Canceled Orders (24h) for the selected period.
At a glance
A rolling 24-hour count of Square orders that moved into a cancelled state. A handful is normal operational noise (a customer changed their mind, a duplicate was voided), but a sudden cluster usually points to a fulfilment problem, an oversell, or a payment issue. It is an early operational warning that pairs directly with the oversell and drift alerts.
| What it counts | The number of orders whose Square Order state is CANCELED with a cancellation timestamp in the last 24 hours, across all channels and locations. Sourced from the Square Orders API. |
| Channel / source treatment | All channels combined. Cancellations can originate from Square Online web orders, Square POS voids, or Square Invoices. Filter by source.name to see which channel is driving them. |
| Currency / unit | Count of cancelled orders (whole number). The card focuses on count; the lost value is captured on the revenue-at-risk and refund cards. |
| Time window | 24H (rolling last 24 hours) |
| Alert trigger | Fires when more than 3 cancellations occur in the 24-hour window. The threshold is deliberately low because a cancellation cluster is a fast-moving operational signal. |
| Roles | owner, operations |
| Cancelled vs refunded | A cancellation stops an order before fulfilment; a refund reverses money after the fact. An order can be cancelled without a refund (never paid) or refunded without cancellation (returned after delivery). |
| What a cluster usually means | An oversold item triggering forced cancellations, a payment processor issue, a fulfilment outage, or a pricing / listing error that customers are reversing. |
Calculation
Calculated automatically from your Square Online data. See the At a glance summary above for what the metric tracks and the worked example below for a typical reading.Worked example
A US outdoor gear merchant on Square. One store plus a Square Online storefront. The card is checked at the start of trade on 14 Mar 26, covering the previous 24 hours.| Channel | source.name | Cancellations (24h) | Typical baseline | Likely cause |
|---|---|---|---|---|
| Square Online Store (web) | Square Online Store | 6 | 1 | Oversold tent model |
| Square POS (in store) | Square POS | 1 | 1 | Duplicate void at register |
| Square Invoices | Square Invoices | 0 | 0 | none |
| Canceled Orders (this card) | 7 | ~2 |
- The threshold of 3 is comfortably breached. Seven cancellations against a typical baseline of around two is a clear cluster, and the alert fires. The single POS void is routine; the six web cancellations are the story.
- The channel split points straight at the cause. All six web cancellations are for one tent model, which matches an oversell on that item. The next move is to open Oversell Risk and POS to Online Inventory Drift Alert to confirm and stop further orders.
- The 24-hour window makes it actionable, not historical. Unlike a 30-day cancellation rate, this card is tuned to catch a problem while it is still happening. By the time a monthly metric moves, the damage is done; here, six cancellations in a day prompt action the same morning.
Sibling cards merchants should reference together
| Card | Why pair it with Canceled Orders (24h) |
|---|---|
| Oversell Risk (negative on-hand projected) | The most common upstream cause. Oversold items force cancellations; if this card spikes, oversell risk is the first place to look. |
| POS to Online Inventory Drift Alert | Drift causes oversell, which causes cancellations. A cancellation cluster on one item often traces back to a parity gap. |
| Refund Rate | The money side. Cancellations before payment cost nothing; cancellations after payment turn into refunds. Watch both to see whether revenue is actually leaking. |
| Order Completion Rate | The positive mirror. A cancellation cluster pulls completion rate down; pairing them shows the net health of fulfilment. |
| Orders by Status | Gives the full state breakdown, open, completed, cancelled. Useful for seeing whether cancellations are rising as a share of all orders. |
| Total Orders | Context for whether the cancellation count is alarming relative to overall volume. Seven cancellations is very different on 50 orders than on 5,000. |
Reconciling against Square
Where to look in the Square Dashboard: Square Dashboard, Items & Orders, Orders. Filter the order list by a cancelled status and set the date range to the last day. The count of cancelled orders should match this card closely. Open individual orders to read the cancellation note and the channel (source.name), which tells you why each one was reversed.
Other Square Dashboard views that look like the same number but aren’t:
- Orders, filtered to cancelled, last 24h: this DOES match closely. It is the direct comparison.
- Reports, Refunds: counts money returned, not orders cancelled. An order cancelled before payment never appears here.
- Reports, Sales summary: nets out cancellations from sales but does not count them as a standalone figure.
- Transactions / Payments: counts payment events. A cancelled, never-paid order has no payment to show.
| Reason | Direction of divergence |
|---|---|
| Time-zone and the rolling window. This card uses a rolling 24 hours in UTC; the dashboard filter uses calendar days in the location’s local time zone. Orders near the boundary fall differently. | Boundary differences of a few orders |
| State timing. An order cancelled then later edited, or a draft abandoned, may be classified differently. We count orders whose state is cancelled with a timestamp in the window. | Minor differences around edited orders |
| Channel filtering. If you filter the dashboard to one location or channel, it will be lower than this all-channel card. | Dashboard lower when filtered |
| Sync lag. The most recent cancellations may take a short cycle to reach our index. | Self-resolves within minutes |
| Card | Expected relationship | What causes legitimate divergence |
|---|---|---|
| Refund Rate | Paid cancellations should also appear as refunds | A cancellation before payment never generates a refund, so cancellation count can exceed refund count. |
google_analytics.conversion-rate | No direct relationship | GA4 records the web purchase event at checkout, not the later cancellation. A web order cancelled after the fact still counts as a GA4 conversion, so the two will not tie out. Treat Square as the source of truth for cancellations. |