Projects · Personal Growth
Why good design engineers leave, and how to keep them

Engineering companies in a growing economy face a strange squeeze. Order books are full, but experienced design engineers are scarce, and the ones they have keep leaving. The average experience of design teams falls year after year as companies rely more on fresh graduates.
Many engineers move to engineering-services firms, where work is sold by the hour, so workloads are controlled and the pressure is lower. Those who stay are asked to work late, which leads to burnout, which leads to more people leaving.
It looks like an unsolvable problem: too few experienced engineers, too much work. But there is an old saying: if you’re in a hole, stop digging. Most engineering companies are still digging.
Start immediately, finish late
The default rule in most design departments is: start the design as soon as an order arrives. The pressure comes from all sides: long-lead items must be specified early; customers want basic designs to integrate with other systems or to start their own civil work.
With limited people, this rule creates constant switching. When a new order arrives, designers drop what they are doing and start on it, often without complete inputs. To keep moving, they assume the missing information and release initial drawings.
The results are predictable:
- Drawings are revised again and again, five or six revisions is not unusual.
- Detailed design is interrupted while basic designs go round in circles.
- Procurement is delayed, and manufacturing loses capacity.
It is worse where design disciplines depend on each other in both directions. Say systems design produces loads that electrical design needs, but electrical routing affects the cut-outs systems design must provide for, and those can only be finalised once the loads are known. Under time pressure, each department assumes the other’s numbers. The assumptions conflict, and a cascade of iterations follows.
Is it really inexperience?
When errors appear, managers usually blame inexperienced designers. But try this simple exercise.
Take two simple tasks you know perfectly: writing the word MULTITASKING, and writing the numbers 1 to 12.
- Round 1: alternate between them, one character at a time: M, 1, U, 2, L, 3…
- Round 2: write the whole word, then all the numbers.
Time both rounds and count your mistakes. Most people are noticeably slower in Round 1, and many make mistakes, on tasks they could do in their sleep. Each switch forces the mind to rebuild its context, and errors slip in.
Now imagine the same switching on a complex engineering design, where a designer must hold dozens of interrelated variables in mind. Every switch means rebuilding that understanding. Under pressure, some of it isn’t rebuilt, and mistakes follow. For a new designer, it is worse still.
Errors in a multitasking environment aren’t mainly a skill problem. Even experts make them when they keep switching.
Supervision suffers too. The most experienced people in the department spend their time explaining delays, chasing other departments, or doing urgent design work themselves. They have no time left to guide and check others’ work. Mistakes by junior designers go unnoticed until late, and the same mistakes are repeated across projects.
So capacity is wasted three ways:
- Set-up time every time a designer switches.
- Iterations from assumptions made while waiting for information.
- Rework from errors made while switching.
Picture a design department of 20 engineers that releases about 100 drawings a month, but each drawing goes through five revisions on average, three of them caused by missing inputs or errors while switching. Cut those three avoidable revisions, and each drawing needs roughly 40% less design effort. The same 20 engineers could release well over 150 drawings a month, more than enough to absorb a growing order book without hiring.
With several iterations per drawing, the wasted capacity is enormous. But because it is accepted as normal, the blame falls on the order book and the job market.
Stop digging: form a queue
The root assumption is that starting earlier means finishing earlier. With limited people shared across many projects, it doesn’t. It wastes both time and capacity.
Think of how a traffic officer clears a jam: he stops vehicles from all entering at once and lets them through in turn. More vehicles get through, faster, on the same road.
A design department can do the same:
- Limit the number of open designs. In departments with heavy switching, open work can often be cut by half or more without anyone running out of work. The rest wait in a queue.
- One out, one in. A new design enters only when an active one is complete, with clear closure criteria, otherwise the department slides back into chaos.
- Use the released capacity of senior designers to prepare the queue. They check that every waiting design has its full kit of inputs before it starts. For designs that depend on other disciplines, they agree interfaces and design criteria with those teams up front: for example, structural and civil limits agreed during the initial systems design. This cuts iterations sharply.
These two steps (limiting work in progress and insisting on a full kit) can release a large share of a design department’s capacity. Design lead times fall, revisions drop, outsourced design becomes unnecessary, and engineers stop working late every night. And when the pressure eases, fewer of them leave.