Stable, predictable demand
Consumption varies within a narrow band across the history window.
Optimize automatically
Calculate, rank, approve in one step and write back.
Inventory & Working Capital Optimization
Rules nobody has reviewed in years are still ordering your stock.
A governed workflow that turns consumption history into scored, ranked replenishment parameter recommendations — approved by the planner in one step, then written back into SAP automatically, without rebuilding a single planning process.
What the workflow runs on
Feeds you already have
ConsumptionMRP parametersUnit costSupplier termsSite master
Governed decision logic
Material-site grainStability classesFour priority tiersConfidence
What reaches the team
Ranked SKUsCapital freedApproval queue
The challenge
Parameters set at go-live, never revisited
Replenishment parameters set when the system was implemented and never revisited quietly tie up working capital across every site in the network. Assortments changed, supplier lead times moved and demand patterns evolved — while the underlying lot sizes and safety stock levels stayed exactly where they were. The gap between what the system says to order and what demand actually requires widens every cycle.
Static rules, months of coverage
Fixed parameters order far more demand coverage per cycle than current consumption justifies, across a significant share of the portfolio.
No systematic view of the excess
Planning teams have no data-driven read on which materials are overstocked, by how much, or at which sites.
Supplier minimums taken at face value
Minimum order quantities are accepted without the consumption evidence that would support renegotiating them.
Corrections are ad hoc
Parameters get fixed one at a time, by whoever noticed, instead of through repeatable optimization logic and scored priorities.
The result is excess stock in some locations, service risk in others, and an inventory cost structure that resists every margin improvement initiative aimed at it.
The leadership question
One question, asked in every review
“How do we systematically identify and correct the inventory parameters costing us the most — without disrupting daily planning operations or requiring a new system?”
The split
Not every SKU should be automated
Applying one rule to the whole portfolio is what created the problem. Classifying demand stability first is what makes automation safe — the stable majority can be optimized without a human in the loop, and the volatile tail should never be.
Volatile or intermittent demand
Consumption is lumpy, seasonal or sparse enough that a calculated parameter would mislead.
Flag for manual handling
Surface it to a planner rather than automating a bad answer.
How the two axes resolve
Demand stability against capital tied up
Capital impact alone would send planners at the least predictable items. Stability alone would send them at the smallest. The scoring carries both, which is why the queue is worth working top-down.
Demand stability →
Low · StableAutomate
Safe to optimize, small prize. Batch it and move on.
High · StableWork it first
Safe to optimize and the largest prize. Top of the queue.
Low · VolatileLeave alone
Neither predictable nor material. Do not spend planner time.
High · VolatilePlanner review
Material, but too lumpy to automate. Needs judgement, not a formula.
Capital tied up →
What is included
Six capabilities, six deliverables
Each one is a packaged capability — the data model, the scoring logic and the deliverable already designed around the decision it serves. They are built to run together, but each produces an artefact that stands on its own.
01
Unified Demand & Parameter Model
Demand Signal Foundation
A shared dimensional model joining two years of consumption history to current replenishment parameters, unit costs, supplier terms and site master data at material-site grain — one normalized demand signal powering every downstream optimization and confidence calculation from a single governed source.
02
SKU Stability & Eligibility Engine
Demand Stability Classification
Transparent classification of every SKU by demand variability across the history window. Stable materials route to full automated optimization, moderate items go to planner review, and volatile or intermittent SKUs are flagged for manual handling — so the right logic is applied to the right material on every run.
03
Order Quantity & Safety Stock Layer
Lot Size & Safety Stock Optimization
Defined formulas deriving the economically optimal order quantity by balancing actual holding and ordering costs, then calculating statistically defensible safety stock from demand variability and observed lead times — rounded to supplier pack sizes and checked against minimum order quantities before anything is proposed.
04
Capital Impact Scoring & Tier Model
Opportunity Scoring & Prioritization
A composite score weighted by capital freed, months-of-supply excess, demand confidence and lead-time risk. Every eligible SKU is ranked into one of four priority tiers, so planners always work the highest-value opportunities first and leadership sees an attributable pipeline of working capital recovery by site and supplier.
05
Planner Approval & Update Layer
Governed Approval & Write-Back
A structured approval workflow presenting ranked recommendations to the responsible controller for a single sign-off before any parameter changes. Materials constrained by a supplier minimum are separated and routed to procurement with a data-backed negotiation brief, and every posted change creates a full change document for audit.
06
Data Quality & Governance Playbook
Data Confidence & Governance
A data confidence index per SKU reflecting demand history completeness, cost data recency, lead-time source quality and the age of the minimum-order record — tagging every recommendation high, medium or low, alongside model versioning, planner override tracking and a defined recalibration cadence.
The artefacts
What actually arrives, and who it arrives for
Six named deliverables rather than access to a tool. Each is versioned, traceable to the model that produced it, and routed to the team that acts on it.
Deliverable 01 · Demand Signal FoundationUnified Demand & Parameter Model
A shared dimensional model joining two years of consumption history to current replenishment parameters, unit costs, supplier terms and site master data at material-site grain — one normalized demand signal powering every downstream optimization and confidence calculation from a single governed source.
Deliverable 02 · Demand Stability ClassificationSKU Stability & Eligibility Engine
Transparent classification of every SKU by demand variability across the history window.
Deliverable 03 · Lot Size & Safety Stock OptimizationOrder Quantity & Safety Stock Layer
Defined formulas deriving the economically optimal order quantity by balancing actual holding and ordering costs, then calculating statistically defensible safety stock from demand variability and observed lead times — rounded to supplier pack sizes and checked against minimum order quantities before anything is proposed.
Deliverable 04 · Opportunity Scoring & PrioritizationCapital Impact Scoring & Tier Model
A composite score weighted by capital freed, months-of-supply excess, demand confidence and lead-time risk.
Deliverable 05 · Governed Approval & Write-BackPlanner Approval & Update Layer
A structured approval workflow presenting ranked recommendations to the responsible controller for a single sign-off before any parameter changes.
Deliverable 06 · Data Confidence & GovernanceData Quality & Governance Playbook
A data confidence index per SKU reflecting demand history completeness, cost data recency, lead-time source quality and the age of the minimum-order record — tagging every recommendation high, medium or low, alongside model versioning, planner override tracking and a defined recalibration cadence.
Previews are schematic. They carry no figures, because the numbers on them would be ours rather than yours.
The output
Every SKU lands in one of four tiers
Planner capacity is the real constraint, not analytical capability. Tiering is what concentrates limited time on the SKUs where a parameter change actually returns capital.
TIER 1Tier 1
Large capital opportunity on stable demand. Work these first — the queue is sequenced for it.
TIER 2Tier 2
Meaningful opportunity with good confidence. Batch approve once Tier 1 is clear.
TIER 3Tier 3
Real but modest. Worth taking when the run comes round, not worth a special effort.
TIER 4Tier 4
Constrained, volatile or thin on data. Route to review or procurement rather than automate.
The workflow
From consumption history to SAP
Nothing writes back without a human approving it, and every posted change leaves an audit trail. That is the difference between an optimization engine and a script.
1
JoinConsumption history to current parameters, costs, terms and master data
2
ClassifyEvery SKU scored on demand stability and routed accordingly
3
CalculateEconomic order quantity and statistically defensible safety stock
4
ConstrainRounded to pack sizes and checked against supplier minimums
5
ScoreRanked by capital freed, excess coverage, confidence and lead-time risk
6
ApproveOne planner sign-off, then automatic write-back with a change document
Materials constrained by a supplier minimum are separated automatically and routed to procurement with a data-backed negotiation brief instead of being silently accepted.
What changes
Four things that stop being true
The workflow does not add another report. It replaces four sentences that get said in every review with something specific enough to act on.
“That’s just what the system orders.”
Every parameter traces to a calculated recommendation on current consumption, with a confidence rating.
“We’ll clean up the parameters next year.”
The optimization reruns on a cadence, so alignment is continuous rather than a project.
“The supplier minimum is the supplier minimum.”
Constrained materials arrive at procurement with the consumption evidence to renegotiate them.
“We can’t automate parameter changes safely.”
Nothing posts without a planner approving it, and every change leaves a full audit document.
Who it is for
One workflow, four seats
The same governed output, presented for the decision each team actually makes. Nobody is asked to interpret somebody else’s view.
Planners & Controllers
A ranked queue rather than an ad hoc review, with a single approval step and no requirement to learn a new system to use it.
Procurement
Ranked, evidence-based renegotiation briefs that quantify the cost of a current supplier minimum and propose a data-backed target quantity.
Finance
Working capital released from excess cycle stock and oversized buffers, attributable to cost of goods and operating expense with no new infrastructure.
Supply Chain Leadership
A clear, attributable pipeline of working capital recovery by site and by supplier, and parameters that stay current as conditions change.
How it is governed
What makes it repeatable
Anyone can produce this read once. What makes it a capability is that the grain, the logic and the model version are fixed and stored, so the next cycle is measured the same way as the last one.
Grain
Material × site. Every recommendation, score and confidence rating resolves to it.
Eligibility
Demand stability decides what may be automated at all. The volatile tail is excluded by design rather than by exception.
Confidence
Every recommendation is tagged high, medium or low on history completeness, cost recency, lead-time quality and record age.
Auditability
One planner sign-off before anything posts, a full change document after, model versioning and planner override tracking throughout.
Start with one site.
Bring one distribution centre and two years of consumption history. We’ll classify the portfolio, calculate the parameters and show you the ranked capital opportunity it produces — before you commit to anything wider.