Docs

Workspaces

A workspace is one store or one client. Its orders, analyses and chats live there, apart from every other workspace. This page covers getting orders in, reading the figures at the top of the workspace, and what the data quality pips mean for your reports.

What a workspace is

Workspace page header: Your analyses, workspace name, data quality, date range, last order, and five store figures with sparklines
  • One store per workspace. A merchant with two storefronts, or an agency with ten clients, keeps each in its own workspace so reports never mix customers.
  • Switch from the workspace name in the header. Manage workspaces lists every workspace you can open, grouped by organization.
  • Everything inside is scoped to it: imports, analyses, spaces, custom segmentations and chats.

Your plan sets how many workspaces you can create and how many orders each one holds. The workspace switcher shows usage against both.

Create a workspace

Open the workspace switcher and choose New workspace, or use the New workspace tile on the workspaces page.

Store website (optional)
The site's icon becomes the workspace's logo, and the name is suggested from it.
Name
Up to 40 characters. Shown in the switcher and on shared reports.

After Create workspace you land in the new workspace with the import dialog open. No export at hand? An empty workspace offers Explore a demo store first: four synthetic stores (a garden centre, a coffee roaster, a skincare brand and an apparel brand) with reports ready to open.

Import orders

There are three ways to get orders into a workspace:

CSV upload
Import data from the workspace name menu, or Import new orders next to the date range. One row per line item.
Webhook sync
Orders flow in as they are placed in Shopify, WooCommerce or Haravan. Set it up under Integrations.
API
Push orders from your own systems. See the API reference.

CSV columns

Shopify, WooCommerce and Haravan exports are recognised and mapped automatically. Shoper’s export is reshaped to one row per line item in your browser. Any other CSV works too: match your columns to the fields on the Configure columns step.

Required
Order ID, Order Date, Product ID (the product name or code).
Customer ID
Optional, but needed for customer segments, retention, reorder timing and audiences. An email address works best: audiences can go straight to your email tool.
Optional
Product SKU, Price, Quantity (1 if missing), In Stock, Category.
Custom fields
Any other columns you want to filter reports by later, such as brand, region or channel.
Upload data dialog: drop a CSV, required fields Order ID, Order Date, Product ID, optional fields

Good to know

Every import merges into what is already there: rows with the same order and product are updated, new ones are added. Re-uploading an overlapping export is safe. To start over, use Clear data first (owners only).

Step-by-step export guides for Shopify, WooCommerce, Magento, Shoper, Haravan and Shopee are under Guides in the dashboard’s profile menu.

The store figures

Five figures sit at the top of every workspace, each with a 12-month sparkline. Hover the sparkline for any month’s value and change.

orders
Orders in the data. The change line compares the last 30 days with the 30 before.
avg order value
Average revenue per order, same comparison.
customers
Identified customers. The change line is new customers in the last 30 days against the total before.
repeat rate
Share of customers who ordered more than once. Amber under 10%: segmentation needs a second order to work with.
customer LTV
Revenue from identified customers to date, divided by customers.

A dash means the figure cannot be computed from this data, with the reason next to it: no prices, no customer IDs, or no orders in the last 30 days.

How to read it

“Last 30 days” means the last 30 days of your data, not of the calendar. The Last order label next to the date range tells you how fresh the data is. It turns amber, with an Import new orders link, once the newest order is more than a week old.

Data quality

The Data quality pips grade how much each analysis will have to go on in this workspace: Bought together, Customer segments, Repeat-purchase patterns, Reorder timing and Audiences. Dark green is strong, teal fair, amber weak, grey unavailable. Hover the pips for the reason behind each grade.

Weak does not mean broken: the analysis still runs, with fewer or less certain patterns. Unavailable means a column is missing, usually the customer ID.

Data quality hover card: every analysis runs, some with less to go on, with a line per analysis

Each analysis takes the grade of its weakest input:

Multi-item orders
Share of orders with 2+ products. Strong from 35%, fair from 15%. Drives Bought together.
Orders
Strong from 2,000, fair from 300. Drives Bought together.
Repeat rate
Strong from 25%, fair from 10%. Drives Customer segments, Repeat-purchase patterns and Reorder timing.
History
Strong from 180 days, fair from 60. Drives Repeat-purchase patterns and Reorder timing.
Customer ID coverage
Strong from 70% of orders. Without any, everything except Bought together is unavailable.

What to do with it

  • Low multi-item share: basket rules will be thin. Start with a customer segments report and reorder timing instead.
  • Short history: import older orders. Reorder cycles need at least a few cycles to be visible.
  • No customer IDs: re-import with the customer or email column mapped. It unlocks four of the five analyses.

Analyses

Analyses list: cards grouped by week and month, each with a headline figure, date, rules count and date range

Every report you run appears here, newest first, grouped into This week, Last week and earlier months. Each card leads with its headline figure: the reachable revenue for a market basket report, the customer count for customer segments. Switch between cards and a table with the toggle on the right. Once you have spaces, the Not in a space tab shows the analyses not filed yet.

Hover a card for its actions:

  • Re-run on fresh data after importing new orders. Earlier runs stay available in the report’s run picker.
  • Schedule automatic re-runs, so the report stays current.
  • Copy link: always opens the latest run.
  • Share, Move to space, Pin to top, Delete.

Read what each report shows in the market basket and customer segments guides.

Chat

Chat view: Where should we start? with suggested questions and a composer

Chatin the header opens an assistant that works on this workspace’s data. It reads your reports, looks up products, runs new analyses when a question needs one, and builds audiences. Its answers come from your numbers, not guesses.

  • Chats are private to you and belong to the workspace.
  • The folder chip under the composer files any analysis the chat runs into a space. Opening chat from a space preselects it.
  • The same tools are available in Claude, ChatGPT and other clients over MCP. See Merchandising Engineer.

Team and settings

Workspaces belong to an organization. Everyone in the organization works in the same workspaces, analyses and reports. Invite teammates by email from Organization in the profile menu.

Owner
Billing, members, API keys, clearing and deleting workspaces.
Member
Everything else: imports, analyses, spaces, audiences and reports.

From the workspace name menu: Import data, Preview data (order activity and sample rows), Integrations, Refresh stats, and for owners Clear data (removes all orders and imports, keeps the empty workspace) and Delete workspace. Both cannot be undone.

Spaces →Group analyses around one question and record what you decided.Market basket report →Your first analysis: what sells together and when.