Let me tell you about Marcus, not his real name, but he’ll know who he is if he reads this. Marcus worked in ops at a mid-sized logistics firm in the Midlands. Every week, he’d submit a report that took him four hours to produce. Not because the data was complicated. Because the internal system was broken, the export function hadn’t worked since 2019, and IT had been looking into it since approximately the last World Cup. So, Marcus built his own, a Google Sheet. Some VLOOKUP formulas held together with hope and a prayer. It produced the same report in twenty minutes; he never told anyone, just got on with it.
Your best people are routing around the system, right now, while you’re reading this over a coffee that’s probably gone cold.
This isn’t a niche problem. Research suggests that roughly 50% of work isn’t supported by enterprise software, which forces employees into manual workarounds and shadow processes. Half of all the actual work being done in organisations today is happening through channels and systems that leadership can’t see, didn’t sanction, and definitely didn’t plan for. And here’s the real stinger: it’s almost always your highest performers doing this.
It’s easy to picture this as dramatic, whistleblowing, sabotage, some kind of corporate rebellion. It’s not; it’s mundane and quiet. It’s a Post-it note on a monitor with someone’s mobile number because the ticketing system takes three days to respond. It’s a WhatsApp group called “ACTUAL project updates” that runs parallel to the official project management tool nobody uses properly. It’s a shared spreadsheet on someone’s personal OneDrive because the approved system is clunky, slow, and requires six clicks to do something that should take one.
Beneath every formal structure lies an unofficial organisation that operates through unwritten rules and invisible networks, complex, dynamic lattices rather than hierarchical ladders. Organisational psychologists have been banging on about this for decades, but the urgency hasn’t always landed with the people who most need to hear it. Founders, senior leaders, entrepreneurs, the people who built the official system often have a blind spot here. They see the organisational charts but fail to see the workarounds holding it together.
I once did a piece of work with a small professional services firm where the ops director, a brilliant woman and very capable, told me with complete confidence that their project delivery process was robust. She showed me the flowchart; it looked great, all laminated and everything. Then I spent a day in the team’s shared office. By lunchtime I’d spotted four separate processes that bore no resemblance to the laminated flowchart whatsoever. There was also a cup of volcanically hot tea from a machine involved, not really tea, is it? But that’s not the point.
The point is: the map is not the territory.
Here’s what makes this psychologically interesting. Workarounds aren’t created by disengaged, half-hearted employees who can’t be bothered to follow the rules. They’re created by people who care enough to get the job done anyway.
Think about what it takes to route around a broken system. You must:
Recognise that the official process is failing (awareness)
Want the outcome badly enough to do extra work (motivation)
Have the capability to design an alternative (competence)
Feel confident enough to make that call unilaterally (agency)
That’s not the profile of someone who doesn’t care. That’s the profile of someone you desperately need to keep.
When workflows are clunky, slow, or unclear, people feel it long before the metrics show it. Your disengaged employees don’t build workarounds; they just wait. They follow the broken process, produce the suboptimal output, then go home. It’s the engaged ones who improvise. And then, ironically, the organisation often has no idea this is happening, because the improvisation works well enough that the bad system never gets flagged as broken
This is becoming more visible, and increasingly urgent right now, partly because of AI. UK workers are embracing AI tools at pace, with more than half of employees now describing their feelings as optimistic, excited, or confident about AI, up from just 34% at the start of 2025. But a lot of that enthusiasm is happening outside official channels.
The EY 2025 Work Reimagined Survey found that 83% of UK employees now use generative AI at work, though this is largely limited to basic, process-driven tasks. And a substantial chunk of that usage is unsanctioned. People aren’t waiting for IT to approve a tool. They’re just using it. Because it works. Because the approved alternative doesn’t exist yet.
Shadow IT, the use of applications and tools employees prefer, rather than what’s officially assigned, accounts for roughly 42% of the average company’s applications. Forty-two per cent. Nearly half the tech stacks your people are using is stuff leadership either doesn’t know about or hasn’t approved. And yet the work gets done. Often better.
There’s a particular organisational blind spot at play here, and it’s worth naming it directly. Senior people tend to experience systems at a low level of friction. They have admin support and override privileges. They also have the political capital to pick up the phone and get something sorted without going through the formal process. The system rarely breaks for them in the same way it breaks for everyone else.
So, when a mid-level manager describes a process as broken, it can land as complaining, resistance to change, even as someone not trying hard enough. When actually it’s a signal. A useful, honest, often quite specific signal that the official route isn’t fit for purpose.
Research has found that when meeting technology and collaboration systems fail, it doesn’t just cause mild annoyance; it undermines wellbeing. Workarounds then show as symptoms. The disease is systems that weren’t built with actual users in mind, or that were built for a version of work that no longer exists.
There’s also a cultural dimension: shadow systems provide speed when official processes are slow, and they can create belonging when the larger organisation feels impersonal. But when left unexamined, they can undermine official policies by favouring insiders over newcomers or certain backgrounds over others. This is the bit that keeps me up at night. Because workarounds that start as useful often calcify into unofficial rules. They become the thing only certain people know, and that information asymmetry is not always benign.
Here’s a scenario. Your best operator, the one who knows how everything actually works, decides to leave. They hand in their notice. Their line manager writes a lovely leaving card. HR does the exit interview, but nobody asks: “Can you show us how you’ve been doing your job?”
Because the answer would reveal that the job as officially described and the job as actually done are completely different things. And that gap represents knowledge that walks out the door on the last Friday of the month, never to return.
Shadow processes, spreadsheet trackers, and side conversations often become the real system of record in organisations, with leaders believing one thing is happening while frontline workers know something different entirely. When the person who holds that real system of record leaves, organisations discover very quickly how much invisible infrastructure they’d been relying on.
I’ve seen this play out more than once. The most painful version was a finance team that lost their longest-serving analyst. She’d been there eleven years and had built fifteen spreadsheets that connected systems which the organisation had been told were already integrated. Turns out: they weren’t. She was the integration. When she left, it took four months and an external consultant to work out what she’d been doing.
None of this means workarounds are good, exactly. They carry real risk, as research makes clear, workaround-centric data activities carry costs that organisations often only recognise after the fact. A personal OneDrive holding live client data is a GDPR incident waiting to happen. A WhatsApp group managing project decisions is a governance nightmare. The fact that it works doesn’t make it safe.
But the answer isn’t to ban the workarounds. It’s to get curious about them.
Some practical places to start:
Ask the people who do the work. Not in a survey but in an actual conversation. Sit with someone for an hour and ask them to walk you through how they complete a key task. You will learn things that no process map will ever show you.
Treat workarounds as diagnostic data. Every unofficial spreadsheet is a bug report. Every informal WhatsApp group is a product review. Every bit of shadow IT is someone telling you, clearly, what the official system can’t do. The question is whether you’re listening.
Be honest about the gap between official process and actual practice. In most organisations, there’s a gap. Acknowledging it isn’t embarrassing. Pretending it doesn’t exist is.
Don’t punish the workaround; fix the underlying problem. If you shut down Marcus’s spreadsheet without fixing the broken export function, you haven’t solved anything. You’ve just taken away the thing that was keeping the wheels on, and made Marcus look for a new job.
Your best people are routing around the system. That fact alone tells you something important. It tells you they’re resourceful, motivated, and capable. It also tells you the system has gaps.
The organisations that thrive over the next decade won’t be the ones with the most polished process documentation. They’ll be the ones that stay genuinely curious about how work actually happens, not how it’s supposed to happen, but how it does. They’ll be the ones who close that gap not by eliminating the workarounds, but by learning from them and building something better.
The laminated flowchart is lovely. But it’s the person who built the secret spreadsheet you really need to talk to.