Every decision lands on the CEO’s desk because the company was built to send it there. Nobody has redesigned it since. Approval chains, management layers, and reporting cycles are all working exactly as built. They were built for control. Speed was never the goal.
Jürgen Dauk, author of The Leadership Operating System, spent more than 25 years inside Oracle, OpenText, and Avaya watching that play out. He treats slow decision-making as a design flaw, not a discipline problem.
His diagnostic takes about a minute. How long does it take you to change direction? How much middle management are you carrying to get a decision communicated? Are you spending more energy policing KPIs than running experiments? The uncomfortable part of his answer is where the design comes from.
Dauk’s view is that most command-and-control structures survive on fear:
- fear of losing control
- fear of missing the number
- fear of disappointing shareholders
A leader operating from there will talk about transformation while keeping it strictly cosmetic.
When a System Rewards the Wrong Things
It looks like a company where everyone is working hard on the wrong scoreboard. Sales blames marketing and marketing blames sales. Each side can prove it’s doing its job, because “everybody’s defending their own KPIs instead of moving the needle.” The system isn’t broken in the sense of malfunctioning. The system rewards people for defending their turf.
Then there’s the “trust math,” which Dauk finds funny in the way a thing can be funny and expensive at the same time. A company hires a capable person in their forties, someone who has bought a house, taken on a mortgage, and decided whether to have children. “You don’t trust them to make a decision worth a thousand bucks,” he says, describing a policy most companies would defend as prudent financial governance.
What that policy hides is the price of waiting. A bad decision is visible. Somebody made it. There may be a postmortem. A decision that sits for six weeks is harder to see, even when the delay costs real money. An executive who has never been penalized for being slow has no reason to get faster.
Years ago I heard the story of Lee Iacocca asking his engineers at Chrysler how long it would take to build a convertible prototype. They came back with months for this and that, the standard engineering answer. He told them to take a LeBaron, cut the lid off, and put a ragtop on it. That was around 40 years ago, and I’m not convinced most companies would move faster today.
What Happens When 12,000 Employees Report to Two Managers?
Dauk points to a nursing company running almost 12,000 people with two managers. Teams cap at 12, and when a team outgrows that it splits into two teams of six. Each team handles its own hiring, firing, and office space. They track a small set of measures, one of which is customer-facing time. A team below the threshold knows it’s costing the company money and goes looking for a coach on its own, without anyone summoning them to a performance conversation.
Here’s how you know the model is real. Somebody wakes up sick and doesn’t call a manager to report it. They call their colleagues and ask who can cover, because there are patients waiting and the team owns those patients. The day isn’t a list of tasks to get through. It’s a group of people someone is responsible for.
The larger version of this is Haier, the Chinese appliance maker that was near bankruptcy before it rebuilt itself as hundreds of small self-managed enterprises operating under one umbrella. It got faster, better at experimenting, and outran its market.
Haier later acquired GE Appliances, giving Dauk a chance to watch what happened when a more decentralized model met a traditionally managed business. He expected the conversion of the GE side to trigger an exodus. Engagement went up instead, because people took on ownership and, in his words, “They felt like they are needed and they have the power to influence something.”
That result should bother anyone who has justified a heavy structure on the grounds that people want direction. The result suggests people didn’t need more management. They needed real ownership.
Start With the Approval Chain
Asked where he’d start with a 90-day mandate, Dauk doesn’t reach for a framework. “I would remove approval chains.” Set clear guidelines on how decisions get made, then delete the queue of signatures those decisions currently wait in. You can say you want faster decisions, but if five people still have to sign off, the system hasn’t changed.
The organizing principle is that the fastest companies “move decisions to where the knowledge is,” rather than moving information up to where the authority sits. They also run on a handful of transparent measures visible to everyone, which removes a category of meeting whose only purpose was to discuss numbers most of the room couldn’t see.
Ownership is the test of whether any of it took. Dauk’s example is a good customer who’s unhappy with your product. On a cross-functional team close to that customer, somebody calls them, finds out what went wrong, and stays with it until it’s resolved.
In a company still running approval chains, the same problem produces a ticket, a routing decision, and a sincere belief on everyone’s part that they did their job.
I got the customer’s view of that machinery last week. My insurance card promises a nurse helpline 24 hours a day, seven days a week. When my hand swelled up with an infection on a weekend, the line wasn’t available. I went to urgent care, saw one of their own approved providers and got a prescription.
A few days later the company emailed to inform me that prescriptions require pre-approval, which takes two to three days. In what version of the world does a swollen hand wait two to three days for a form to clear?
Fix the Work Before You Automate It
Dauk watched a manufacturing client run the standard playbook. Legal and compliance drove the requirements, IT got the order to roll a solution out company-wide, millions went out the door, and almost nothing improved. Switching on a copilot inside an office suite doesn’t change how work gets done.
His sequence runs the other way. Get the vision and purpose clear first, and make it something beyond growing revenue 20% a year, since nobody empowers themselves in service of a growth target. Then go to the teams and ask what repetitive work is standing between them and their best work. It’s a question a company can only ask if decisions haven’t already been centralized away from the people doing the job.
One of his clients, a small business, tested that on a single function. An outside firm automated the bookkeeping department’s repetitive tasks and cut them by 80%. Nobody was fired. The bookkeepers were redeployed to call the company’s best customers and ask what was going wrong and what would help.
Both customer satisfaction and employee engagement climbed. What Dauk keeps running into is the question executives ask first, which is “How much can we save?” The better one is “What else can we achieve?”
He cites NVIDIA’s Jensen Huang making the harder version of the point: an executive whose first instinct with AI is cost reduction has “lost their imagination.”
I’d put it in terms of a hospital bill. When my son was in the hospital, the doctor might have spent half an hour with him over several days. That didn’t mean the rest of the care was wasted. It meant the hospital knew which work actually required a doctor. Most companies haven’t done that exercise for their own people. Instead, they’re buying software to speed up work nobody should be doing.
Ask Your People for Solutions
Given one piece of advice to leave with owners and executives, Dauk said, “Ask your people for solutions. They know better.” He pairs it with one condition: empowerment without a clear vision and real guardrails produces confident decisions pointed in unhelpful directions.
The bottleneck usually doesn’t look like a bottleneck. It looks like a capable employee waiting for permission. Maybe it’s the same 40-year-old who signed a mortgage on Saturday but can’t spend $1,000 on Monday without a signature.
“Ask your people for solutions. They know better.”