — Unit Economics Engine |

### ACT 3 — THE BUSINESS CASE (Slides 17–21)

| Slide | Content |
|---|---|
| 17 | Client verticals — how each feature maps to logistics, school transport, ISP/telco, courier, NGO |
| 18 | Revenue expansion — *"You earn $X per device per month today. Here is what the intelligence layer adds."* (fill ARPU before sending) |
| 19 | How it works with your stack — white-label ready, Tracksolid/Jimi API compatible, no rip-and-replace |
| 20 | Partnership models — 4 options, flexible |
| 21 | Close |

### ACT 4 — OPTIONAL ADD-ONS (Slides 22–25)

| Slide | Content |
|---|---|
| 22 | Intro: *"Two optional add-ons for clients who need more than intelligence — they need storage and history."* |
| 23 | Blob Storage (RustFS) — pain question, what it stores, pitch angle |
| 24 | Data Warehouse (DuckDB) — pain question, what it enables, pitch angle |
| 25 | Tier model — Intelligence / +Storage / +Warehouse / Full Stack |

### CLOSE — Slide 21

> *"We already run this for an 80-vehicle telco field service fleet across Nairobi, Mombasa, and Kampala. The stack is live. The question is whether your clients should have it next."*

**CTA:** Book a 45-minute working session. We pull live data from one of your client's fleets and show you what we see.

---

## 4. Partnership Models (Slide 20 Content)

Present all four — let them name the one they want.

| Model | Description | Best for |
|---|---|---|
| **White-label** | Intelligence layer runs under their brand; they bill clients directly | Resellers wanting full client ownership |
| **Revenue share** | Fireside builds and operates; reseller earns % of intelligence tier revenue | Resellers with no technical capacity |
| **Technology licence** | Reseller buys the stack and runs it on their own infrastructure | Resellers with in-house tech team |
| **Joint GTM** | Both companies co-sell a combined tracking + intelligence product to new clients | Resellers wanting to expand upmarket |

---

## 5. Intelligence Layer — Technical Summary (Slide 10 Content)

Three sentences for the deck:

> Fleet Intelligence is an analytics and alerting layer that sits on top of any GPS tracking feed. It transforms raw position data into operational decisions: driver scores, service forecasts, utilisation reports, SLA compliance, and unit economics — all in real time, configurable per client. It is white-label ready and does not require replacing your existing hardware or tracking platform.

**Stack (internal reference, not for deck):**
- Ingestion: Python → TimescaleDB/PostGIS (60s position poll, 30m high-res trail, webhook push)
- Analytics: SQL aggregates → `dwh_gold` layer → Grafana dashboards
- Deployment: Docker Compose on Coolify; self-hosted, client-isolated
- API source: Tracksolid Pro / Jimi IoT Open Platform (compatible with any Jimi-based reseller stack)

---

## 6. One-Pager Layout (Meeting Leave-Behind)

**Front — The 6 Questions**
- Headline: *"6 Questions Your Clients Cannot Answer Today"*
- Q1–Q6 listed cleanly, large type, no answers
- Footer: *"We built the answers. Ask us."*

**Back — The Intelligence Layer**
- Section 1: What it is (3 sentences from Slide 10)
- Section 2: The 6 features in 2 sentences each
- Section 3: Partnership models (4 options, one line each)
- Footer: David Kiania · Fireside Communications · contact details

---

## 7. Preparation Checklist Before the Meeting

- [ ] Fill in reseller's actual device ARPU on Slide 18 before sending the deck
- [ ] Confirm their primary client verticals (logistics-heavy? school transport? ISP?) and reorder Q slides to lead with their most relevant pain
- [ ] Prepare one live Grafana dashboard screenshot per feature for the meeting (not the deck)
- [ ] Have a 45-minute working session slot ready to offer at close
- [ ] Know the answer to: *"Which of your clients would you pilot this with first?"* — be ready to suggest one

---

## 8. Optional Add-On Products

### Add-on 1 — Blob Storage (RustFS)

**Pain question (Slide 23):**
> *"Where does your client's dashcam footage go after a road incident — and can they retrieve it in court?"*

**What it stores:**
- Dashcam video from camera-capable devices (JC400P)
- Trip report exports (CSV/PDF)
- Driver documents — licences, PSV badges, insurance certificates
- Vehicle photos — condition reports, incident evidence
- Webhook event payloads for long-term audit

**Pitch angle:** Your client generates video evidence every day. Without a storage layer, that evidence is gone in 72 hours. With RustFS, they have a retrievable, organised archive — per vehicle, per driver, per incident. Self-hosted, S3-compatible, no third-party cloud dependency.

---

### Add-on 2 — Data Warehouse (DuckDB)

**Pain question (Slide 24):**
> *"Can your clients run a 12-month fleet performance report without calling you, hiring an analyst, or waiting overnight?"*

**What it enables:**
- 12–24 months of position, trip, and event history queryable in seconds
- Cross-period reporting — this month vs same month last year
- Export-ready datasets for finance, insurance, and compliance
- Feeds directly from RustFS-stored CSV/Parquet files — the two add-ons compound each other

**Pitch angle:** TimescaleDB handles live operations. DuckDB handles history. Together they mean a client never loses a data point and can answer any question about any vehicle on any day going back two years — with no data team required.

---

### Tier Model (Slide 25)

| Tier | Includes | Target client |
|---|---|---|
| **Intelligence** | 6 core features + Grafana dashboards | Any fleet 5–200 vehicles |
| **Intelligence + Storage** | + RustFS blob (video, docs, exports) | Camera-device fleets, compliance-heavy clients |
| **Intelligence + Warehouse** | + DuckDB historical analytics | Finance, insurance, multi-branch reporting |
| **Full Stack** | All three layers | Enterprise, telco, school networks |

---

## 9. Success Criteria

The pitch succeeds if the prospect:
1. Books the 45-minute working session, OR
2. Names a preferred partnership model, OR
3. Names a client they want to pilot with

Any one of these three is a win. The deck and one-pager are designed to generate at least one of them.