Today Actual Forecast Last year Spring Holiday
CybexAI MERCHANDISING
Merchandising and replenishment
for specialty retail
Cybex Retail Data Platform · AI Insights
Buying from last year's spreadsheet.
Measured on live data at a five-store specialty apparel retailer, twelve weeks into the fall season.
+12%
stock against last year, while units sold fell 7%
18.6
weeks of supply, against 14.2 at the same point last year
195%
of their demand already held by the style-colours that sell
27%
of letter-size stock in XL, a size that sells 23% of units
Too much of what doesn't sell, too little of what does. Static reorder points and spreadsheets can't keep up.
From data to decisions, automatically.
One platform carries every number from the till to the purchase order.
1
Data platform

POS, ERP, e-commerce and purchasing in one model

2
BI analytics

Sales, stock, receipts and open orders by week

3
Statistical extract

Selling rate, season index, store and size indexes

4
Insight datasets

Season, size, buy, markdown and assortment

5
AI Insights viewer

Findings, exceptions and actions in plain language

6
Applications

Replenishment
Suggested POs
Open-to-buy
Size, markdown, assortment

Starting with Merchandising & Replenishment
Every sale feeds tomorrow's forecast
Fully automated: re-forecast every night, buyers approve the exceptions
01 · Forecasting
A forecast that knows the season.
Each style's own recent selling rate, shaped by a weekly season index. Built for every SKU at every store.
Weekly season index
52 retail weeks · 1.00 = an average week
1.00 Week 47 · holiday · 1.80
Base rate · last 6 weeks
12/ wk
×
Season index · wk 47
1.80
=
Forecast · wk 47
21.6/ wk
02 · Store and size indexes
Every store.
Every size.
A store index splits the chain forecast by store. A size index splits it by size, so size runs stay complete where they sell.
Store index
Share of the chain's demand
Size curve
Share of units by size
Store AStore B
03 · Suggested purchase orders
Purchase orders that write themselves.
Forecast need over the vendor's lead time plus safety stock, less stock on hand and on order. The buyer approves.
Suggested PO · Vendor 0412Lead time 6 wks + 2 wks safety
Built overnight
Suggest = forecast × 8 weeks − on hand − on order
Style · Colour · SizeFcst / wkOn handOn orderSuggest
Total 0 unitsWithin open-to-buy
Approve PO
04 · Open-to-buy
Open-to-buy that stays honest.
Plan, actuals and open orders in one view, refreshed every night. It shows which months to hold and where to chase.
Open-to-buy = plan end stock − (opening stock − sales − markdowns + on order)
05 · AI Insights
Insights that explain themselves.
Season, size, buy, markdown and assortment, written up in plain language every morning.
Cybex AI Insights · Merchandising
Insights
Season & open-to-buy
Size analysis
In-season buy
Stock rebalancing
Markdown strategy
Spring assortment

Season & open-to-buy · fall weeks 1 to 12

All stores · refreshed 06:00 · stock by department, last year = 100
Sales vs last year
-7.2%
Stock vs last year
+12%
Weeks of supply
18.6
AI insight
Automation and Optimization.
Every night it re-forecasts, re-indexes and rebuilds the orders. Your buyers start the day on the exceptions.
↑
In-stock

Best sellers kept in stock by store and size

↑
Turns & GMROI

Money in the stock that actually sells

↓
Markdowns

Less of the wrong size in the wrong store

→
Buyer time

From spreadsheets to exceptions and strategy

CybexAI MERCHANDISING
Buy what sells. Stock where it sells. Automatically.
cybexai.ca
cybexai.ca