The executive brief
- Hidden friction makes work more expensive than it looks. Information hunting, rework, workarounds and unnecessary handoffs consume time and capacity.
- Efficiency in one function can create friction elsewhere. Leaders need to look across the organization, not just at individual team metrics.
- The real process often differs from the official one. Workarounds can reveal where processes no longer reflect how work actually gets done.
- Clear ownership matters. Assigning accountability for outcomes—not just individual steps—can help organizations reduce friction and create more value.
Join 100,000+ subscribers for the latest tech news
Stay up to date on the most important—and intriguing—industry news on AI, automation, data, quantum, infrastructure and security with the Think Newsletter, delivered twice weekly.
Some of the most expensive work in an organization looks like good work.
A report gets another review. Someone pulls information from three systems because no single source has what they need. A team keeps a spreadsheet alongside the official system. None of this looks like failure. It looks like people making sure the work gets done.
That kind of adaptation is increasingly becoming the way work evolves. In the IBM Institute for Business Value research, 52% of employees say the tasks they are assigned have changed over the past year as AI reshapes their day-to-day responsibilities. Yet organizations are still more likely to let work evolve through day-to-day adaptation than deliberate work redesign.
But these behaviors often reveal hidden friction in the way work gets done. When that friction becomes part of the routine, it consumes time, slows decisions and limits capacity. That is where the hidden friction becomes a business problem.
For business leaders, the question, then, isn’t simply “How can we make this work faster?” It is, “What are people doing to keep the work moving that shouldn’t have to be part of the work?”
Here are five familiar but invisible forms of friction that can help reveal the answer.
1. People spend more time finding information than using it
Information tends to sit across many systems, documents and teams. Before anyone can act on it, they often have to track down the right sources, assemble the relevant pieces and confirm the information can be trusted. As a result, highly skilled people can lose a large share of their time to what Emily Lansdowne, AI Strategy Lead for Enterprise Transformation at IBM, calls “information hunting and collating.” This leaves less room for the analysis, judgment and storytelling where they add the most value. For example, Emily is working with a client that has limited visibility into the AI solutions already being used across the organization. Before they can decide which initiatives to scale, the client needs to piece together information from cybersecurity tools, employee surveys, AI champions and software contracts—a process that is consuming significant time before any meaningful analysis or decisions can begin.
When this kind of friction becomes routine, people often learn to work around it rather than challenge the underlying problem. One reason, says Lansdowne, may be how ownership is distributed. When people are accountable for an outcome — not simply one step in a process — they have more incentive and authority to question why the work takes so much effort in the first place. Giving the people closest to the work greater ownership can surface inefficiencies earlier, creating a stronger incentive to remove them rather than simply work around them.
2. Every function becomes more efficient. The work still gets harder.
Productivity programs tend to track gains function by function. Marketing generates more demand, finance closes its reports faster and operations moves more inventory, so each team can point to progress on its own metrics. But a productivity gain in one part of the organization can unwittingly add work somewhere else. For example, more output can create more reviews, decisions and coordination for the teams receiving it.
The effects of that progress travel across the organization. As Lansdowne observes, “If you’re seeing the volume of output from one particular team spike above others, it’s likely that pressure is flowing to other parts of the organization.”
How people are measured can reinforce this effect. “People will automatically try and game the system to try and hit the numbers that are being measured,” says Lansdowne. That matters when organizations encourage greater use of AI. “We’re not being intentional enough about the incentives that we are designing in,” she says, particularly if the ambition is to drive “an increase in quality, not just an increase in volume.”
The broader view is to follow the work across functions, recommends Lansdowne. Where does additional output go? What demands does it create downstream? A team can become more efficient on its own measures, while the overall flow of work becomes more complicated.
3. The official process and the real process have drifted apart
There’s how work is supposed to get done — and there’s how employees actually get it done. The gap between the two can reveal where a process has stopped reflecting the reality of the work.
A spreadsheet appears because a system misses something important. An email chain develops around an approval. A colleague becomes the person everybody calls because they know how to get something unstuck.
It is tempting to file these adaptations under anomalies or one-offs. Look more closely, and they often point to the places where the formal process has fallen out of step with the way work is really done.
“The best way that I found to do this is sit and observe the team getting the work done,” Lansdowne says.
Watching the work happen can reveal what process documentation misses — where people leave the system, rely on another source of information, create a workaround or use knowledge that exists nowhere in the formal process.
But Lansdowne sees another question emerging, as more execution work moves to AI agents. The issue may no longer be whether an AI system can follow the human process step by step. Instead, leaders may need to describe the outcome they want to achieve, provide the agent with the appropriate data, tools and access, and then observe how it gets there.
That could mean the path an AI agent takes, which could look very different from the path people have traditionally followed. An agent may find information in a different system, skip steps that humans have always followed or expose dependencies that were never formally documented.
Viewed in this light, the informal process is worth more than its inefficiencies suggest. It captures how work really gets done, which makes it a practical place to begin when deciding what should change as the work evolves.
4. Work gets done. Then someone does it again.
Rework can look responsible. When the stakes are high, it can feel entirely reasonable to have another person review a report, recreate an analysis or check work that has already been checked. The trouble is that these layers of review can gradually become part of the way work gets done, until nobody questions whether they are still adding anything or simply repeating what has already happened.
Lansdowne points to two reasons behind much of this repetition: trust and accountability. “The rework here is because somebody doesn’t trust that the work’s been done properly the first time,” she says. There is another reason, too. Someone may feel they need to check the work themselves because “my name is on the hook if it goes wrong.”
That raises a different question as AI takes on more work. If people already recheck work because they don’t trust it, introducing AI into the process doesn’t automatically resolve that underlying problem. People may still feel the need to check, validate or redo the work before they are willing to rely on it.
So, the question for leaders isn’t simply whether a review can be removed. It is worth understanding why the review exists in the first place. Is it there because the work genuinely needs another layer of scrutiny, or because people don’t have enough confidence in what happened before? That distinction can help leaders focus their attention on the source of the rework, rather than simply trying to eliminate the visible step.
5. Everyone owns a step. Nobody owns the outcome.
Many of the outcomes that matter most cut across organizational boundaries. Customer onboarding, employee experience, compliance and strategic initiatives are examples that rely on several teams delivering part of the work. With responsibility spread this widely, it can be hard to tell who is accountable for the outcome.
Lansdowne sees this regularly when she asks clients how a process works. For example, she may be pointed to three different people or departments, each responsible for a different part. Sometimes, there isn’t even a clear owner. Instead, there is one person who has been around long enough to simply know how the whole thing works — and when that person is away, the process can quickly become difficult to navigate.
That is a different kind of organizational risk. The work may have owners at each stage, but there is no one person responsible for making sure all those stages come together to produce the intended outcome.
Lansdowne also looks at where the money and reporting lines sit. Who funds the people doing the work? Where do they sit? Who owns the transformation budget? Who paid for the change in the first place? Those details can reveal where an organization believes ownership sits, even when the process itself crosses several functions.
That raises a different question — who is accountable for the value an outcome is meant to create, and who is accountable for the risks that come with changing how the work gets done?
Lansdowne suggests that these responsibilities do not necessarily need to sit with the same person. In fact, creating some intentional good friction between them can be useful. The person accountable for the value created by an outcome may be incentivized to take risks to capture more of that value. The person accountable for risk can provide the counterweight, making sure those risks are understood and the right mitigations are in place.
The goal isn’t to eliminate that tension. It is to make it productive — so decisions about changing how work gets done balance the opportunity to create value with the responsibility to manage risk.
Most organizations are designed to make work move. Over time, they also accumulate the extra work people do to keep things moving. That hidden effort has a business cost. It consumes capacity that could otherwise go toward customers, decisions, innovation and growth. For business leaders, the bigger issue is that this effort often sits between functions, outside formal processes and beyond the measures used to manage individual teams.
Answer those questions, and leaders can begin to reclaim that capacity, redirecting time and energy from managing friction to creating value for customers and the business. The opportunity isn’t simply to make work move faster. It’s to make more of the work matter.
Brand and Content Strategist
Thought Leadership
