On Monday, an operator reports a leaking seal on a circulation pump. The repair itself takes four hours. Yet the order doesn't reach TECO (technically completed, the SAP status that closes the work) until six weeks later. Nobody did anything wrong: the notification waited for review, the order for approval, the part for a purchase order, the job for a suitable window, and the paperwork for someone with time.
Most maintenance organizations will recognize this. The lead time of a maintenance order is mostly waiting, not working. And that waiting rarely shows up in reporting. Backlog overviews show where an order stands today, not how long it took to get there. KPIs such as MTTR (mean time to repair) average everything out. And each queue has a different owner: planning, approval, purchasing, operations, administration. Everyone sees their own piece; nobody sees the whole chain.
The good news: the data to make this visible is already in your ERP system, such as SAP. Every status change, purchase requisition and confirmation is stored with a timestamp. Put them in sequence per order and you can see exactly where time is lost.
When you do, the same five places keep coming up. Here is why each one stalls and what to measure.
1. The notification waits for review
Notifications pile up with a single reviewer, are incomplete, or get reported twice.
Measure: time from notification to order by priority, and the share of notifications closed without an order.
2. The order waits for planning and approval
Approval sits with one person, or the order goes back and forth because the estimate or scope is off.
Measure: time from CRTD (created) to REL (released), time spent in the approval user status, and orders that receive the same status more than once.
3. Materials and purchasing
Often the biggest queue. The purchase requisition follows purchasing's release strategy, not the urgency of the order. And an order always waits for its last part.
Measure: time from purchase requisition through release and purchase order to goods receipt, the critical part per order, and time between receipt and execution.
4. Released, but not executed
The asset isn't available, the permit isn't ready, or urgent work takes priority. Note: releasing orders immediately at creation makes this waiting time invisible.
Measure: time from REL to first confirmation, and how often the planned start date was changed.
5. From technically complete to TECO
Hours, damage codes or a contractor invoice are still missing. The result: a polluted backlog, costs on open orders and less reliable damage codes.
Measure: time from last confirmation to completion, and month-end spikes in TECOs.
Finally
Back to the pump. Four hours of work, six weeks of lead time. Once you split those six weeks into the five places above, the conversation changes. It's no longer "maintenance is slow", but "orders spend eleven days waiting for a purchase requisition to be released" or "half our backlog is work that's done but not closed". That is a problem with an owner, and a problem you can fix.
The gain isn't in faster wrenching, but in shorter waiting. And most of that waiting sits outside maintenance: in purchasing, operations and administration. That's why the biggest improvements rarely come from a maintenance project alone, but from making the whole chain visible to everyone who is part of it.
So why track this at all? Because waiting time is never just an administrative number. It has a direct effect on the plant:
- Risk and downtime. The longer a repair waits, the longer an asset runs in a degraded state or on a temporary fix. A small defect that waits weeks can become a breakdown.
- Reliable planning. If you don't know how long orders really take, you can't plan shutdowns, contractors or capacity. Backlog figures only mean something if you know what is truly waiting and why.
- Costs and working capital. Uncertain lead times lead to rush orders, expediting costs and "just in case" spare parts on the shelf.
- Safety and compliance. Safety-critical and statutory work that gets stuck in a queue becomes overdue, often without anyone noticing in time.
- Better data for reliability. Orders closed quickly and properly give you reliable damage and cause codes, the basis for tackling repeat failures.
- Trust on the shop floor. When operators see that reported issues are fixed quickly, they keep reporting them. When they don't, notifications dry up and problems go unseen.
You don't need to tackle all five at once. Start with one order type, find the largest queue, and take the numbers to the team that owns it. One concrete figure on the table usually does more than a year of general complaints about lead time.