# EXAMPLES: three reports from a fictional store
A small online shop that sells ceramic mugs. Every number below is invented for the example. Amounts are in plain currency units.

## Example 1: contribution margin for one order, with a gap

The owner asks: "Did the order with two blue mugs make money?"

```
Recommendation: yes before ads, but the figure after ads is incomplete until the card fee rate is filled in.

| Line                          | Value   | Source                                   |
|-------------------------------|---------|------------------------------------------|
| Net sales excl. tax           | 64.00   | orders export, order #1042               |
| Shipping charged              | 5.00    | orders export                            |
| Cost of goods (2 x 11.50)     | -23.00  | CONTEXT, margin inputs                   |
| Packaging and fulfilment      | -2.50   | CONTEXT                                  |
| Carrier cost                  | -6.20   | CONTEXT, zone 1                          |
| Payment fees                  | no data | card fee rate missing in CONTEXT         |
| Platform fee per order        | -0.90   | CONTEXT                                  |
| Refunds and returns           | 0.00    | refunds export, none for this order      |
| Margin before ads             | 36.40 + payment fees (incomplete) |                |
| Ad spend allocated            | -9.80   | ad report, split by revenue              |
| Margin after ads              | 26.60 + payment fees (incomplete) |                |

What I checked: orders export for 1 to 7 March, store timezone; order total matches the admin order page.
No data: card fee rate. Blocks every margin figure until filled in.
Decisions for you: 1. Add the card fee rate to CONTEXT.md, or tell me to use a default and label it assumed.
```

What makes this correct: the missing fee stays visible instead of being filled with a typical rate, and the ad allocation method is named.

## Example 2: reorder signal

```
Recommendation: reorder the blue mug this week; the green mug can wait.

Inputs from CONTEXT: lead time 21 days, safety days 7, cover days wanted 28, window N = 28 days.

| Product    | Stock | On order | Sold in 28 days | Velocity/day | Reorder point | Days of cover | Suggested qty |
|------------|-------|----------|-----------------|--------------|---------------|---------------|---------------|
| Blue mug   | 90    | 0        | 112             | 4.0          | 112           | 22            | 134 -> 144 (minimum 48 per colour) |
| Green mug  | 160   | 0        | 56              | 2.0          | 56            | 80            | none          |
| Grey mug   | 0     | 48       | 30              | no data      | no data       | no data       | no data       |

Grey mug: out of stock for 12 of the 28 days, so sales understate demand. Velocity not calculated; decide with the supplier's next delivery date.
Decisions for you: 1. Place the blue mug order (144 units)? 2. Grey mug: extend the window, or use the 16 in-stock days only?
```

Check the blue mug by hand: 4.0 x (21 + 7) = 112, and 90 on hand is below it. Suggested: 4.0 x (21 + 7 + 28) - 90 = 134, rounded up to a multiple of the supplier minimum.

## Example 3: reply draft for customer service

Customer message: "My mug arrived cracked. Can I get a new one before the weekend?"

```
Draft (for you to send):
Hello, we are sorry the mug arrived cracked. Could you send a photo of the mug and the box? As soon as we have it, we will send a replacement at no cost to you. It usually ships within 2 working days.

What this commits the store to: a free replacement (covered by the damaged-in-transit policy) and dispatch within 2 working days (the dispatch time in CONTEXT.md). It does NOT promise delivery before the weekend; that depends on the carrier and is not guaranteed by the policy.
```

What makes this correct: the draft answers the question, quotes only what the policy allows, and tells the owner which promise it avoided.