Apparel Pack Optimiser — Documentation

Apparel Pack Optimiser is a Snowflake Native App from SCIP Solutions. It takes your sales plan by store and size and works out the most profitable way to pack the buy: single ratio packs, two- and three-ratio assortments, single-size packs and mixed strategies. Every option is costed on operating profit after supply chain handling, store handling and markdown, then ranked.

The app runs entirely inside your own Snowflake account. It makes no external network calls, holds no credentials, and reads only the tables or views you explicitly connect.

Before you start

You need a Snowflake account and a running warehouse. An X-Small warehouse is enough. You do not need to move any data: the app runs where your data already lives.

1. Install the app

Install Apparel Pack Optimiser from the Snowflake Marketplace. After it installs, grant the app's user role to the roles that should open it:

GRANT APPLICATION ROLE <app_name>.APP_USER TO ROLE <your_role>;

Replace <app_name> with the name you gave the app when installing, and <your_role> with the role your planning team uses. Until you run this, the app will report that it is not accessible.

Then open the app from Apps in Snowsight.

2. Load a sales plan

There are three ways to get a plan into the app.

Try demo data. A sample seasonal range is built in. One click and you have a full result in under a minute, with no data preparation. Use this first to see how the app works.

Upload a CSV. The app accepts a CSV with one row per store and one column per size, in size order, smallest to largest. An optional first column names the product so you can load a whole range at once.

product,store,size_6,size_8,size_10,size_12,size_14,size_16,size_18,size_20
Jersey tee,101,12,26,39,44,29,21,12,8
Jersey tee,102,9,19,31,38,27,18,10,6

A template is available from the upload screen.

Connect a Snowflake table or view. Press Connect a table, pick the object in the Snowflake dialog, and the app reads it through a reference. The app only ever gets read access to the object you choose, and you can disconnect it at any time.

The table must have the same shape as the CSV: one row per store, one column per size, in size order. If your planning data is one row per store and size, create a view that pivots it first. For example:

CREATE OR REPLACE VIEW planning.apo_sales_plan AS
SELECT * FROM (
    SELECT style_code, store_id, size_code, planned_units
    FROM planning.store_size_plan
    WHERE season = 'AW26'
)
PIVOT (SUM(planned_units) FOR size_code IN ('4','6','8','10','12','14','16','18'))
    AS p (product, store, size_4, size_6, size_8, size_10, size_12, size_14, size_16, size_18);

Replace the schema, table and column names with your own, and list your size codes in the IN list in size order — the aliases after AS p become the app's size columns. If your size codes are numbers rather than text, drop the quotes.

3. Set your inputs

Before running, set the assumptions the costing uses:

  • Cost and retail price per unit.
  • Units per pack. A pack must hold at least one unit of every size, so the minimum is the number of sizes in your plan.
  • Markdown depths. Four bands, applied to stock in excess of the sales plan: 0-20% excess, 20-50%, 50-100% and 100%+. These decide what excess stock is worth when it clears.
  • Rainbow ratios. On by default. Restricts every ratio to a single-peaked curve, which is what a merchandisable size curve looks like. Turn it off only if you are willing to consider ratios that rise and fall across sizes.

These inputs drive the answer. The recommendation is only as good as the costs you give it, and it is worth agreeing them with your planning and supply chain teams before you rely on a result.

4. Read the results

The app compares every pack strategy on operating profit and ranks them:

  • Single ratio. One ratio for every store. Results are provably optimal: every legal split of the pack across sizes is evaluated.
  • Two- and three-ratio assortments. Stores can be supplied from a mix of ratios. Results are the best found over a strong candidate set, with every store's exact pack mix solved for the best-scoring combinations. The app labels these "best found".
  • Single-size packs. Cartons of one size only, in several carton sizes.
  • Mixed strategies. A ratio pack topped up with single-size cartons.

Every plan the app proposes covers planned demand in full for every store and size. What differs between strategies is how much excess stock is needed to do it, and what that excess costs after markdown.

For each product you see the recommended strategy, the units required, full-price sell-through, the operating profit, and the ratio packs themselves. The Profit vs Inventory view plots profit against units ordered, so you can see which options buy the most profit for the least stock — useful when the most profitable option is not the leanest one.

5. Export

Three exports are available from the results screen:

  • Results (CSV) — every strategy with its full costing, for your own analysis.
  • Report (PDF) — a summary suitable for a buying meeting.
  • Store pack workbook — packs by store and units per size, which is the file your supplier needs.

Run times

Run time depends mostly on the number of products. As a guide, on an X-Small warehouse a 20-product range across 40 stores completes in about 30 seconds, and 5 products across 300 stores in a few seconds. Larger ranges scale roughly in proportion to the number of products.

Your data stays in your account

The app runs inside your Snowflake account, on your warehouse. It makes no external network calls, holds no credentials, and reads only the tables or views you explicitly connect, with SELECT access only. SCIP Solutions has no access to your data. You can review and revoke what the app can see at any time from the app's Security tab in Snowsight.

Reproducibility

The same input always produces the same answer. A result can be reproduced exactly from the sales plan and the inputs used, which matters when a recommendation is questioned in a buying meeting.

Support

Email contact@scipsolutions.com.au. If you are reporting a problem with a result, include the sales plan and the input settings you used, and we can reproduce it exactly.

About SCIP Solutions

SCIP Solutions is a supply chain, inventory planning and retail analytics business founded by Oliver Blombery, former General Manager of Inventory and Supply Chain at Kmart Australia.