Numbers Cheat Sheet — Memorize These
Numbers Cheat Sheet — Memorize These
Purpose: Freddy must know these cold. No notes. No hesitation. Rule: ONLY data from our interviews + Chain IQ dataset. No external research numbers.
People — Who Said What
| Person | Role | Source |
|---|---|---|
| Stijn de Leuw | Project Manager | 40-min interview → people-transcript/stijn-de-leuw-pm/ |
| Tim Paridaens | CTO | Fireside chat → people-transcript/tim-paridaens-cto/ |
| Daniel Ringsma | Head of AI Product & Platform | Workshop → people-transcript/daniel-ringsma-head-of-ai/ |
Chain IQ Numbers (all from interviews)
| Number | What | Who Said It |
|---|---|---|
| 700+ | Chain IQ employees | Daniel Ringsma (workshop) |
| 60+ | Clients (UBS, AXA, KPMG, FedEx, IBM…) | Daniel Ringsma (workshop) |
| 49 | Countries served | Daniel Ringsma (workshop) |
| 70B USD | Total addressable spend | Daniel Ringsma (workshop) |
| ~800 | Requests per month (range 600-1,000) | Stijn de Leuw (PM) + employee follow-up |
| 640 | Automatable requests/month (80% of 800) | Calculated from Stijn’s 80/20 |
| 160 | Strategic requests/month (20% of 800) | Calculated from Stijn’s 80/20 |
| 80% | Volume = tail spend (automatable) | Stijn de Leuw: “Those 80% small tickets — could be automated” |
| 20% | Volume = strategic (needs humans) | Stijn de Leuw: “The other 20% has to be escalated” |
| 24h | SLA for request pickup | Stijn de Leuw: “Within 24 hours we make sure it’s picked up” |
| 30% | Requests misclassified by clients | Stijn de Leuw: “70% classified correctly. 30% still goes to review” |
| 8% | Annual savings target promised to clients | Tim Paridaens (CTO): “We promise 8% a year” |
| 40% | Cumulative savings over contract duration | Tim Paridaens (CTO): “40% cumulative over 5-year contracts” |
| 5 years | Typical contract duration | Tim Paridaens (CTO, fireside chat) |
| 25 | Spot agents Chain IQ has today | Tim Paridaens (CTO, fireside chat) |
| 50 | Companies Chain IQ evaluated for their AI solution | Tim Paridaens (CTO, fireside chat) |
Time Benchmarks (from Stijn de Leuw, PM)
| Tier | Manual Time | With Sift | Time Saved |
|---|---|---|---|
| Marketplace (tail spend) | Same day | Seconds | ~1 day → seconds |
| Technical | ~20 working days | ~90 seconds | 20 days → 90 seconds |
| Strategic | Up to 60 working days | Minutes + human review | 60 days → minutes + review |
Calculated impact on 800 monthly requests:
- 640 marketplace/technical requests × 20 days saved each = 12,800 working days saved per month
- 160 strategic requests significantly accelerated (60 days → minutes of agent prep + human review)
Dataset Numbers (Chain IQ provided data)
| Metric | Value |
|---|---|
| Requests in dataset | 304 (cherry-picked from ~1 year) |
| Suppliers | 40 |
| Pricing tiers | 599 |
| Policy rules | 145 (16 approval thresholds + 61 preferred + 5 restricted + 10 category + 8 geography + 8 escalation + others) |
| Historical awards | 590 |
| Languages | 6 (EN, DE, FR, ES, PT, JA) |
| Categories | 30 (4 L1: IT, Facilities, Professional Services, Marketing) |
| Countries | 19 |
| Currencies | 3 (EUR, CHF, USD) |
Demo-Specific Numbers (REQ-000004 — the edge case)
| Detail | Value |
|---|---|
| Request | 240 docking stations, DE |
| Budget stated | EUR 25,199.55 |
| Cheapest unit price | EUR 148.80 (Bechtle) |
| Minimum total | EUR 35,712 |
| Shortfall | EUR 10,512.45 |
| Days to deadline | 6 |
| Fastest supplier (expedited) | 17 days |
| Policy triggered | AT-002 (2 quotes required >EUR 25K) |
| Escalations | 3 (Requester, Procurement Manager, Head of Category) |
Revenue Model (from Stijn de Leuw + Tim Paridaens)
- Chain IQ charges based on savings they generate (Stijn)
- “Our revenue is decided by how much we save our clients” (Stijn)
- “If our client goes with the most expensive one, that’s the moment we start crying — that’s where we lose money” (Stijn)
- “We promise 8% a year. 40% cumulative over 5-year contracts” (Tim, CTO)
- “Savings methodology can destroy your entire business” (Tim, CTO)
- Every documented saving = Chain IQ revenue
- Every override forced to document the deviation reason
- = audit protection + revenue protection in one feature
Quotes to Use in Pitch
From Stijn de Leuw (PM) — use in opening + demo narration
| Quote | When to Use |
|---|---|
| ”Traceability is one of the most important things. That’s where things get lost.” | OPENING |
| ”Somebody leaves, the knowledge walks out the door.” | CLOSE |
| ”If it’s not documented, that’s where audits tend to fail.” | During edge case demo |
| ”Our revenue is decided by how much we save our clients.” | During savings/override demo |
| ”Those 80% that are the small tickets — that’s the one that could be automated.” | During Tier 1 demo |
From Tim Paridaens (CTO) + Daniel Ringsma (Head of AI) — use in close + Q&A
| Quote | Who | When to Use |
|---|---|---|
| ”The superpower is not the model — it’s structured data grounding.” | Daniel Ringsma | CLOSE |
| ”If they don’t speak about a semantic graph… then I know they haven’t understood.” | Tim Paridaens | Q&A — architecture |
| ”Point my attention to what has changed.” | Tim Paridaens | Q&A — HITL pattern |
| ”Savings methodology can destroy your entire business.” | Tim Paridaens | Q&A — savings |
| ”The house doesn’t burn down because it’s reckless. The house burns down because it’s obedient.” | Daniel Ringsma | Q&A — guardrails |
Validations (confirmed by CTO + Head of AI)
- Structured rules as JSON = correct for the case
- 7-step audit trail with policy citations = exactly what compliance needs
- 3-track autonomy system (marketplace/technical/strategic) = correct pattern
- Mono-agent with file system > multi-agent orchestration
- They SAW our demo live and engaged with it
- Procurement = AI Act high-risk → audit trail is compliance requirement