Most late deliveries in a workshop are decided weeks before anyone notices. The job was never late on the day it should have shipped. It was late on the day it should have started. Working out that earlier date, honestly, for every open line is one of the most useful things a planner can do.
Start from the date you are actually aiming for
There are usually two dates on a line: the date the customer asked for, and the date your team now plans to despatch. If a planner has agreed a later date with the customer, that is the date the shop floor is working to, so it is the date to count back from. If no plan has been set, fall back to the required date, and treat that line as unplanned.
Count working days, not calendar days
A ten-day job does not start ten calendar days before despatch. Count back through working days only: skip weekends and public holidays, and add any shutdown days your site takes. A job that must be despatched on Friday 20 November and takes ten working days must start no later than Friday 6 November, because the ten working days back from the 20th are the 19th, 18th, 17th, 16th, 13th, 12th, 11th, 10th, 9th and 6th.
Around Christmas, Easter and the August bank holiday the same job needs noticeably more calendar days, which is exactly when a flat rule of thumb lets you down.
One make time for every job is only a starting point
A standard make time, say ten working days, is a sensible default for a first pass. It is also wrong for most jobs. A simple bracket and a large welded frame with machining and NDT do not take the same time. Set a figure for each job where you know it differs, and leave the default in place only where you do not.
The default is a planning assumption, not a routing. Say so on the schedule, so nobody mistakes a rule of thumb for an engineering estimate.
Know the difference between late and not yet started
A line past its start-by date is not automatically a problem. Record why, in a way the whole team can see:
- In process: work has started, so it should not be flagged as a late start.
- On hold by the customer: the delay is theirs, and the schedule should say so.
- Awaiting material: the job cannot start, which is a purchasing action, not a shop-floor one.
- Complete: finished, awaiting despatch or paperwork.
Lines with no status and a missed start-by date are the ones that need a conversation this week.
A weekly routine that fits in half an hour
- List every open line whose start-by date has passed or falls within the next five working days.
- For each, confirm the status: started, held, awaiting material or genuinely behind.
- For the ones that are genuinely behind, decide: add resource, resequence, or agree a new date with the customer.
- Record the new planned date where everyone can see it, so the schedule and the conversation stay in step.
Where software helps, and where it does not
The arithmetic is easy to automate and tedious to do by hand across hundreds of lines. Nova’s Job Book does exactly this: it works out the start-by date for every open line, flags late starts and keeps planned dates and job statuses in one shared place. Deciding what to do about a late start is still a planner’s judgement, and it should stay that way.
You can try the calculation yourself on the Nova AI page.
Conclusion
Count back from the date you are really working to, use working days, set make time per job where you can, and record why a late start is late. That turns a vague worry about delivery into a short list of lines that need a decision today.