Section 1 Data
Data we need for the Proof of Concept
The Proof of Concept requires five years of store-level history, using the fields listed below. Sales and inventory must be daily, by store, SKU and size, because coarser data reduces forecast accuracy. All transfers are encrypted, and access is limited to named people on your team and ours.
| Data set | Required fields | Grain | History |
|---|---|---|---|
| Sales (system of record) | Store ID, SKU, style, color, size, transaction date (local store date), units sold, net sales after discount, units returned or negative sales | Store × SKU × size × day | 5 years |
| Inventory and supply | Daily on-hand by store, SKU and size; on-hand by DC, SKU and size; open purchase orders with expected receipt dates; in-transit stock (shipped, not yet received) | Store × SKU × size × day | 5 years |
| Product details | Style, color, size, category, season or lifecycle label, launch date, MSRP, active or discontinued flag | Per SKU | Full list, including discontinued items |
| Store details | Store ID, name, open date, square footage, region or market, climate zone (if set), existing cluster ID (if set) | Per store | Full list, including closed stores |
| Promotions and pricing | Start and end dates, promotion type, scope (store, region or chain), markdown schedules or rules (if available) | Per promotion | 5 years |
| Events (optional) | Store ID, event dates, event type (VIP, activation, local marketing) | Per store and date | 5 years, if available |
If a data set has less than five years of history, send what you have and tell us the earliest date.
Section 2 Transfer
How to send the data
We accept data in two ways. Secure file transfer (SFTP) is the preferred method for the Proof of Concept: structured CSV exports, encrypted and sent on a set schedule. Only named people on your team can send data, and only through the preferred channel.
- 1
Secure file transfer (preferred)
Structured CSV exports over SFTP, encrypted, sent on a set schedule.
- 2
API
An authenticated service account, with endpoint documentation and documented rate limits.
Preferred Channel
Preferred Channel: Optierra-managed SFTP
Send Proof of Concept data over SFTP only. Email attachments and consumer file-sharing links are not approved channels, because they do not meet our security requirements. Your access is limited to named people on your team, and every transfer is encrypted.
Channel
SFTP over SSH, port 22. Files are encrypted in transit and at rest on our side.
Recipient
The Optierra approved location. The Optierra support team sends the host, the account name and your company's folder. Access is set up for named people only.
Questions and access
The Optierra support team handles access requests, format questions and delivery confirmations at support@optierra.com.
Steps for sending data
- 1
Ask for access
Email support@optierra.com and tell us who on your team will send the files. We will email you the SFTP host, an account name, a temporary password, and the folder for your company.
- 2
Sign in and change your password
Connect with any SFTP client (FileZilla, WinSCP, Cyberduck, or the command line) over SSH on port 22. You will be asked to change the temporary password the first time you sign in.
- 3
Export one CSV per data set
Use UTF-8 text, comma-separated, with field names in the first row and no merged cells or summary rows. For sales and inventory, use one row per store, SKU, size and day, with local store dates. Send inventory as three files: store, DC and inbound.
- 4
Name each file
Use company, data set, start date and end date. Example: acme_sales_20210101_20260101.csv.
- 5
Upload to your folder
Put the files in the incoming folder we give you. Never overwrite an earlier delivery. For every update, upload a new dated file.
- 6
Confirm the delivery
Email support@optierra.com when the upload is done. We check that the files arrived, open, and have the right field names, then confirm back to you.
- 7
Send the sample first, then the full history
Send 2 to 4 weeks of sales and one full inventory snapshot first. Once we confirm it, send the full five years on the agreed schedule, so the Proof of Concept runs on current data.
Section 3 Checklist
Questions to confirm before we load the data
Your answers tell us how to read your data correctly. Tick each one when it is done, or write the answer under it. We do not begin modeling until these answers confirm the data quality we need.
Section 4 Process
Your allocation and replenishment process
Describe how you allocate and replenish stock today. This lets us build the Proof of Concept around your process, not a generic one. Write your answer in the right-hand column. Short answers are fine. If you are not sure of an answer, leave the row blank and we will go through it with you.
| Topic | What we need to know | Your answer |
|---|---|---|
| Who decides | Who sets store allocations and replenishment orders | |
| Allocation rules | How stock is split across stores today (rules, tiers, size curves, sales history) | |
| Allocation timing | When allocation happens (on receipt, at season start, weekly) | |
| Replenishment trigger | What starts a replenishment order (min and max, weeks of cover, reorder point, manual) | |
| Replenishment cadence | How often orders run, and the cutoff time for each run | |
| DC to store lead time | Days from DC dispatch to store receipt | |
| Supplier to DC lead time | Days from order to DC receipt | |
| Constraints | Minimum or maximum units per store, pack sizes, minimum drop size | |
| Store grouping | Store tiers or clusters used today, and how they are set | |
| Tools | Systems or spreadsheets used for allocation and replenishment | |
| Sign-off | Who approves allocation and replenishment decisions | |
| Success measures | What you measure (sell-through, stock-outs, margin, inventory days) |
Section 5 Outcomes
What the Proof of Concept will show you
The Proof of Concept tests each of these outputs on your own store data, so you can see what they show for your business.
- Early warning of stock-outs, store by store
- Store-specific size curves
- Allocation that accounts for stock already on order, before it arrives
- Signals when inventory is building up
- Markdown timing by store cluster
- Less manual allocation work
- Better margin from each unit of stock