The diagnostic  /  Data specification

What to export, field by field.

One page for your IT or MIS team, before the first call. Your column names do not have to match ours: we map your export to these fields in the first week and tell you what is missing. Send files exactly as your system produces them.

Fresh categories only: fruit and vegetables, dairy, bakery.

The rules

Format

CSV (UTF-8) or Excel, one file per table. Dates as YYYY-MM-DD; times as HH:MM, store local time. Quantities in the unit you sell in, with a unit column (kg or pieces). Money in rupees.

How much history

Twelve months if you have it. Six months is enough for where the money goes and a backtest on those months. A year is needed to compare festival seasons and measure seasonality from your own history; with less, those effects come from our calendar and are flagged as assumptions.

What not to send

No customer data of any kind. Remove staff names, employee IDs and phone numbers before export; if any arrive, we delete them on receipt.

How to send

Only after the NDA and data-sharing agreement are signed, through a secure link we provide. Data is held in an India cloud region and deleted at the end of the diagnostic unless you ask in writing to keep it.

Required

Four files.

1. Daily sales and stock one row per store × item × day

FieldMeaningExample
dateTrading date2026-09-28
store_idYour store codeS08
item_idYour item code (or PLU for loose produce)FV07
unitkg or pcskg
soldQuantity sold that day23.5
closing_stockStock on hand at close, as the system records it12.0
receivedQuantity received that day (from DC, collection centre or supplier)30.0
wastedQuantity written off that day (expiry, damage), if recorded1.5

2. Item master one row per item

FieldMeaningExample
item_id, nameCode and descriptionFV07, Lemon
categoryFruit & veg, dairy or bakery (your own hierarchy is fine)Fruit & veg
unitkg or pcspcs
shelf_life_daysTypical days the item can be sold after receipt7
case_packUnits per crate, case or tray as ordered50
price, costSelling price and purchase cost per unit, in rupees5.00, 3.20

3. Store list one row per store

FieldMeaningExample
store_id, name, cityCode, name and cityS08, Ameerpet, Hyderabad
opened_onOpening date, if under two years old2025-06-01
size_sqft, formatSelling area and format, if you keep them6500, Supermarket
supplied_byThe DC or collection centre that supplies its freshDC1

4. Orders one row per store × item × order date

FieldMeaningExample
order_date, store_id, item_idWhen and what was ordered or allocated2026-09-27, S08, FV07
ordered_qtyQuantity ordered or allocated40
placed_byStore or head office/DC (a role, never a person's name)Store
If you have them

Each one sharpens a finding.

FileFieldsWhat it adds
Sell-out timesdate, store_id, item_id, last_sale_time on days the item ran outEstimates demand you missed from when the shelf emptied, not just that it did
ERP reorder settingsstore_id, item_id, min, max, reorder level, review cycleThe backtest against the rule your ERP actually uses
Promotion calendaritem_id, store_id (blank = all), start date, end date, type, discount %Keeps promotion days out of normal demand
Write-off reasonsReason code on each write-offSplits waste into ordering, handling and damage
DC or collection-centre stockdate, node, item_id, usable stock at day endSeparates store ordering problems from supply shortfalls
Audited write-off line"Inventory written off" for the last three years, and its valuation basisReconciles store-level waste to your accounts

Questions about a field? Write to [email protected]. A food brand planning launches needs different data; it is listed on the launch review page.