lumtify
Türkçe

Retail Planning

Retail planning map

A season in fashion retail is dozens of coupled decisions. This map lays all of them out: which algorithm answers which question, and at which stage it comes in.

Algorithms
38
Algorithm groups
152 phases
With an open write-up
1

Retail planning is not one problem but a chain that feeds itself. A budget decision made before the season determines what you can ship to which store during it; the initial price decision narrows the room markdown timing will have.

The map below shows that chain end to end: seven stages before the season, eight during it. Algorithms that expand carry a short note on what they do; the ones with a dot beside them have an open technical write-up published.

Pre-season

7 stages · 12 algorithms

Decisions made before the season starts. Mistakes here are expensive: a badly built range or a misallocated budget cannot be fully corrected once the season is running.

  1. 01

    Merchandise Financial Planning

    • Long-range demand forecast (macro)

      Revenue and unit targets at category and budget level before the season.

  2. 02

    Range Plan

    • Product lifecycle curve forecasting

      Predicting a product’s weekly sales curve across the season.

  3. 03

    Product strategy

    • Consumer trend tracking

  4. 04

    Collection development

    • AI product design recommendation

    • Product performance forecasting

    • SKU rationalization

      Trimming products that cannibalize each other; narrowing range width.

  5. 05

    Assortment Planning

    • Initial price setting

    • Mid-range demand forecast

    • Size analysis (PrePack optimization)

      Deciding the size mix inside the pack that ships to each store.

  6. 06

    Store clustering

    • AI store grouping

      Grouping stores that behave alike; allocation and pricing decisions run on these groups.

  7. 07

    Sourcing

    • Supplier performance and risk tracking

    • Supplier selection recommendation

Season starts

In-season

8 stages · 26 algorithms

Decisions made once the season is running. Speed decides the outcome here: a week of delay turns sellable stock into markdown.

  1. 08

    Allocation — initial shipment

    • Automated initial allocation

      Which store gets how many units and when, under production, weather and capacity constraints.

    • Allocation strategy recommendation

  2. 09

    Replenishment

    • Forecast-driven automatic replenishment

      Optimal dispatch of limited stock by store capacity and potential.

  3. 10

    RPT

    • RPT requirement forecasting

    • RPT quantity calculation

  4. 11

    Inter-store transfer

    • Inter-store product optimization

      Moving a slow-selling product to a store that can sell it, under logistics cost and sell-through probability.

      Read the technical write-up →

    • Block, single-unit and broken-size decisions

    • Store opening/closing impact

  5. 12

    Markdown, promotion and campaign

    • Price elasticity model

      Measuring how a price change moves sales and margin, by product hierarchy.

    • Markdown optimization

      Which product gets which discount depth in which week, under end-of-season targets.

  6. 13

    Business intelligence

    • Stock-out and lost sales calculation

      Quantifying sales missed at SKU level during stock-out periods.

    • Broken-size and idle stock analysis

    • Algorithm performance measurement

  7. 14

    Warehouse and logistics

    • Network design optimization

    • Warehouse layout optimization

    • Picking optimization

    • Route optimization

    • Return packing optimization

  8. 15

    CRM and customer analytics

    • Customer segmentation

    • Churn prediction

    • Lifetime value prediction

    • Recommendation systems

    • Dynamic pricing

    • Product ranking algorithm

    • Review classification

    • Campaign audience building

dot: open write-up published

We explain some of these algorithms openly — on a synthetic dataset, with the mathematical model and an end-to-end SQL and Python solution:

perakendeanalitigi.com