Operations cascade · with Deena

Forecasting that stays current as the business moves.

CPG teams forecast in spreadsheets that go stale the moment someone opens them. Modus plans bottom-up from every distribution point — Retailer × DC × Product — and keeps the number live as velocity, promos, and doors change.

app.moduscpg.com/demand-planning Deena monitoring
RETAILER × CDC × PRODUCTW45W46W47W48W49W50
Kroger · Dallas · Vanilla Pint1,8401,8601,9102,1402,1802,210
Kroger · Atlanta · Vanilla Pint1,1201,1351,1501,2901,3101,320
Target · Chino · Choc Bar 12ct9609401,0101,0201,0401,050
Whole Foods · NE · Oat Creamer720735741748755760

W48 · velocity supersession applied — +200 doors, +8,400 cases/wk · past weeks locked

87%
Forecast accuracy
−2.1%
Bias
25+
Live metrics
4
Drift flags this wk

01 · The stale number

The forecast was right the day it was built. Then velocity moved, a promo hit, 200 doors landed — and every downstream decision kept running on the old number.

02 · The living plan

Modus plans bottom-up from every distribution point — Retailer × DC × Product — with velocity, drivers, and promotions layered per week. When anything changes, the whole plan recalculates. Past weeks lock; history stays queryable.

03 · The watch

Deena tracks accuracy automatically — WAPE, bias, tracking signal — flags drift by name, and drafts the baseline adjustment. You approve from the punchlist.

Distribution-point grid

Bottom-up, from every shelf.

The forecast is built where demand actually happens — velocity × store count per retailer, DC, and product. When a door opens or a velocity shifts, the plan follows automatically.

app.moduscpg.com/demand-planning DP grid
Kroger · Dallas · Vanilla Pint2,140 /wk · W48 supersession
Kroger · Atlanta · Vanilla Pint1,290 /wk
Target · Chino · Choc Bar1,020 /wk
WF · NE · Oat Creamer748 /wk
Velocity × store count per Retailer × DC × Product — past weeks lock, history stays queryable

Model Builder · why this changes the industry

The right model for every segment. Proven before it ships.

Every segment of your business behaves differently — so every segment gets the model that fits it. Models apply by criteria: Holt-Winters where there’s deep history, launch curves for new items, promo-aware models where trade is heavy. And every change earns its way live through the promote gate — backtested on your own history, promoted only when it measurably improves accuracy.

app.moduscpg.com/demand-planning/model-builder Composed stack
Holt-Winterswhere: 52+ wks history · stable velocity
Baseline velocitywhere: new items · < 12 wks history
ARIMAwhere: promo-heavy segments
+ seasonality · weather · promoslayered signals
The promote gateevery promotion earns it
Backtestfull production engine · your history
Accuracy checkmust improve WMAPE ✓
Scope checkone model per scope ✓
Promotedversioned · provenance on every number
criteria
gated model usage
backtest
before promote
proven
accuracy gains only
versioned
provenance per number

Forecast Center

Know when you’re wrong — by name.

WAPE, bias, and tracking signal computed continuously, from portfolio down to a single distribution point. Drift gets flagged with its cause, and Deena drafts the correction.

app.moduscpg.com/demand-planning/accuracy WAPE 12.8%
Portfolio WAPE12.8% · improving
Bias−2.1% · within band
Kroger · Vanilla Pintdrift flagged · +12% vs plan
Deena’s draftbaseline +8% from W48
Accuracy tracked automatically — drift flagged by name, fix drafted

New-distribution wizard

200 doors in 60 seconds.

The parking-lot moment lives here: pick the retailer, the products, the DCs, the negotiated velocity — and the win becomes forecast, supply signal, and trade calendar in one pass.

app.moduscpg.com/demand-planning/new-distribution Wizard
RetailerProductsDCsVelocityApply
Kroger SE · +200 storesblown out per DC
Vanilla Pint6.2 /store/wk · negotiated
Forecast impact+8,400 cases/wk from W48
The 60-second cascade — supply and trade see it immediately

Snapshots & signoff

S&OP with receipts.

Lock a point-in-time plan, route it for multi-stakeholder signoff, and compare any month against any other — with every variance attributed to its driver.

app.moduscpg.com/demand-planning/snapshots S&OP
June S&OP snapshotlocked · signed by 4
July working plan+3.2% vs June snapshot
Variance driverKroger expansion · promo shift
Point-in-time locks — compare any two plans, every change attributed

A forecast that’s still right on Thursday.

Featured · Model Builder

Compose the forecast itself: a statistical base — Holt-Winters, ARIMA, exponential smoothing — with signals stacked on top: seasonality, weather, promos, overrides. Backtest against your history, promote through the gate, provenance on every number.

87%
forecast accuracy
−2.1%
bias
25+
derived metrics
60s
cascade to supply

And everything around it

Distribution-point grid — velocity × stores, effective-dated
New-distribution wizard — the 60-second cascade
Scenario builder — model the deal before it closes
Variance analysis — portfolio down to DP
Snapshots & S&OP signoff
Forecast Center — accuracy, stock requirements, review
Velocity suggestions from scan + shipment data
External signals — weather, events, market data

In the cascade: Demand Planning Order Management Supply Planning & MRP

See it on your data.

Book a Demo