AI process redesign is where the return sits. Automation shortens the work; most processes lose their days to waiting. Learn how to tell them apart.
Everyone says AI saves time. Fewer people ask which time it saves.
That distinction matters because most organisations measure productivity at the task level while customers experience performance at the process level. AI can write, classify, summarise, and respond in seconds, yet a customer request can still spend days sitting between teams, waiting for an approval, or returning for a correction. The work became faster, but the outcome did not.
Everyone says AI saves time. Fewer people ask which time it saves.
That distinction matters because most organisations measure productivity at the task level while customers experience performance at the process level. AI can write, classify, summarise, and respond in seconds, yet a customer request can still spend days sitting between teams, waiting for an approval, or returning for a correction. The work became faster, but the outcome did not.
This is the mistake many AI projects make. They optimise the visible work without redesigning the invisible waiting around it. The result is impressive demonstrations, enthusiastic teams, and dashboards showing minutes saved, while the number the business actually cares about, cycle time, turnaround time, or customer response time, barely moves. AI process redesign begins by measuring both clocks before deciding where AI belongs.
This is the mistake many AI projects make. They optimise the visible work without redesigning the invisible waiting around it. The result is impressive demonstrations, enthusiastic teams, and dashboards showing minutes saved, while the number the business actually cares about, cycle time, turnaround time, or customer response time, barely moves. AI process redesign begins by measuring both clocks before deciding where AI belongs.
BCG's fourth annual AI at Work survey, published 3 June 2026 from 11,749 workers across 14 markets, found that 42% of frontline employees who use AI regularly save at least a full working day every week. Frontline means white-collar staff without managerial responsibility; regular means daily or several times a week. It is self-reported, so treat it as direction rather than measurement.
BCG's fourth annual AI at Work survey, published 3 June 2026 from 11,749 workers across 14 markets, found that 42% of frontline employees who use AI regularly save at least a full working day every week. Frontline means white-collar staff without managerial responsibility; regular means daily or several times a week. It is self-reported, so treat it as direction rather than measurement.
The same survey found that 66% of the people saving that time get limited or no guidance on what to do with it.
The hours came out of the tasks and went nowhere, because nobody redesigned the process the tasks sit inside. A saving that is never re-planned at the level of the process is a personal result, not a business one.
An hour saved by everyone and planned by no one leaves the organisation exactly where it was.
The same survey found that 66% of the people saving that time get limited or no guidance on what to do with it.
The hours came out of the tasks and went nowhere, because nobody redesigned the process the tasks sit inside. A saving that is never re-planned at the level of the process is a personal result, not a business one.
An hour saved by everyone and planned by no one leaves the organisation exactly where it was.
The first clock is touch time: the minutes somebody spends working on it. The second is elapsed time: request to result, waiting included.
Automation shortens the first clock, which in most processes is the smaller number by a wide margin. The gap between the two is where the days live, and four things usually account for it:
The first clock is touch time: the minutes somebody spends working on it. The second is elapsed time: request to result, waiting included.
Automation shortens the first clock, which in most processes is the smaller number by a wide margin. The gap between the two is where the days live, and four things usually account for it:
None of the four is typing. All four are structure: who does what, in what order, with whose approval, and by when. That is what workflow means, and why a faster typist fixes none of them. McKinsey put it plainly in August, describing the organisations getting a return as those that "fundamentally redesign workflows that are enabled by AI rather than insert AI into existing ones."
None of the four is typing. All four are structure: who does what, in what order, with whose approval, and by when. That is what workflow means, and why a faster typist fixes none of them. McKinsey put it plainly in August, describing the organisations getting a return as those that "fundamentally redesign workflows that are enabled by AI rather than insert AI into existing ones."
Automating a bad process, the saying goes, only makes the badness happen faster. That is close, and the precision matters, because it is the whole difference between automation and AI process redesign. Speeding up a step is harmless when that step is the constraint, and harmful when it feeds a constrained one, because you are then filling a queue faster than it drains.
Automating a bad process, the saying goes, only makes the badness happen faster. That is close, and the precision matters, because it is the whole difference between automation and AI process redesign. Speeding up a step is harmless when that step is the constraint, and harmful when it feeds a constrained one, because you are then filling a queue faster than it drains.
The invoice automation did not save four hours a week. It moved them into the approval queue, where they stopped being work and became waiting. Waiting is invisible in a way work is not. No timesheet shows it, no dashboard counts it, and it belongs to the customer. That is why the project was reported as a success while the number did not move.
The invoice automation did not save four hours a week. It moved them into the approval queue, where they stopped being work and became waiting. Waiting is invisible in a way work is not. No timesheet shows it, no dashboard counts it, and it belongs to the customer. That is why the project was reported as a success while the number did not move.
Once you can see the gap, the redesign has somewhere to go. In our own Align sessions the argument worth having is almost never about which tool to buy, but about which of these three a team will change.
Once you can see the gap, the redesign has somewhere to go. In our own Align sessions the argument worth having is almost never about which tool to buy, but about which of these three a team will change.
This is the work the Align stage of our AI Success Workshop is built around. Vision Academy helps organisations answer three questions in order, how to start, how to implement and how to succeed. This post is about the first. Teams bring their own candidate processes and score them on the AI Impact Matrix, impact against feasibility, until the room can repeat one prioritised list back without arguing. It sits one level below the three decisions a leader cannot delegate: that post asked who decides, this one what they are looking at.
This is the work the Align stage of our AI Success Workshop is built around. Vision Academy helps organisations answer three questions in order, how to start, how to implement and how to succeed. This post is about the first. Teams bring their own candidate processes and score them on the AI Impact Matrix, impact against feasibility, until the room can repeat one prioritised list back without arguing. It sits one level below the three decisions a leader cannot delegate: that post asked who decides, this one what they are looking at.
Twenty minutes, no budget, no vendor. Take the process your AI project is aimed at.
Twenty minutes, no budget, no vendor. Take the process your AI project is aimed at.
If the gap is small, you have a good automation candidate. If it is large, the tool was never going to be the thing that moved the number, and you have saved yourself a pilot that everyone except the customer would have called a success.
If the gap is small, you have a good automation candidate. If it is large, the tool was never going to be the thing that moved the number, and you have saved yourself a pilot that everyone except the customer would have called a success.
The organisations getting a return from AI are not the ones with better tools. They are the ones who saw that most of a process's duration was waiting rather than working, changed the structure that produced the waiting, and put AI where it made the new structure possible. That is AI process redesign. It starts with the waiting, not the work, and another pilot will not make that decision easier.
The Align stage of the AI Success Workshop works your own candidate processes down to a prioritised AI Impact Matrix: the ranked shortlist, first process circled, the number it should move beside it. Workshops run across Asia, with dates by region. If you would rather start smaller, we wrote about which strength to point AI at first, the same question from the other end.
The organisations getting a return from AI are not the ones with better tools. They are the ones who saw that most of a process's duration was waiting rather than working, changed the structure that produced the waiting, and put AI where it made the new structure possible. That is AI process redesign. It starts with the waiting, not the work, and another pilot will not make that decision easier.
The Align stage of the AI Success Workshop works your own candidate processes down to a prioritised AI Impact Matrix: the ranked shortlist, first process circled, the number it should move beside it. Workshops run across Asia, with dates by region. If you would rather start smaller, we wrote about which strength to point AI at first, the same question from the other end.
Copyright © 2026 Vision Academy. All rights reserved.
