Nine days 'in production'
A customer asks, politely, when their hoodie will ship. You check the partner dashboard: the order has been 'in production' for nine days. Their usual time for hoodies is a few days. There is no message explaining why. You open a support ticket and learn that the blank in that colour was out of stock at the facility, and the order was waiting for a delivery.
When you look further, several other orders at the same partner have been sitting for a week. Nobody told you, because each order was technically still in progress.
Slow orders look like normal orders
Partner dashboards show status, not how long an order has been in that status compared with normal. A delayed order and an on-time order both say 'in production'.
- Partners quote typical production times, but individual orders vary.
- Holds for stock, file checks or quality issues are not always announced.
- Backlogs at one facility affect many orders at once.
- Marketplaces measure your dispatch time, not your partner's.
- Checking order ages by hand across partners takes time nobody has.
So the first alert is usually a customer message, which means the delay is already a problem.
What unseen delays cost
Late dispatch counts against you on marketplaces and makes customers anxious. Customers who ask before you know look like they are being ignored. Orders stuck for a reason you could fix, such as a colour you could substitute with the customer's agreement, stay stuck. And a partner running slow for weeks can damage your reviews before you notice the pattern.
At peak times, it is worse: every partner slows down, and knowing which one is slowest is the difference between moving orders and losing them.
The ageing watch we build
- Each order's status history is collected from the partner APIs: received, in production, printed, packed and shipped, with times.
- Normal times are worked out for each partner, product type and stage from your own order history, including seasonal differences.
- Orders taking noticeably longer than normal at any stage are flagged, with the partner and product.
- When several orders at one partner are running slow at once, that partner is flagged as a whole, which can pause them in a routing layer if you use one.
- For each flagged order, the watch shows any reason the partner has reported, and drafts a support query or a customer update for you.
- A weekly summary shows each partner's actual production times by product, next to what they quote.
| Signal | What it means | Suggested action |
|---|---|---|
| One order slow | An individual hold or issue | Query the partner, update the customer |
| Many orders slow at one partner | Backlog or facility issue | Consider pausing or rerouting |
| One product slow everywhere | Blank shortage | Check stock and substitutes |
| Seasonal slowdown | Expected peak effect | Adjust your stated processing time |
Normal times come from your own history, not the partner's marketing, so the watch reflects what you actually experience.
Delays you see first
When an order runs past its normal time, you know within a day, and you can ask the partner what is wrong before the customer asks you. When one partner slows down across the board, you see it and can move new orders elsewhere. Customers affected by a delay get an update from you, rather than discovering it themselves.
The weekly summary helps with bigger decisions, such as which partner to use for each product at peak times.
Signs production delays are hidden from you
- Customers tell you about delays before you know.
- You check order ages at partners by hand, now and then.
- Partner backlogs have affected many orders before you noticed.
- You do not know each partner's real production times.
- Your marketplace dispatch figures suffer at peak times.