teradata. AI ROI · Sales Play 03

Wave 1 · v1.0 Internal · Naming per Guide D044422 Open the Parable
Sales Play 03 SP-03 · AI ROI

Efficiency that compounds.

The CFO-facing play. For AI leaders where the invoice is now measurable and the board is asking pointed questions. Framed by The Cost of More — a factory parable built on published reporting: the flagship that had everything except the price of one finished unit, and the rival that prices every layer of a valid cell and earned the margin to scale. Backed by Nucleus Research: 427% ROI on a marginal-cost basis, sourced.

Buyer · CFO · VP Infra · FinOps Every number · Nucleus-sourced Wave 1 · Get in the room
Scroll
The play at a glance

A marginal-cost argument — every number sourced, every answer architectural.

AI ROI is not a POC-uplift question. It is a marginal-cost question. Every enterprise AI stack spends across six cost axes, and most of that spend is unattributed. This play turns that structural fact into a diagnostic — and hands the CFO office a defensible written ROI baseline in two hours, with every number traceable to Nucleus Research or a public reference.

🎯 Anchor asset

The Cost of More parable

Interactive · in this pack. ~5-minute interactive read. A production line you can walk station by station, the cost anatomy of one valid cell across seven layers, and a reveal that maps the story to the model. EN + DE.

  • Companion Talk Track Guide (4 sections · ~12 min base script)
  • Companion Executor (3 CFO-adjacent personas)
  • Companion Quick Start (3-min CFO briefer)
🧭 Funnel filter

Where the invoice already exists

This play works where the AI invoice is large enough to be visible and the CFO office is actively looking at it.

  • Named AI capex line item on the P&L ✓✓
  • FinOps team in place (any maturity)
  • Board-level question about AI ROIC
  • 3+ named AI workloads in production or late POC
👤 Primary buyer

CFO · VP Infra · FinOps

Financial or infrastructure owner. Reads the invoice line. Owns the KPI. Answers to the board about marginal cost per useful outcome.

  • Secondary: Head of Data Platform · Deal Desk
  • Anti-signal: pure R&D shop with no P&L accountability
  • Anti-signal: AI budget not yet its own line
427%
Nucleus ROI
Marginal-cost basis
6+1
Axes + compounding
Structural framework
4/4
Wave 1 shipped
Parable · Guide · Quick Start · Executor
EN·DE
Languages live
Parable · Quick Start · Executor
The wave train

Assets ship in the order the seller experiences the deal.

Same wave logic as the Sovereignty play. Sales-cycle-moment order, not dev-effort order. Wave 1 (Get in the room) is where we are — four artifacts shipped, two DE translations in flight. The model diagnostic (Wave 2) is where the biggest build sits.

Wave 1

Get in the room

CFO outreach · positioning · first-meeting framing
Without W1, no CFO / VP Infra / FinOps conversation starts. Four artifacts shipped in the ROI green accent, two DE translations in flight.
The Cost of More — parable EN/DE (illustrated scene · six reveal pairs)
Talk Track EN (5-section CFO presenter script)
Quick Start 1-pager (A4 landscape · CFO-forwardable)
ROI Executor (CFO · VP Infra · FinOps)
🟡Talk Track DE (translation from finalized EN)
🟡Executor DE prompts (variant)
Wave 2

Frame & diagnose

Interest → scored ROI baseline
The interactive AI Cost Model — the play's model, same UX pattern as the AI Maturity Model. Built and bundled in this pack — in review: the content still needs domain-expert sign-off before customer use. Six cost axes, per-line-item spend, the phase ladder and a two-layer recoverable-spend figure. The Nucleus reference-mapping is still open.
Interactive six-axis ROI diagnostic
Baseline capture (per-workload)
Auto-generated compounding scenario
🔴Nucleus reference-mapping engine
Wave 3

Design & scope

Show the customer what "Optimized" looks like
Target-state architecture across Teradata Cloud + Teradata Factory + Teradata Fabric. Consumption-model designer. Business Value Calculator bridge.
🔴Target-State Architecture designer
🔴Consumption-model designer (Active + Elastic Compute)
🔴Business Value Calculator bridge (teradata.com integration)
🟡Industry payloads (FS · Telco · Retail · Manufacturing · Insurance)
Wave 4

Justify & defend

Survive procurement · deal desk · competitive TCO
Competitive TCO, discovery Qs, objections, ROI-claim disclaimers, coaching frameworks.
🟡Competitive TCO — Snowflake · Databricks · Pinecone · hyperscaler-native
🔴Discovery question set (cost-signal listening)
🔴Objection handling library (top-10)
🔴ROI-claim disclaimers · Nucleus sourcing rules
🔴Manager coaching framework (P&L-signal red flags)
Wave 5

Close, expand, sustain

Scale · refresh · executor · routing
Deal-desk packaging, metrics telemetry, refresh cadence, FDE feedback loop, partner co-sell wiring, Executor persona expansion.
🔴Deal-desk packaging (platform + FinOps engagement)
🔴Metrics & telemetry v1
🔴Refresh cadence (Nucleus updates · KPC-style change log)
🔴FDE feedback loop → PM · FinOps signal
🔴Partner co-sell (FinOps consultancies · CFO-office advisory)
🔴Executor extension: +6 personas (Head of Data · CPO · Deal Desk)
🔴Cross-play routing (ROI → Strategy · ROI → Sovereignty)
🔴FR / ES localization
Wave 1 · Shipped

Six artifacts. Field-usable today.

Six self-contained HTML files — no build step, no dependencies, each opens from a local folder. The parable opens the conversation and the model runs the diagnostic; the four sales assets around them carry the choreography. House-of-brand guardrails applied; naming compliant throughout.

Live
📖
Parable · The anchor
The Cost of More
The anchor story. A production line you can walk station by station, the cost anatomy of one valid cell, and the instrument that ends the guessing. EN + DE. ~5-min read.
why-the-cost-of-more.html
Live
📊
Model · Interactive diagnostic
AI Cost Model
Six cost axes under one denominator: cost per unit of finished work. Your baseline, the phase ladder, self-assessment and a two-layer recoverable-spend figure — every figure from your own invoice.
dist/AI_Cost_Model.html
Live
🎙️
Parable Talk Track · Beat-by-beat script
The Cost of More — beat by beat
The script for the parable itself: one section per beat, verbatim opening lines, the lines to land in bold. EN.
why-the-cost-of-more-talktrack.html
Live
🎤
Talk Track Guide · Sales choreography
Sales choreography
Companion to the parable talk track, not a replacement: persona-to-beat pairing, play additions between beats, the turn to the model, objection handling, workshop hand-off.
talktrack-guide.html
Live
📄
Quick Start · Board-forwardable
Board-forwardable briefer
A4 landscape single-page briefer. Print-ready. 3-minute skim. Six cost axes, the compounding curve, sourced proof points, 2h workshop CTA.
quickstart.html
Live
✉️
Executor · Persona outreach
CFO-tuned outreach
Two-stage AI-assisted workflow. Stage 1 · persona essay generator. Stage 2 · outreach with hook + marginal-cost bridge + concrete ask. 3 personas: CFO · VP Infra · FinOps Lead.
executor.html
The framework

Six cost axes. One denominator. Every axis attributed compounds — every month.

Every enterprise AI stack has the same six cost axes. Each maps to a persona reading a different invoice line. Every axis you attribute is one you can act on — every month, for the life of the workload. The Teradata answer works at every axis, under one denominator, with unit-based pricing you can defend to your board.

Unit-cost invoice · six cost axes + compounding curve
Your enterprise AI is spending unattributed across six cost axes. Every one is attributable.

This isn't a marketing framework — it's a diagnostic. The parable is a story about two factories building the same cell — the flagship that could never say what one finished unit cost, and the winner that priced every layer of it, week by week, and earned the margin to scale. The AI Cost Model then does the pricing: cost per unit of work today, at the autonomous enterprise, and which axis decides whether the climb survives — every figure computed from the customer’s own invoice, not from ours.

A1
Inference
Per-call inference charges
×1.00
A2
Re-tokenization & sync
Vector store <→> source of truth
×0.60
A3
Movement
Data leaving your platform to be processed
×0.50
A4
Idle capacity
Reserved capacity you don't use
×0.10
A5
Contention & retries
Workloads fighting for capacity
×0.80
A6
Version drift
Model version changes · quietly
effort
And the four moves that change them
re-time
when it runs
re-place
where it runs
re-choose
what runs it
re-use
whether it runs at all

The judgement vocabulary the parable ends on and the model carries: when, where, what, whether — the four questions about a unit of work. Transparency ends the guessing; these four are what actually move the number, and none of them takes work away.

📊 Approved proof points — every number sourced

These are the only ROI numbers to use in-field. Every one is analyst-grounded (Nucleus) or a public benchmark. The Talk Track and Executor enforce this discipline in-prompt.

427%
ROI — marginal-cost basis, cross-customer aggregate
Nucleus Research
14%
Year-3 cost reduction under consumption pricing
Nucleus Research
$0.0009
Cost per query vs $0.0686 Snowflake
Teradata public benchmark
62×
Query volume — 197,366 vs 3,144 (Snowflake)
Teradata public benchmark
The 14-slot asset spine

Every sales-play slot with a wave assignment and a status.

Same 14-slot spine as Sovereignty. Extended from the mind-map's original 7 slots (A–G) to 14 (H–N) after the completeness pass. Every ROI number in-field must trace to slot K (Legal / ROI-claim disclaimers).

SlotContentStatusWave
A · High-level graphicillustrated production line + six reveal pairswhy-the-cost-of-more.html · reusable in decks✅ LiveW1
B · Differentiation"One platform, six answers · one denominator · unit-based pricing"Talk Track Guide · the Teradata answer✅ LiveW1
C · Proof pointNucleus Research 427% ROI over 3 years, 11-month payback (July 2025, NASBA methodology) — reported in the model’s final tab only✅ LiveW2
D · Concrete next steps2h AI ROI Baseline Workshop → per-workload baseline → compounding scenario🔴 BuildW3
E · Mapping into AI Servicessix cost axes ↔ Teradata Cloud · Factory · Fabric · AI Studio · Active/Elastic Compute · in-DBCapability-Gap-Analysis (293-node model)✅ LiveW3
F · Industry-specific touchFS · Telco · Retail · Manufacturing · Insurance · Auto🟡 AdaptW3
G · Feedback loop into R&DROI baseline deltas → PM · FinOps signal — channel undefined🔴 BuildW5
H · Objection handling libraryTop-10 (POC-first · already on hyperscaler · consumption fear · migration cost · Nucleus challenge)🔴 BuildW4
I · Competitive positioningvs Snowflake · Databricks · Pinecone · hyperscaler-native TCOKPC 2026 · Cost Calculator · Vector Store BoF🟡 AdaptW4
J · Discovery question set15–20 Qs — cost-signal listening (invoice line ownership · marginal-cost question · POC vs prod)🔴 BuildW4
K · Legal / ROI-claim disclaimersEvery ROI claim sourceable · Nucleus attribution rules · anti-fabrication rules🔴 BuildW4
L · Localization planEN + DE at W1–W2 · FR + ES at W5🔴 BuildW5
M · Metrics & telemetryLanding-page visits · workshop bookings · CFO meetings sourced · Business Value Calculator handoffs🔴 BuildW5
N · Lifecycle & refresh cadenceNucleus refresh annually · comp row monthly · cost/query numbers quarterly🔴 BuildW5
Naming (Guide D044422) · Use throughout: Teradata Autonomous Knowledge Platform · Teradata Cloud · Teradata Factory · Teradata Fabric · Teradata AI Studio · Teradata Database · Tera · Context Engine · Active Compute / Elastic Compute · Unified Data Protection · Connected Data Foundation. Never use: AK Cloud · AI Factory · Artemis · AI Workbench · ClearScape Analytics (as product name) · QueryGrid · Enterprise Cluster.
Canonical Seismic mapping

The play in Revenue Enablement's native format.

How AI ROI maps to Teradata's canonical sales-play template (Quick Start → Phase 1 What to Know → Phase 2 What to Say → Phase 3 What to Do → Phase 4 Manager Enablement). This is how Dan Carr's team will accept it into Seismic.

QUICK

Quick Start

ICP: CFO / VP Infra / FinOps with visible AI invoice and board-level ROIC question.

✅ Live
P1

What to Know

six cost axes + the model’s computed outputs. Vendor evidence appears in the final tab only.

✅ Live🟡 Assemble
P2

What to Say

The Cost of More parable + Talk Track (EN + DE) + CFO outreach templates.

✅ Live🟡 DE
P3

What to Do

ROI diagnostic (model) → baseline capture → workload replay → BVC handoff.

🔴 W2–3
P4

Manager Enablement

Co-sell with CFO office + FinOps + coaching frameworks for cost-conversion.

🔴 W4–5
Wave-ordered backlog

Ordered by when in the deal cycle — not by dev effort.

Every item unlocks a moment the seller must win. Waves 1 items are shipping this quarter. The model is built and bundled; Waves 2–3 deepen it and add the BVC bridge. Waves 4–5 are the defense and sustain infrastructure.

Wave 1

Get in the room · 4 shipped · 2 in flight

  • The Cost of More parable EN/DE — illustrated production line, the instrument on-stage, six reveal pairs, no asserted figures.Live
  • Talk Track EN — 5-section presenter script with ROI-claim discipline block.Live
  • Quick Start 1-pager — A4 landscape · 4 Nucleus proof cells · CFO-forwardable.Live
  • ROI Executor (executor.html) — CFO · VP Infra · FinOps personas.Live
  • 🟡 Talk Track DE — translation from finalized EN.~2d
  • 🟡 Executor DE prompts — parallel variant.~2d
Wave 2

Frame & diagnose · model deepening

  • 🔴 ROI diagnostic (the model) — six-axis framework + baseline capture + compounding scenario.~2 dev-weeks
  • 🔴 Nucleus reference-mapping engine — customer inputs map to closest Nucleus reference case.~1 week
Wave 3

Design & scope · workshop-ready

  • 🔴 Target-State Architecture designer — Teradata Cloud + Factory + Fabric consumption model.~1 week
  • 🔴 Consumption-model designer — Active Compute + Elastic Compute scenario.~1 week
  • 🔴 Business Value Calculator bridge — teradata.com/business-value-calculator integration.~3d
  • 🔴 AI ROI Baseline Workshop template — 30m baseline + 60m mapping + 30m compounding.~1 week
  • 🟡 Industry payloads — FS · Telco · Retail · Manufacturing · Insurance · Auto.~1w / industry
Wave 4

Justify & defend

  • 🟡 Competitive TCO row — vs Snowflake · Databricks · Pinecone · hyperscaler-native.~1 week
  • 🔴 Discovery question set — cost-signal listening (15–20 Qs).~3d
  • 🔴 Objection handling library — top-10 (POC-first · Nucleus challenge · migration cost).~4d
  • 🔴 Legal / ROI-claim disclaimers — Nucleus sourcing rules, anti-fabrication guardrails.~3d
  • 🔴 Manager coaching framework — CFO / VP Infra conversation red flags.~1 week
Wave 5

Close, expand, sustain

  • 🔴 Deal-desk packaging — ROI (platform + FinOps engagement) commercial shape.~1 week
  • 🔴 Metrics & telemetry v1 — workshop bookings · BVC handoffs · CFO-meetings sourced.~1 week
  • 🔴 Refresh cadence — Nucleus annual · comp row monthly · cost/query numbers quarterly.~3d
  • 🔴 FDE feedback loop — ROI baselines → PM · FinOps signal channel.~1 week
  • 🔴 Partner co-sell wiring — FinOps consultancies · CFO-office advisory.~2 weeks
  • 🔴 Executor +6 personas — Head of Data Platform · CPO · Deal Desk Lead (extends to 9 total).~1 week
  • 🔴 Cross-play routing rules — ROI → Strategy · ROI → Sovereignty.~3d
  • 🔴 FR / ES localization.~1w / lang
Ready to circulate

Six artifacts. One coherent Wave 1 pack. Ready for LT and Sales to react to.

Open any of the four sibling assets to see the play in the seller's hands. Or forward this landing page to your reviewers as the strawman — every ROI number sourced, every axis architecturally addressable.