Stop automating bad processes.
Automating a broken workflow only makes it fail faster. The larger prize is redesigning the work itself.
There is an old instinct in operations that feels like progress and is usually the opposite: take the process you already have, and make it faster. Automate the steps. Remove the waiting. Push the same work through the same path at greater speed. It is intuitive, it is easy to fund, and it is one of the most reliable ways to entrench a problem you should have removed.
Most workflows inside an established business are archaeological. They are not designed so much as accumulated — layer upon layer of workarounds built around constraints that, in many cases, no longer exist. A step survives because someone added it during a crisis in 2014. An approval exists because of a person who left years ago. A handoff persists because the two systems it bridges were never meant to talk.
Automating a broken process does not eliminate the problem. It industrialises it.
Redesign before you automate
The larger prize is almost always in the redesign, not the automation — and the two are frequently confused. Automation asks how do we make this step faster. Redesign asks a more uncomfortable question: why does this step exist at all? The order of operations is the whole game. Map where value is actually created. Interrogate every handoff, every approval, every re-keying of the same information — and remove what cannot justify itself. Only then automate what remains.
The results of doing this in the right order do not look incremental. A quoting process that took five days and seven manual handoffs does not become a four-day process; redesigned around the outcome, it becomes a two-step workflow that clears in hours.
Why firms skip the hard part
If redesign is where the value is, why do so many firms reach for automation instead? Because automation is a purchase and redesign is a decision. One can be delegated to a vendor; the other requires leadership to confront how the work is really done.
The Stansbury view
Redesign first. Automate what survives. Anything else is paving the cow path — faster, at scale, with a higher error rate.