Assortment and space optimisation
Decides what each location should carry using local demand, display space, substitution and online behaviour — not a national planogram.
- Moves
- Sales per square metre
- Twin
- Merchandising Twin
- Capability
- Optimisation
Industries
04 — Retail & Consumer
Every unit priced, placed and proven.
Demand, price, availability and fulfilment as one engine — agents that move margin daily, inside brand, pricing-law and consumer-protection guardrails.
North Star
Gross margin return on inventory investment
The value engine, end to end
Discover → Model → Twin → Agents → Govern → Compound
The same six moves we run everywhere, expressed in the physics of this industry.
01
Discover
Where does margin actually leak?
Margin waterfall by category
02
Model
What drives demand and availability?
Elasticity and availability model
03
Twin
What if we change price, buy or space?
Category and network twin
04
Agents
Who executes daily?
Pricing, replenishment, content agents
05
Govern
Is it fair, legal and on-brand?
Pricing and claims control file
06
Compound
Does GMROII improve?
Outcome contract
Where the value is
Value pools and their drivers
Before a twin is built, we agree where the money sits and what moves it.
Value pool
Price & promotion
Driver
Elasticity, promo depth, markdown timing
Measured as
Margin points per category
Value pool
Availability
Driver
Forecast error, replenishment latency
Measured as
Lost sales recovered
Value pool
Supply cost
Driver
Freight, handling, returns
Measured as
Cost per unit delivered
Value pool
Customer value
Driver
Repeat rate, basket mix
Measured as
Contribution per active customer
Use cases
Where value is already sitting in Retail & Consumer
Each one is a number somebody already owns, a twin that models it, and an agent that moves it inside a defined authority. Filter by function.
7 use cases
Price and markdown optimisation
Sets price and markdown timing on measured elasticity and seasonality, so margin is protected instead of discounted away at the end of the season.
- Moves
- Gross margin per unit sold
- Twin
- Pricing Twin
- Capability
- Elasticity modelling
Out-of-stock and availability risk
Predicts the stockout before it shows on the shelf and triggers the replenishment or substitution that saves the sale.
- Moves
- Lost sales from availability
- Twin
- Supply Chain Twin
- Capability
- Forecasting
Promotion and trade budget allocation
Measures what each promotional channel actually caused at regional and national level, and moves the next dollar to the one that paid.
- Moves
- Incremental margin per promo dollar
- Twin
- Marketing Twin
- Capability
- Causal measurement
Returns and reverse-logistics reduction
Finds the products, sizes and journeys that generate returns and fixes them upstream, where the cost has not yet been incurred.
- Moves
- Return rate and cost per return
- Twin
- Process Twin
- Capability
- Root-cause analytics
Shrink and transaction anomaly detection
Separates operational error from deliberate loss at till, stock and supplier level, and quantifies both.
- Moves
- Shrink as percent of sales
- Twin
- Financial Crime Twin
- Capability
- Anomaly detection
Product content and catalogue enrichment
Generates and standardises product attributes, descriptions and imagery metadata so search and recommendation have something to work with.
- Moves
- On-site conversion rate
- Twin
- Merchandising Twin
- Capability
- Multimodal generation
Outcome twins
Twins, levers, features and the agents that run them
Each twin owns a set of levers. Each feature moves one lever. Each agent executes inside a defined authority.
Category Twin
Simulate range, price and space together instead of in three meetings.
Levers
Price elasticityPromo ROIMarkdown timingRange productivity
Features
- — Price simulation
- — Promo evaluation
- — Assortment optimisation
Multi-agent layer
Pricing Agent
Proposes daily price moves within guardrails.
Availability Twin
Put the right unit in the right node before the customer looks.
Levers
Forecast accuracyOn-shelf availabilitySafety stock
Features
- — Demand forecasting
- — Replenishment
- — Store allocation
Multi-agent layer
Replenishment Agent
Rebalances stock across nodes overnight.
Fulfilment Twin
Choose the cheapest node that still keeps the promise.
Levers
Cost to serveSplit shipmentsReturn rate
Features
- — Order sourcing
- — Returns prediction
- — Carrier selection
Multi-agent layer
Sourcing Agent
Routes each order to the lowest true-cost node.
Customer Twin
Grow contribution per customer, not raw traffic.
Levers
Repeat rateBasket mixDiscount dependency
Features
- — Offer personalisation
- — Churn prediction
- — Content generation
Multi-agent layer
Offer Agent
Selects the cheapest offer that still converts.
Unit economics
ROI calculator — per feature, per unit
Set your volume, switch features on or off, and move each impact to your own evidence. Everything is expressed per order.
Orders per year
8,000,000
Contribution per order
$11
Per-order contribution; swap in per-basket or per-SKU economics as required.
Annual program investment
$3.5m
Category Twin
Unit revenue — Revenue per order
2.0% — Elasticity-aware moves instead of blanket discounts.
$9.9m
annual value
Availability Twin
Volume — Orders captured
3.0% — Availability recovered on high-velocity lines.
$2.6m
annual value
Fulfilment Twin
Unit cost — Cost per order
4.0% — Fewer splits, cheaper nodes, same promise.
$16.3m
annual value
Category Twin
Leakage — Markdown leakage
20.0% — Cuts earlier and shallower where the curve says so.
$11.9m
annual value
Fulfilment Twin
Leakage — Returns leakage
10.0% — Prevents the orders that were always going to come back.
$6.0m
annual value
Value by feature
Annual value contributed by each active feature at current settings.
Value and ROI vs volume
How engineered value and ROI move as orders per year change. The marker is your current setting.
Cumulative value vs investment
Where the curve crosses the investment line is payback.
Value mix by lever
Which lever the engineered value is actually coming from.
Baseline contribution
$88.0m
Engineered annual value
$46.7m
ROI on program
1235%
Payback
0.9 months
Directional model. In an engagement every number here is replaced by your measured baseline, signed off by finance, and written into the outcome contract before any agent goes live.
Scale
How this compounds from one unit to the enterprise
01
One category
Prove elasticity and margin lift against a holdout.
02
One banner
Roll the same twin across categories with local guardrails.
03
Group
Shared runtime across markets, one margin definition.
Integrations
Domain platforms we connect
Vendor-agnostic by design. Connectors are added per engagement — no platform lock-in.
Commerce
- Shopify
- SAP Commerce
- Salesforce Commerce Cloud
- commercetools
Merch & supply
- Blue Yonder
- RELEX
- Oracle Retail
- SAP IBP
Data & CDP
- Snowflake
- Databricks
- Segment
- Bloomreach
Ops
- Manhattan
- NetSuite
- Zendesk
- Dynamics 365
Regulations and standards
What we are held to in this industry
Compliance is a design input, not a review gate. Every agent action is logged, attributable and reversible.
GDPR / ePrivacy
Consent-based personalisation and profiling limits.
EU AI Act
Transparency for AI-generated content and recommender behaviour.
Omnibus / price-indication
Prior-price rules on any automated markdown.
Consumer protection law
No dark patterns in agent-generated offers.
PCI DSS
Payment data boundaries in agent tooling.
Standards supported
GS1EDI / EDIFACTISO 27001SOC 2 Type IINIST AI RMFWCAG 2.2
Next industry
Energy & Utilities