# Deck 13 · Competitor Intelligence (Plex-CI) Productised

**Status:** Draft (not yet built)
**Saved:** 2026-06-28 by Jeff (verbal)
**Owner:** Plex / OMX Commercial + Analytics

---

## One-line thesis

**Productise the Plex-CI scraper pipeline — already running, already capturing 545k URLs and 471k change events across 59 competitors — from an internal tool into a strategic differentiator OMX can deploy commercially and (optionally) licence to other distributors.**

## The wedge — why now

- The pipeline EXISTS and WORKS — sitemap channel unblocked 2026-06-27, 545,789 URLs + 471,991 url_changes captured against 2026-04-24 baseline
- Direct Python → Postgres path bypasses the broken SQLite chain
- Plex-CI competitor scan revealed: **none of NXP / DiscountOffice / McGreals / Hurdleys publish custom competitive intelligence** — genuine moat
- The 4 clean suppliers (NXP, DiscountOffice, McGreals, Hurdleys) provide reliable data; the 12 brand-owner duplicates are noise to be filtered, not features
- The IP, infrastructure, and operational tradecraft are already paid for — productisation is incremental

## What this deck covers

1. **What Plex-CI does today** — scrape, normalise, change-detect across 59 competitors; raw_urls JSON kept legacy-compat
2. **What it could become internally** — promo-aware pricing, off-range alternative suggestion, share-of-wallet attack
3. **What it could become externally** — licensed dataset/service for other NZ B2B distributors (Angle C from Deck 07 — platform-licence)
4. **The 4 clean suppliers** — why NXP/DO/McGreals/Hurdleys form the verified spine
5. **Channel build-out roadmap** — sitemap (done) → products → news → social
6. **Economics** — internal pricing-decision uplift, optional licence revenue stream

## What this deck explicitly does NOT do

- Not promo planning (Deck 02 — PPSS owns)
- Not catalog management (Deck 05 — Global Catalog owns SKU data)
- Not product matching at SKU level (Deck 20 — Product Matching owns)

---

## Problem framing (what's broken)

- **Pricing decisions are blind** — without continuous competitor signal, contract-review or pricebook updates lag market by weeks
- **No off-range alternative** — when an OMX SKU is OOS, no automated "competitor still has it" or "we have an equivalent" path
- **Switching evidence is anecdotal** — Stage 3 Switching needs "you paid $X at competitor; OMX has same SKU at $Y"; today's competitor data is one-off scrapes
- **Channels broken** — products, news, social all parked at 2026-06-27 status
- **Plex-CI memory limitations** — `reference_plex_ci_vendor_field_classification` notes brand-owner contamination (acquire/elive/pbtech etc); analysis must filter

## Benefits (the value story)

| Lever | Mechanism | Sizing approach |
|---|---|---|
| **Pricing margin protection** | Live competitor signal → faster pricebook moves | Industry: 0.5-2% margin recapture from continuous CI vs quarterly |
| **Switching evidence** | "Competitor $X vs OMX $Y" data in Stage 3 (Deck 09) onboarding | Direct enabler of switching narrative |
| **Off-range alternative** | Automated equivalent-SKU suggestion when OMX OOS | Reduces lost-sale; integrates with Deck 05 Catalog |
| **Promo intelligence** | Detect competitor promo launches before customers do | Defensive — preserves promo lift in Deck 02 (PPSS) |
| **Licence revenue (optional)** | Other NZ distributors pay for the dataset | Platform-licence model — Angle C from Deck 07 |
| **N3 + EXCO ammo** | Quantified market intelligence for buying-group and exec discussions | Strategic — credibility multiplier |

---

## Layout candidates from the gold standard

- **Problem vector grid (4)**: Blind-pricing / No-OOS-alternative / Anecdotal-switching / Vendor-noise
- **Live stats hero**: 545,789 URLs · 471,991 changes · 4 clean + 12 noise · 59 competitors — running counter
- **The 4 clean spine + 12 brand-owner footnote**: visual hierarchy showing what's signal vs noise
- **Internal use cases**: 4 boxes — Pricing / Switching / Off-Range / Promo
- **External use case (optional)**: Licence-to-distributors visual
- **Channel roadmap**: Sitemap (done) → Products (next) → News → Social — stage-gate timeline
- **Architecture pipeline**: Competitor sites → Python sitemap fetcher → Postgres → Snowflake → Lens dashboards + downstream products
- **The ask**: Build budget to unblock products/news/social channels; optional licence framework

---

## Open questions to resolve

1. **Internal vs external priority** — does OMX want this as internal differentiator only, or both?
2. **Products channel unblock plan** — what's the next blocker after sitemap?
3. **Pricing decision integration** — who consumes the signal? Commercial? Buying? Pricebook?
4. **News + social value** — is the additional channel signal worth the build vs sitemap-only?
5. **Licence model** (if Angle C) — pricing tier, customer profile, support model, exclusion clauses
6. **Legal / ethics** — public scraping defensible (sitemap is public) but need a position
7. **Anti-scrape defence** — if competitors block, can pipeline degrade gracefully?
8. **Connection to Deck 20 (Product Matching)** — Plex-CI feeds the matching engine; do they ship together or separately?

---

## Audience

**Primary:** CFO + Chief Commercial Officer. Margin protection + switching enablement.
**Secondary:** Pricing + Buying teams — daily users.
**Tertiary:** Strategy + Board — optional licence-revenue narrative.

## Reference

- Memory: **Plex-CI scraper pipeline state** (`project_plex_ci_scraper_state`) — 2026-06-27 sitemap unblock; 545k URLs + 471k changes captured
- Memory: **Plex-CI vendor field classification** (`reference_plex_ci_vendor_field_classification`) — 4 clean + 12 brand-owner duplicates
- Memory: **Deck 07 Angle C** — platform-licence marketplace pattern (parallel to this deck's external option)
- Memory: **Deck 02 (PPSS)** — Plex-CI competitor scan revealed none publish custom promo/demand planning; PPSS is differentiator

---

## Research deepening (background-agent, 2026-06-28)

### What Plex-CI is competing against — vendor landscape

| Vendor | Positioning | Pricing (USD/yr) | Notes for OMX |
|---|---|---|---|
| **Prisync** | SMB-friendly ecommerce CI | $99-$399/mo ($1.2k-$4.8k/yr) | Caps at 5,000 products — too thin for OMX 50k+ SKU range |
| **DataWeave** | AI-powered pricing intelligence | $99/mo entry, enterprise custom | Strong on global retail; NZ B2B coverage thin |
| **Competera** | Elasticity-based pricing | $15k-$50k+ entry, enterprise custom | Heavy on price-optimisation modelling vs raw signal |
| **Wiser Solutions** | Online price tracking + MAP monitoring | Custom enterprise | US-centric |
| **Crayon / Klue** | Full competitive intelligence (battlecards) | $12.5k-$60k/yr | Sales-enablement angle, not pricing signal |
| **Intelligence Node** | Enterprise retail CI | $50k+/yr | Heavy enterprise; over-engineered for 4-supplier spine |
| **Plex-CI (OMX)** | NZ B2B distributor-specific, sitemap-first | Already built; marginal cost ~$0 | The moat |

**Source citations:**
- Prisync tiers — https://prisync.com/competitive-intelligence-software/
- DataWeave — https://dataweave.com/us/pricing-intelligence
- Competera — https://pryse.ai/blog/pricing-intelligence-software (covers Competera + Crayon/Klue)
- Market overview — https://www.withorb.com/blog/competitive-pricing-tools

### Key takeaway for the deck
**No off-the-shelf vendor covers the NZ B2B distributor wedge.** All US/global retail-focused, all priced for SKU caps that break at OMX scale, all miss the NZ-specific supplier set (NXP / DiscountOffice / McGreals / Hurdleys). Plex-CI's marginal cost is **near-zero** vs $15k-$60k/yr for the cheapest enterprise option — and licence revenue (Angle C) becomes credible at $5k-$25k/yr/distributor.

### Sizing the moat — quantified

| Metric | Value | Source |
|---|---|---|
| URLs captured | 545,789 | Plex-CI 2026-06-27 run |
| Change events vs 2026-04-24 baseline | 471,991 | Plex-CI |
| Competitors monitored | 59 (37 OK / 22 fail) | Plex-CI |
| Clean signal spine | 4 suppliers | NXP / DiscountOffice / McGreals / Hurdleys |
| Vendor channel still parked | Products / news / social | 2026-06-27 status |

### Pricing-recapture sizing benchmark
Continuous competitive pricing intelligence typically delivers **0.5-2% gross margin recapture** vs quarterly manual pricebook reviews (Bain, McKinsey retail pricing literature). At OMX revenue scale, even **0.25% margin lift on $250M ≈ $625k/yr** — a credible Year-1 internal economic case.

### Licence-revenue (Angle C) shape
If OMX licences the dataset to 5-10 NZ B2B distributors at **$10k-$20k/yr each**, that's **$50k-$200k/yr ARR** — modest in absolute terms but **gross-margin >80%** since the data already exists.

---

## Vectors + visuals

### Lucide icons
- **Cover hero:** i-search (continuous monitoring) + i-zap (live signal)
- **Problem grid:** i-search (blind-pricing) / i-shopping-cart (no OOS alt) / i-shuffle (anecdotal switching) / i-bar-chart (vendor noise)
- **4 clean spine:** i-sparkles (signal) — paired with i-life-buoy small (legacy noise)
- **Internal use cases:** i-bar-chart (Pricing) / i-shuffle (Switching) / i-repeat (Off-Range) / i-zap (Promo)
- **Architecture pipeline:** i-laptop (sites) -> i-brain (Python+Postgres) -> i-file-text (Snowflake) -> i-smartphone (Lens)

### Image concepts (cover + 5 key slides)
1. **Cover hero** — Close-up of a NZ office-products warehouse aisle with subtle blue digital scan-lines overlay. Source: Unsplash NZ warehouse / OMX brand library (DC stock photos). Anchor: "545,789 URLs. 471,991 changes. 4 clean suppliers. 59 competitors. Watching."
2. **The 4 clean spine slide** — Honest composite logo wall: NXP, DiscountOffice, McGreals, Hurdleys above the fold; the 12 brand-owner duplicates greyed/below (use logos pulled from each public site). Source: public brand assets, each at <100px footprint to stay nominative-use safe.
3. **Pricing decision slide** — NZ commercial team at desk reviewing pricebook on dual monitor with live competitor signal on screen-right. Source: OMX internal photography or styled stock (Auckland-CBD office context). Note: avoid US suit/tie aesthetic — NZ business-casual.
4. **Switching evidence slide** — Side-by-side product card: competitor SKU $X / OMX SKU $Y. Source: real public listings (sanitised).
5. **Channel roadmap slide** — Four-stage visual: Sitemap (green check) -> Products (amber building) -> News (grey waiting) -> Social (grey waiting). Pure infographic, no photography.
6. **Licence (Angle C) optional slide** — Map of NZ with dots on 5-10 distributor cities (Auckland, Hamilton, Tauranga, Wellington, Christchurch, Dunedin). Source: Mapbox/Leaflet static export.