24 weekly questions for operations, retail and finance teams
By SchemaChat (Data BI LLC) · Published 30 October 2026 · Last reviewed 30 October 2026 · 7 min read
Eight questions each for logistics operations, retail and e-commerce, and finance teams to ask of their own data every week. Each lists the data it needs, what a good answer looks like, and the trap that makes the number wrong. Use them with any tool, or paste them into SchemaChat.
Formulas below are standard definitions; we give no benchmark ranges, because good values depend on your business. Always reconcile numbers that leave the building to your source of record.
Logistics operations
| Question | Data needed | A good answer shows | The trap |
|---|---|---|---|
| What was our on-time delivery rate last week, by carrier? | Delivery date, promised date, carrier | On-time deliveries ÷ total deliveries × 100, per carrier, with counts | Missing delivery timestamps counted as late (or on time) |
| Which lanes had the most late shipments? | Origin, destination, delay | Top lanes with number of shipments, not just percentages | Tiny lanes where every shipment was late (2 of 2) |
| What is our cost per mile this quarter vs last? | Transport cost, miles driven | Total transport cost ÷ miles, with both periods | Cost and miles from different date bases |
| Which shipments have no delivery timestamp? | Delivery date | A count and a list | Treating blanks as zero days |
| Which customers' orders are late most often? | Customer, delay | Count and share, with a minimum-volume note | Ranking customers with very few orders |
| What is our average delivery distance? | Distance per delivery | Total distance ÷ number of deliveries | Including cancelled deliveries |
| How did capacity utilisation change by week? | Loaded vs available capacity | Week by week, in calendar order | Weeks ordered alphabetically |
| Which carriers had the biggest change in late deliveries? | Carrier, week, delay | Change per carrier with the counts behind it | Comparing a full week with a partial one |
Retail and e-commerce
| Question | Data needed | A good answer shows | The trap |
|---|---|---|---|
| What is the return rate by product, and what share of revenue is it? | Orders, returns, product | Return rate and revenue share together | Returns recorded as negative orders and double-counted |
| What is gross margin by category after returns? | Revenue, cost, returns | Margin per category with the formula used | Cost missing for some products |
| What is the repeat-purchase rate this month? | Customer ID, order date | Customers with ≥ 2 orders ÷ customers | Guest checkouts with no customer ID |
| Which products are in the top 10% of revenue? | Product, revenue | A ranked list with cumulative share | Variants of one product counted separately |
| How does revenue compare month over month? | Order date, revenue | Months in calendar order with percent change | A partial current month compared with a full one |
| Which channels have the highest refund rate? | Channel, refunds, orders | Rate and counts | Channels with very few orders |
| What is average order value by customer segment? | Segment, order total | AOV per segment with order counts | Outlier orders skewing the average |
| Which products sold well last month but have no sales this month? | Product, order date | A list with last month's totals | Seasonal products flagged as problems |
Finance
| Question | Data needed | A good answer shows | The trap |
|---|---|---|---|
| Total spend by category, this month vs last? | Spend lines, category, date | Both months side by side | Uncategorised spend hidden |
| Which categories moved by more than 10%? | Same | Categories with the amounts and direction | A threshold applied to tiny categories |
| Which vendors account for that movement? | Vendor, amount | Vendors ranked by contribution | The same vendor under two names |
| What is gross margin by product? | Revenue, cost of goods | Margin with the formula | Costs booked in a later period |
| Which invoices are overdue by more than 30 days? | Invoice date, due date, paid status | A list with ages | Paid invoices not marked paid |
| What is our revenue by customer, top 20? | Customer, invoice amount | Ranked with share of total | Credit notes ignored |
| How does this month's total compare with the ledger? | System total, ledger total | Both figures and the difference | Different cut-off dates |
| Which entries have a missing cost centre? | Cost centre | A count and a list | Blank vs “N/A” counted differently |
How to use these
- Start with the questions your data can answer; skip the rest.
- Ask them in the same order each week so changes are visible.
- When you ask the same question three weeks running, make it a dashboard or an alert.
- Reconcile the first run of each to a figure you already trust.