Is this workflow worth engineering?
Not every broken process deserves a build. Some deserve a meeting, a spreadsheet, or honest acceptance. Here is how we read the difference before recommending anything.
4 min read
The only question that matters
Where is valuable work getting stuck, what is that costing the organisation, and can a materially better flow be engineered? Work that is valuable but moves badly is worth attention. Work that moves badly but is not valuable is noise. The two get confused constantly, because stuck work feels expensive to the people living it - the test has to be economic, not emotional.
Five signs the answer is probably yes
1. People move information between systems by hand
Re-keying, copying fields from one tool to another, reconciling two spreadsheets that should agree but don't. Every manual transfer is a place work waits, errs, or gets lost - and it is effort spent moving data rather than using it.
2. An inbox is acting as a queue
If work arrives, waits and gets claimed through a shared mailbox, you have an invisible queue: no owner, no ageing, no view of backlog. The work is real - the process holding it is incidental.
3. Exceptions eat the day
The routine cases could run themselves, but the unpredictable minority consumes the team. Nobody has ever separated which work is routine and which genuinely needs judgement - so everything gets judgement-time.
4. The knowledge lives in one person
"Ask Sandra - she knows how that works." When a process depends on one person's memory, it does not scale, does not survive absence, and cannot be audited. The fix is not documentation alone - it is turning the knowledge into a system.
5. Volume grows only by hiring
If processing more work strictly requires more people, the operating model is linear and margins compress exactly when the business grows. Worth engineering if the volume is real - the economics compound.
And when the honest answer is no
The process changes faster than you could build
If the workflow itself is still being invented, engineer the process first. Automating a moving target buys rigidity, not flow.
The volume is too small
A workflow that stalls twice a month costs irritation, not money. If the stuck work is not valuable or not frequent, the honest answer is leave it alone.
Nobody owns the outcome
Engineered flows need an owner on your side - someone accountable for whether the new way of working is actually better. Without one, even a good build drifts.
If you are not sure
That uncertainty is itself diagnosable, and cheaply. A Workflow Engineering Review looks at one nominated workflow - how it really runs, where it stalls, what the stall plausibly costs - and returns a written engineering assessment. Fixed scope, fixed fee, yours to keep whether or not it leads anywhere. Including when the finding is “leave it alone”.
Short, infrequent pieces on what is actually worth engineering in an operation - and what is not.
Nominate the workflow you keep tripping over.
About the Review