Designed Without the Doers: What the Exclusion of the Frontline Reveals About How Organisations Change
Exclude it and you are not being decisive; you are working blind and calling it strategy.
Executive Summary
Every large transformation carries an unexamined assumption: that the work can be redesigned by people who do not do it. The programme is drawn up by a steering group and a team of advisers, translated into a target operating model, and then handed down to the people at the counter, the desk and the line to be implemented. This essay is a reflection on that habit — where it comes from, why it persists across eras and industries, and what its persistence tells us about the nature of organisational change itself.
The argument moves through several turns. It begins with the oldest instinct in industrial management, the separation of the thinking from the doing, and traces how each generation of reform — the quality movement, lean production, reengineering — has both attacked that separation and quietly reproduced it. It takes seriously the case for designing without the frontline, because that case is stronger than its critics admit: the people closest to the work are not always its best reformers. But it argues that the knowledge held on the floor is not an optional input to be consulted or skipped; it is the part of the map the plan cannot draw for itself. When a transformation excludes the doers, it does not merely risk poor morale. It proceeds on a false picture of its own subject.
The deeper point is a mirror. How an organisation transforms reveals what it truly believes about where competence lives. An enterprise that treats its frontline as an implementation surface — something change is applied to — has already told you it thinks the work is simple and the workers interchangeable. It is usually wrong on both counts, and the transformation is where that error comes due.
The report was green
There is a particular kind of silence that settles over a place of work while a transformation is being planned somewhere above it. The people at the counter keep serving. The clerks keep clearing the queue. Nothing on the floor announces that anything is coming, and for a long while the only evidence of the programme is the occasional visitor with a lanyard and a clipboard, asking to observe for an afternoon and then leaving.
Meanwhile, three floors up, the picture is confident. The plan has a name, a logo and a timeline. The steering committee sees a status report each month, and for many months that report is green. The design is progressing. The new operating model has been signed off. The system build is on schedule. Everyone who reviews the programme is looking at the same set of indicators, and the indicators say the change is under control.
Then the new process goes live, and the queue does not move. What had taken a practised clerk ninety seconds now takes six minutes, because the redesigned screen asks for information in an order nobody collects it in, and refuses to proceed without a field that, in half of all real cases, does not yet exist at the point of entry. The workarounds begin within a week. By the second month there is a shadow system — a spreadsheet, a shared folder, a set of informal rules passed between colleagues — running quietly alongside the official one, doing the work the official one cannot. The report is still green, because the report was never measuring the work. It was measuring the plan.
I have watched this sequence enough times to believe it is not an accident of execution but a feature of design. The programme did not fail because the people resisted it. It failed because it was built on a description of the work supplied by everyone except the people who do it.
A very old habit
The instinct to separate those who design work from those who perform it is not a modern managerial vice. It is close to the founding idea of industrial management. When scientific management proposed, a century ago, that there was one best way to perform any task and that it was the job of a specialised planning department to determine it, the frontline worker was recast from a source of method into an executor of method. The knowledge was to be gathered up, studied, systematised and handed back as instruction. The worker’s own understanding of the work was treated as unreliable folk knowledge — the very thing the new science was meant to replace.
That settlement has never really been overturned; it has only been periodically embarrassed. Each major reform movement of the last several decades has, in part, been a rediscovery of the intelligence on the floor — followed, more often than not, by a fresh way of ignoring it.
The quality movement was the first great correction. Its insistence that the person doing the job is the person best placed to see how it fails, that variation must be understood at its source, that fear must be driven out so that problems can surface rather than hide — all of this was a direct rebuke to the planning-department model. The most striking symbol of the philosophy that grew out of it is the cord above the assembly line that any worker may pull to halt production the moment something is wrong. It is a formal declaration that the person at the point of work holds authority the plan does not: the authority to say this is not right and to stop the whole line until it is. An organisation that trusts its frontline to stop the line is making a profound statement about where it believes competence resides.
Lean thinking carried that conviction further, elevating the place where the work actually happens into a principle. Go and see for yourself. Do not manage the work from a description of the work. Respect for the people who perform a process is treated not as a courtesy but as a source of competitive advantage, because they are the ones who can see the waste that no process map records.
And yet the habit reasserts itself, often in the very movements that set out to break it. Reengineering, in its purest form, arrived with an explicit instruction to disregard how the work was currently done — to wipe the slate, start from the outcome, and redesign the process from first principles without being anchored by the messy reality of the existing one. There was real insight in that: incremental tinkering with a broken process often only entrenches it. But the method contained, in its own logic, a licence to exclude. If the current way of working is to be treated as an obstacle to imagination rather than a repository of hard-won knowledge, then the people who embody that current way become obstacles too. It is telling that the most honest voices of that movement later acknowledged they had underweighted the human dimension — that they had built a powerful engine for redesigning processes and forgotten that processes are performed by people who know things the redesign did not.
Each generation of reform rediscovers that the people who do the work understand it better than the plan does — and then finds a new reason to design around them anyway. The pattern is not ignorance. It is a habit that keeps surviving its own refutation.
We are, in this respect, remarkably poor students of our own history. The lesson has been taught by every serious school of management thought for the better part of a century. It does not stick, and the reason it does not stick is worth understanding, because it is not stupidity.
Why the habit persists
If excluding the frontline were simply a mistake, it would have been corrected long ago. It persists because, at the moment of decision, it is the easier and more comfortable path in several genuine and reinforcing ways.
The first is speed. Involving the people who do the work is slow. There are a great many of them, they are busy doing the work, and their knowledge is not written down anywhere; extracting it means sitting with them, watching, asking, and listening to things that do not fit the categories the programme has already established. A design team under pressure to show progress finds it far quicker to interview a handful of managers, read the procedure manuals, and proceed. The procedure manuals describe the work as it is supposed to be done, which is exactly the version that is easiest to redesign and least like the truth.
The second is legibility. Senior decision-makers need a picture of the organisation they can hold in their heads and act upon. The frontline reality — the exceptions, the local knowledge, the informal arrangements that make the official process actually function — is precisely the detail that a simplified, governable picture must strip out. There is a deep and permanent tension here between the view required to run a large organisation and the view required to understand any particular piece of its work. Transformation is usually designed from the first view, and performed in the second.
The third is the operating model of change itself. Large transformations are frequently delivered by teams — internal programme functions, external advisers — whose expertise is precisely that they are not embedded in the work. Their value proposition is the outside eye, the cross-industry pattern, the methodology. This is genuinely useful, but it carries an occupational bias: a group whose authority rests on general method is not naturally inclined to defer to specific, local, unmethodical knowledge. The way the programme is staffed quietly predetermines whose knowledge counts.
And the fourth, least often admitted, is that opening the box is uncomfortable. To ask the people who do the work how it really gets done is to invite answers the programme may not want. It is to learn that the elegant process on the slide depends, in practice, on three experienced individuals who know how to handle the cases nobody documented. It is to discover that the efficiency the business case promised assumes away the very complexity the frontline spends its day absorbing. It is far more comfortable to design the future than to look too closely at the present.
“The procedure manual describes the work as it is supposed to be done — which is exactly the version that is easiest to redesign and least like the truth.”
None of these forces is foolish. Each is a rational response to a real constraint. Together they explain why intelligent, experienced people, who know in the abstract that the frontline holds essential knowledge, will nonetheless design a transformation without it — and be surprised, every time, when the plan meets the work.
The strongest case for designing without them
It would be too easy to end there, as though the only honest position were reverence for the wisdom of the floor. The case for designing transformation without heavy frontline involvement is stronger than its critics usually allow, and it deserves to be put at its full strength rather than as a straw man.
Begin with the uncomfortable truth that the people closest to a process are not always its best reformers. Deep familiarity breeds a particular blindness: what is habitual becomes invisible, and practices that made sense under conditions that no longer hold are defended precisely because they are familiar. Ask a team steeped in a way of working to reimagine it and you will often get the current process with its worst irritations sanded off — an optimised version of the wrong thing. Genuine reinvention frequently does require someone who is not captured by the existing arrangement.
There is a second, harder point. The frontline sees its own part of the work with great clarity and the whole system hardly at all. A change that is locally sensible — that makes one team’s day easier — may be systemically damaging, shifting cost or delay onto a team downstream that the first never sees. Optimising each station for the people who staff it does not produce an optimised flow; it can produce the opposite. Someone must hold the view of the whole, and that view is not available from any single point on the line.
Third, involvement has a real cost and a real risk. Consultation raises expectations, and expectations that are raised and then disappointed do more damage than never having asked. A transformation that promises the frontline a voice and then overrules it — as, holding the view of the whole, it sometimes must — can poison the very goodwill it sought to build. There are moments when a leadership team, having genuinely weighed the local knowledge, must still make a call the frontline will dislike, and pretending otherwise is its own form of dishonesty.
“The frontline sees its own part of the work with great clarity and the whole system hardly at all.”
This is a serious case, and any honest treatment of the subject has to hold it. But notice what it actually establishes. It establishes that the frontline should not have a veto, that local knowledge is not the same as system judgement, and that involvement must be designed rather than declared. It does not establish that the doers can be safely designed around. Every one of these objections is an argument about how to weigh frontline knowledge against other considerations — which presupposes that the knowledge is in the room to be weighed in the first place. The failure this essay describes is not the failure to let the frontline decide. It is the failure to let the frontline inform. Those are entirely different things, and conflating them is how the strong case for outside judgement gets quietly stretched into a licence for ignorance.
What the doers know that the plan cannot
To see why the frontline’s knowledge cannot simply be gathered from managers and manuals, it helps to be concrete about what that knowledge actually is.
Consider a consolidation of finance operations into a single shared service — the kind of programme that was, and remains, a staple of the efficiency agenda. The logic is clean: many scattered teams doing broadly the same work, brought together, standardised onto one process and one system, run at lower cost. The design team maps the standard process. It identifies, let us say, twelve standard invoice types, defines a single streamlined handling procedure, and models the savings on the assumption that the great majority of volume flows down this clean path with only a small tail of exceptions.
The people who actually process the invoices could have told them the model was wrong before it was built. In the reality of the work there were not twelve invoice types but something closer to forty, and while most were indeed straightforward, a stubborn minority — perhaps one in eight — carried the bulk of the value and every one of the genuine complications: the foreign-currency cases, the ones tied to contracts with non-standard payment terms, the disputes that had to be handled without alienating a major supplier. None of these appeared on the process map, because the map had been built from the procedure manual and a handful of manager interviews, and the manual describes the tidy majority. The complexity lived in the hands of a small number of experienced clerks who had absorbed it over years and who handled it, largely, by knowing what the rules did not say.
| The work as described | The work as done |
|---|---|
| Twelve invoice types on a single standard path | Roughly forty variants, with a critical minority carrying most of the value |
| Exceptions are a small, low-value tail | The “exceptions” are where the money and the risk actually sit |
| Knowledge lives in the procedure manual | Knowledge lives in the hands of a few experienced people |
| Standardisation removes waste | Standardisation removes the accommodations that made the work possible |
When the new operating model went live, the standard path worked as designed and the critical minority fell apart. The experienced clerks, in the logic of the consolidation, had been the most obvious candidates for redundancy — their roles looked, on paper, like exactly the manual work the new model would absorb. With them went the undocumented knowledge, and the tidy process had no way to handle the cases that had never been on the map. Processing times for the complex cases lengthened severalfold; a backlog built; the organisation found itself paying premium rates for temporary staff and, in some cases, quietly re-hiring the very people it had let go, now as contractors, to keep the supplier relationships from breaking. The projected savings did not so much fail to materialise as invert.
The point of the example is not that consolidation is wrong, or that savings are illusory. It is that the knowledge which would have corrected the design existed, in full, inside the organisation — and the way the transformation was conducted guaranteed it would not be heard until it arrived in the form of a crisis. What the doers knew was not an opinion about the change. It was the actual shape of the work, and no quantity of manager interviews or manuals could substitute for it, because the manuals recorded the intended work and the managers, one level removed, believed the manuals.
This is the crux. The frontline does not merely have feelings about the transformation that a considerate leader might wish to acknowledge. It holds the empirical description of the present that any sound design of the future must start from. Exclude it and you are not being decisive; you are working blind and calling it strategy.
Participation is not a referendum
If the answer were simply “involve everyone,” the problem would have solved itself long ago, because everyone agrees with the sentiment. The reason the sentiment fails to translate is that genuine involvement is confused, in both directions, with things it is not.
It is not the same as consultation theatre — the roadshow, the town hall, the feedback form circulated after the decisions have been taken, the workshop whose conclusions were written before it began. People who do the work can tell the difference between being asked and being informed almost instantly, and a pretence of involvement is worse than an honest exclusion, because it spends trust and returns nothing. Much of what passes for engagement in transformation is of this kind: a communications exercise wearing the costume of participation.
But genuine involvement is also not a referendum. It does not mean the frontline decides, or holds a veto, or that every local preference must be honoured. The steelman was right about that. The task is more discriminating: to bring the knowledge of the work into the design as evidence, to weigh it honestly against the view of the whole that the frontline cannot see, and to be transparent about the trade-offs when they cut against local preference. That is a harder and more respectful thing than either ignoring the floor or deferring to it.
In practice the distinction tends to come down to a few honest questions a programme can ask of itself:
- Did we build our picture of the current work from the people who perform it, or from the documents that describe how it is supposed to be performed?
- When the frontline’s account contradicted the plan, did the plan change, or did we explain why the frontline was mistaken?
- Have we identified where the real complexity and undocumented knowledge actually sit — and protected it, rather than treating it as the most obvious cost to cut?
- When we overrule local knowledge in favour of the whole, can we say so plainly, and would the trade-off survive being explained to the people it affects?
An organisation that can answer those questions honestly is not one that has surrendered its judgement to the floor. It is one that has stopped confusing a plan with the work — which is the specific confusion at the root of the pattern.
What it reveals about how organisations change
Return, finally, to the question the topic actually poses: not merely how to run a better transformation, but what the exclusion of the doers reveals about how organisations change.
It reveals, first, that we consistently mistake the map for the territory. A transformation is conducted through documents — models, plans, business cases, status reports — and it is fatally easy for the people conducting it to begin treating the documents as the reality rather than as a lossy representation of it. The frontline is excluded not usually out of contempt but out of this quieter error: the belief that the work is adequately captured in its description, so that whoever holds the description holds the work. The green status report over the stalling floor is the emblem of the whole pattern. The instruments were measuring the plan, and the plan was mistaken for the thing.
It reveals, second, what an organisation truly believes about where competence lives. Every design choice about a transformation is also a statement of belief. To design change for the frontline rather than with it is to have decided, before any evidence is gathered, that the intelligence in the enterprise sits above the work and the labour sits below it — that the doers are hands to be redirected rather than minds to be consulted. Organisations rarely say this aloud. They do not need to. The way they conduct their transformations says it for them, and the people on the floor hear it clearly, whatever the communications plan announces.
And it reveals, third, something about the nature of durable change. The transformations that hold are not, in my experience, the ones with the most elegant target operating models. They are the ones where the people who do the work recognised themselves in the design — where the new way accommodated the real complexity rather than assuming it away, because the real complexity had been in the room while the design was made. Change that is done to people is endured, worked around, and quietly outlasted; the shadow spreadsheet is patient. Change that is done with them can actually take, because it was built on the truth of the work rather than on a description of it.
An organisation that treats its frontline as a surface to which change is applied has already told you what it believes: that the work is simple and the workers interchangeable. The transformation is where that belief is tested — and, usually, where it is found wrong.
The habit of designing without the doers will not disappear, because the forces that sustain it — the pressure for speed, the need for a legible picture, the outside-in operating model of change, the discomfort of looking too closely at the present — are permanent features of large organisations, not passing failures of a particular era. That is precisely why the pattern recurs across the decades and across every wave of reform. But naming it accurately is not nothing. The exclusion of the frontline is not a soft, human-relations footnote to the hard work of transformation. It is a factual error at the centre of it — a decision to redesign a reality on the strength of a description, while the only people who hold the reality stand a few floors below, keeping the work running by means the plan will never see, waiting to be asked.