The order confirmation lands on a Monday: a conveyor frame with a platform, six weeks to delivery. Everyone knows what needs to happen, and yet the drawing package only reaches the shop floor three weeks later. Not because anyone spent three weeks drawing; the actual design and drafting work took less than four days. The rest of the time the order sat waiting: for an answer from sales, for customer approval, for a review, for someone to spend one more afternoon churning out PDFs and BOMs at the very end.
Sound familiar? Then follow along. In this article we trace a single order from quote to production drawing and look at where the days really go. Engineers who want to reduce engineering lead time usually hunt for faster modeling. A shame, because that is rarely where the time sits. The real gains are in waiting, handovers, search time and the repetitive publishing work at the end of the chain. Part of that you fix with agreements, part with tooling such as Thundercad, the toolbox for Autodesk Inventor.
At the end you will find a step-by-step plan that measurably shortens the chain, without anyone working harder.
Follow the order: where fifteen working days go
Take a mid-sized order as an assumption: fifteen working days between a signed quote and a released drawing package. Ask the engineer how much time the work took and he will say: about four days. So where did the other eleven go? Write it out phase by phase and the split usually looks something like this:
- four days of actual design, modeling and drafting;
- three days waiting for technical answers from sales or the customer;
- two days waiting for internal review and release;
- two days sitting still between handovers: from sales to engineering, from engineering to work preparation;
- one day searching: for a similar earlier project, the right revision, the last email with that one dimension;
- one day publishing: exporting, printing, preparing BOMs, sending everything out.
The ratios differ from company to company, but the pattern is remarkably stable: most of the lead time is not work time but wait time. That is good news, because removing wait time is far easier than drawing faster.
Waiting: the silent day eaters
The biggest waits sit at the front. The quote is sold, but the technical details are thin: the layout drawing is missing, the drive is "still being agreed", the customer has yet to choose between two versions. Engineering starts anyway, hits an open question, sends an email, and puts the order aside. Every time an order is picked up and put down again, time evaporates: reading back in, restarting, reopening the same files.
Two agreements go a long way here. One: a fixed intake checklist for the handover from sales to engineering. Only complete orders enter the planning; incomplete ones stay on the sales desk. Two: bundle open questions. Not an email per question, but one decision list per order, with an owner and a date for every item. That prevents an order from stalling five separate times on five loose ends.
The same goes for reviews. A drawing that gets looked at "when there is a moment" sits waiting for days on average. A fixed fifteen-minute review slot every day removes almost the entire queue. The reviewer spends a quarter of an hour a day; the chain gains days.
A third intervention takes a little more nerve: working in parallel. Once the main dimensions and the critical purchased parts are fixed, work preparation can set the long-lead items in motion while engineering finishes the rest. It feels risky, because a late change can then cost money. So limit it to items with a small chance of change and a long delivery time; that is where the chain wins whole weeks.
Handovers and search time: information that does not travel
Information gets lost at every handover. Sales knows why the customer wants that one exception, but it lives in a phone note. Engineering knows which parts are critical for delivery, but work preparation only hears it once the purchase order turns out to be late. The result: people reconstruct information that already existed. It feels like work, but it is waste.
Search time is the invisible variant. Finding the right revision, digging up a similar earlier project, working out which of the three folders with "final" in the name is actually final. Each instance is small, but it happens dozens of times a week. In What does repetitive work really cost your engineering team? we ran the numbers on how those small actions add up to full working days every month.
The remedy is boring but effective: one order file with a fixed structure, the same folder layout for every project, and metadata that is in order, so status and revision live on the document itself instead of in someone's head. Boring, yes. But every minute nobody spends searching is lead time.
The tail of the chain: publishing and exporting
Then the final phase, and it is structurally underestimated: publishing. All production drawings to PDF for the shop floor, DWG or DXF for the sheet metal supplier, a STEP for the customer, the BOM to Excel for purchasing. Do that file by file and an order of forty drawings easily swallows an afternoon. It is risky work too: skip that one drawing, or send an outdated revision, and the mistake travels straight down the chain.
The nasty part is where it sits in the chain: right at the end, when the delivery date is already close. Every hour lost here is an hour of delivery slip. Which is why automation pays back fastest at this spot. With Batch Publish you convert all drawings of an assembly in bulk to PDF, DWG, DXF and STEP, working together with Vault, and with Export BOM the bill of materials lands in your own Excel template, ready for purchasing. The publishing afternoon becomes a coffee break. What that export step looks like in practice, from twelve actions down to one click, is covered in Faster exporting in Autodesk Inventor.
Curious what happens when publishing goes from an afternoon to minutes? Try the Thundercad tools in your own Inventor environment, on your own orders.
Try 30 days freeStep by step: a shorter chain in six moves
You do not shorten lead time with a memo, but with a handful of concrete interventions in the right order. This plan works for a one-person department and for a team of ten.
- Measure one order. Note start and end dates per phase and make wait time visible next to work time. It costs next to nothing and gives you your hardest evidence.
- Complete the intake. Create a fixed checklist for the sales-to-engineering handover and only accept complete orders into the planning.
- Bundle decisions. One open-items list per order, with an owner and a date per item, instead of loose emails that each bring their own waiting time.
- Give reviews a fixed slot. A daily review quarter-hour takes days of waiting out of the chain for the price of fifteen minutes a day.
- Automate the tail. Hand exporting, printing and BOM output over to tooling: it is the most repetitive and most error-prone part of the chain.
- Measure again and lock it in. Repeat the measurement after a few orders, show the difference and write the new routine down, or the old habits creep back.
The pitfall is trying everything at once. Pick the phase where your days actually sit and start there. Usually that is the intake or the tail, because those two touch every single order.
Frequently asked questions
Is halving lead time realistic?
For many departments it is, because most of the lead time consists of waiting rather than working. Remove the waits and automate the publishing tail and the chain shrinks considerably without anyone drawing faster. How much is in it for you is what the measurement from step one will tell.
Where do I start without any data or time tracking?
With one order and a simple list: the start and end date of every phase. Keeping it up takes a few minutes a week and gives you a reliable picture of the biggest delays within a single order. Then pick your first intervention accordingly.
What does automating the publishing step deliver in practice?
Run it with your own numbers: forty drawings exported one by one versus the same forty in a single bulk run. The difference quickly amounts to half a day per order, right when the delivery date is pressing. You can test it without risk: try Thundercad free for 30 days on a real order.