This is the fourth in a series exploring each of the 5 Lenses, the framework I use to help boards and leaders see more of the system when navigating complex change. Last month, I explored the Power lens and what it means when the people with formal authority do not have the capacity to use it. The Power lens asks who can actually move the problem, and what is getting in their way. This month, I turn to the Process lens, because once the Power lens has shown us where the gaps are, the next question is almost always the same: what in the way the work is actually designed is creating those conditions?
Most organisations do not think of their processes as decisions. They think of them as infrastructure. The approval pathway, the reporting cycle, the way referrals are managed, the point at which a frontline worker escalates to a manager: these feel like background, like plumbing. They exist so that the real work can happen.
The Process lens challenges that assumption directly. It treats every process as a set of choices, even when nobody made them consciously. A process that grew through habit and history still decides who checks what, who is consulted, and who picks up the rework when something goes wrong. The difference between a deliberately designed process and one that simply evolved is not that one has choices and the other does not. It is that only one of them has an owner.
This matters most in the kinds of organisations I work with, where processes are rarely designed from scratch. They grow. A founding CEO does something once because it seemed sensible. It becomes the way things are done. Someone joins and learns the pattern without questioning it. By the time the organisation has twenty people, the founding logic is buried, and so is the weight it was carrying all along.
The Process lens does not ask whether your processes are documented. It asks who ends up carrying what, and whether that was a conscious choice.
A few years ago, a charity I work with ran a digital inclusion project and bought a small bank of tablets to lend out to the people they supported. A good idea, well intentioned, genuinely useful. But nobody ever designed a process for what came next: the bookings, the charging, the tracking of who had which device and when it was coming back.
So the person who had championed the project kept doing it. She fielded the requests. She knew which tablet had the cracked case that needed mentioning. She sent the reminder messages. I found out recently, she had been quietly running what amounted to a lending library service for years, entirely informally, on top of her actual job.
The detail that stays with me: she is, by her own cheerful admission, the least technically confident person in the organisation. She did not keep doing it because she was the obvious choice. She kept doing it because she cared, and because there was no handover process to pass it on to anyone else.
This is not a failure of goodwill. The organisation was full of committed people. It is a failure of design. Nobody wrote a process for what happened after the project ended. So the task stayed with the person who started it, quietly, indefinitely.
John Seddon calls this failure demand: the work generated not by genuine need but by the failure of a system to function as intended. When a process does not work as designed, someone still has to deal with the consequences. That work goes somewhere. In most organisations, it finds the people who are most conscientious and least able to redirect it.
The first thing the Process lens asks you to do is map that gap. Not by commissioning a review, but by asking the people closest to the work what they are carrying that nobody officially asked them to.
The strongest academic anchor for this lens is Sharon Parker and Caroline Knight's SMART Work Design model, published in Human Resource Management in 2024 and written up for practitioners in MIT Sloan Management Review the same year. SMART stands for Stimulating, Mastery, Autonomous, Relational, and Tolerable demands: five dimensions of work quality validated across three studies involving more than 1,800 workers.
The dimension most relevant here is the last one: tolerable demands. Parker and Knight define this as the absence of overload, role conflict, and work-home interference. Intolerable demands arise when the effort required to achieve system-level goals exceeds what a reasonable person can sustain. The research is clear that when demands tip into that territory, the consequence is not just burnout in the individual. It is a structural signal that goals and available resources are out of alignment, and that somebody has been quietly making up the difference.
The autonomy dimension adds a second layer. Parker's model shows that autonomy functions as a multiplier: when it is absent, the motivating potential of even deeply meaningful work collapses. A frontline worker who cares deeply about the people they serve, but who has no control over how they do their work and no feedback on whether it is working, will not find more purpose. They will find more frustration. The structure is making the conditions. The person is living in them.
Taken together, these findings make a simple point for the Process lens: when structure and resources are misaligned, process silently turns that misalignment into intolerable demands for specific people.
A 2025 paper in the *Journal of Business and Psychology* adds something important about the limits of individual agency. Job crafting, the idea that workers can reshape their own roles to reduce friction and add meaning, only functions when the top-down structure permits it. In tightly controlled or severely under-resourced environments, the structural conditions defeat the individual effort before it starts. Telling people to find their own solutions to structural problems is not empowerment. It is delegation of an organisation's design failures downward. People cannot craft their way out of a broken process.
The sector context sharpens this. The 2025 Charity Digital Skills Report, produced by Zoe Amar and Nissa Ramsay, found that 76% of charities are now using AI tools, up from 61% the previous year. Only 2% are using AI strategically. More striking still, the proportion of charities with a digital strategy fell from 50% to 44% in the same period, while the proportion of boards with poor digital skills rose by 11 percentage points.
The pattern suggests that parts of the sector are deploying tools they cannot govern into workflows they have not documented. That is not a technology problem. It is a governance problem that technology is making more consequential.
Seen through the Process lens, the same system looks different depending on where you sit.
At board level, process tends to appear as policy. The board sees the documented version: the safeguarding framework, the HR procedures, the financial controls. What it rarely sees is the operational reality: who is doing the informal work that sits around the edges of those policies, what happens when a documented process meets a genuine edge case, and how much is being held together below the level at which the board receives its reports.
This is not a criticism of boards. It reflects a structural gap in most governance arrangements: the information that reaches the board has been curated, summarised, and presented by people with an interest in it looking coherent. The Process lens does not ask boards to distrust their executives. It asks them to add a different kind of question to their repertoire. Not "is the process in place?" but "who is carrying it, and can they sustain that?"
A risk register updated quarterly because it has to be is not the same as a risk register that reflects what the organisation is genuinely worried about. A process that looks robust on paper but is held together informally by one or two people who have simply never stopped is not robust. The board cannot always see the difference from the papers in front of it, but it can ask.
The more useful question is: if this process broke down tomorrow, who would notice first, and how quickly would it reach us?
At manager level, the Process lens becomes immediate and often uncomfortable. Managers are the people who can most clearly see the gap between what the process says and what actually happens. They know who the informal coordinators are. They know which approval pathway always gets stuck at the same point. They know which reporting requirement generates three hours of rework for information that nobody reads.
The failure mode here is covering for the dysfunction rather than naming it. A manager who quietly patches a broken process, who stays late to chase what the system should have caught, who corrects the output before it goes up, is protecting the organisation in the short term and making the problem harder to see in the medium term. The process continues to look functional because the manager is compensating for it.
The better question is not "how do I keep this moving?" but "what would happen if I stopped compensating, and is that information the organisation needs?"
At individual and frontline level, the Process lens is lived experience. It is the point at which decisions that were well-considered, or never really considered at all, arrive in a person's working day. The question is not whether they understand the process, but whether it was designed around how the work actually flows, or around how the organisation wishes it flowed.
Research from the US nonprofit sector on staff burnout makes the point plainly: burnout does not stem simply from heavy workload. It stems from carrying that workload without being given a voice, recognition, or opportunity (Candid, 2024). The patterns are directly recognisable in UK charities. The same volume of work inside a well-designed process, with clear ownership and genuine autonomy, is experienced very differently from the same volume inside one that was never really designed at all. The weight is similar. The sense of meaning and agency is not.
It is also worth saying directly: when organisations decide to review their processes, that review can easily become another thing the same overstretched people are asked to do. Asking someone to map the gap between documented and actual practice, while they are still carrying the undocumented work, is its own form of process failure. The review needs an owner who is not already buried. If there is no such person, that is itself data from the Process lens.
When organisations talk about improving their processes, they almost always mean making them faster, simpler, or more consistent. Efficiency is the frame. The question is: how do we get the same outcome with less friction?
The Process lens asks a different question: who is currently carrying the friction, and what happens to them when we remove it?
This reframe is not anti-efficiency. Reducing waste is genuinely valuable. But efficiency improvements that are not accompanied by an honest map of where the current weight sits will often just redistribute it without anyone noticing. The person who has been informally carrying a task for years does not automatically stop when a new process is introduced. They may carry on, because the new process has a gap they can see, and they still care.
It is also worth naming who benefits from the current arrangement. Process drift is rarely neutral. The person not redesigning the process is often the person who would have to absorb the cost of the redesign, whether in time, political capital, or disruption to a system that currently works well enough from where they sit. That is not bad faith. It is a rational response to an unclear ask. But it is worth seeing clearly, because it means process improvement rarely happens by itself. Someone has to decide it is worth the short-term cost.
W. Edwards Deming, whose thinking underpins much of modern quality improvement, argued that the overwhelming majority of failures in an organisation are caused by the system, its processes, structure, and practices, rather than by individual lack of effort or care. The precise figures he cited, often quoted as 85% systemic to 15% individual, are contested, but the directional point has held up across decades of research: most of what goes wrong is structural, not personal. And yet the instinctive response to process failure is almost always to focus on the individual: the person who did not follow the procedure, the team that missed the deadline, the manager who did not escalate in time.
This inversion is not accidental. It is faster to tell someone to do better than to ask whether the conditions make doing better possible. The Process lens interrupts that instinct. It asks whether the system is generating the outcome before it asks whether the individual is responsible for it.
The same logic applies to AI. The current pattern in the charity sector, layering AI tools onto manual, siloed, undocumented workflows, is not inherently wrong. Tools can help. But if the process was producing the wrong output before the tool arrived, the tool will produce the same wrong output more quickly and more consistently. Failure demand does not disappear when you automate it. It becomes faster, harder to see, and harder to question. If nobody had the standing to challenge the process before, automation rarely creates that standing. It usually removes the friction that made the problem visible. Contestability, the ability to reverse and challenge outcomes, needs to be designed in before deployment, not retrofitted after something goes wrong.
Before turning to the practice questions, it is worth being concrete about what redesigning a process well actually involves, because without a sense of what the alternative looks like, the diagnosis alone can feel overwhelming.
I have seen organisations do this well in a single afternoon. Not by commissioning a consultant or running a restructure, but by gathering the four people who actually do the work in question, asking them to walk through what they each do step by step, and marking anything that has no named owner or no clear trigger. This works in the right conditions: when the right people are in the room, when someone with the authority to act on what they find is present, and when the organisation is genuinely ready to hear the answer. In one case, that conversation identified three tasks that had been sitting with the wrong person for over two years, none of which anyone had flagged because each person assumed someone else had agreed to it.
The redesign was not complicated. It was a shared document, a named owner for each step, and an agreement that if the named owner left, the handover was part of the offboarding checklist. The whole conversation took ninety minutes. The value was not in the document. It was in the fact that for the first time, the people doing the work had been asked.
That is where most process improvement starts: not with a framework, but with a question directed at the right people.
These questions are not a checklist. They are a way of looking at processes you already know, with a different frame.
If you wanted to start somewhere this week, pick the process in your organisation that most depends on one person's institutional memory. Ask them what would happen if they were not there. The answer is usually the beginning of the conversation that should have happened years ago.
The Process lens alone will not resolve unclear authority or rebuild trust. But it is the lens that makes visible what the other lenses can only point to. Purpose tells you why the work matters. People tell you how it lands on individuals. Power tells you who can move it. Process tells you how the load is actually distributed, regardless of what was intended.
If you are carrying a problem that nobody seems to own, look at the process that surrounds it. Not the policy. The actual sequence of events, and who is doing what in the gaps.
The most important question this lens asks is not retrospective. It is not "who decided this?" It is: now that you can see it, what will you do differently?
Next month, I will turn to the Practice lens: how organisations learn, or fail to, from what the work is showing them, and whether the systems for reflection and adaptation are designed to surface the truth or to protect the story.
Candid (2024). Data Shows Why Frontline Nonprofit Workers Face Burnout/Turnover. View arthttps://candid.org/blogs/data-shows-why-frontline-nonprofit-workers-face-burnout-turnover/icle
Parker, S.K. and Knight, C. (2024). The SMART model of work design: A higher order structure to help see the wood from the trees. Human Resource Management, 63(2), pp. 265–291. View paper
Parker, S.K. and Knight, C. (2024). Design Work to Prevent Burnout. MIT Sloan Management Review, Winter 2024. View article
Amar, Z. and Ramsay, N. (2025). Charity Digital Skills Report 2025. Charity Digital Skills. View report
Seddon, J. (2008). *Systems Thinking in the Public Sector.* Triarchy Press. View book
Springer Nature (2025). Top-Down and Bottom-Up Work Design: A Multilevel Perspective on How Job Crafting and Work Characteristics Interrelate. *Journal of Business and Psychology.* View paper
UK Parliament (2024-25). *Public Authority Algorithmic and Automated Decision-Making Systems Bill [HL].* House of Lords Library. View briefing
This article was co-created through a human-led process using several AI models as thinking partners. It reflects our commitment to ethical, transparent, and accountable use of AI, where human judgement, curiosity, and oversight remain central.