9 Areas Where COOs Are Spending More Time Now Than 5 Years Ago
The role of the Chief Operating Officer has shifted dramatically as operational priorities adapt to new technologies and workforce expectations. Through conversations with seasoned COOs and operational leaders, nine distinct areas have emerged where executive attention has intensified compared to half a decade ago. These shifts reveal how modern operations teams balance human capability with technological advancement while maintaining organizational agility.
Invest in Workforce Capability
One unexpected shift has been the amount of time dedicated to workforce capability planning rather than traditional operational oversight. Five years ago, operational efficiency and delivery metrics dominated the COO agenda. Today, a significant portion of time is invested in aligning skills with rapidly evolving business demands, particularly in areas like AI, cybersecurity, and agile transformation. Research from the World Economic Forum indicates that nearly 44% of core skills required for jobs are expected to change by 2027, making continuous upskilling a strategic priority rather than a support function.
This shift has reshaped leadership style toward a more future-focused and talent-centric approach. Decision-making increasingly revolves around long-term capability building instead of short-term operational gains. There is greater emphasis on listening to workforce signals, fostering adaptability, and building learning ecosystems that can respond to constant disruption. The COO role has evolved from driving efficiency to enabling resilience and sustained growth through people.
Remove Drag from Processes
The area where I spend significantly more time than I did five years ago is deciding what not to build. Not in the product sense. In the operational sense. Which processes not to formalize, which meetings not to add, which tools not to adopt, which reports not to generate. Most of my operational work used to be about building the scaffolding a growing company needs. More of it now is about refusing to add scaffolding that would quietly slow the company down.
The shift happened gradually and caught me off guard. Early on, every process we added solved a real problem, so adding processes felt like progress. At some point the math flipped. New processes started creating more friction than they removed, and the cost showed up in slower decisions, longer meetings, and teams that felt supervised instead of trusted. I hadn't noticed the inflection point because each individual addition still seemed reasonable. It was the accumulation that was expensive.
Now I treat operational additions the way I used to treat engineering debt. Every new process, dashboard, or recurring meeting gets a clear purpose, an owner, and a defined review date. If it's not earning its place by the review date, it goes. That discipline sounds simple. It's the single hardest thing to hold because every process has a defender who can explain exactly why it exists, and removing anything feels like a loss even when the accumulation is obviously drag.
The impact on my leadership style has been significant. Five years ago I was biased toward structure. I believed the right answer was usually more clarity, more documentation, more alignment. Today I'm biased toward restraint. I ask whether a team's problem is really about missing structure or about too much of it, and more often than I expected, the answer is the second one.
This changed how I coach other leaders too. When someone proposes a new process, my first question isn't whether it would help. It's what existing process it would replace, and what it would cost us if we didn't add anything at all. That framing forces a different conversation. It pushes people to think about the cost side of operational design, which is the side most of us were never trained to see.
The lesson I'd pass on is that the mature operator's job is usually subtraction. Growth doesn't come from adding more scaffolding. It comes from building the company so the scaffolding has less work to do.
Prioritize Automation Integrity
Five years ago when I was an operator, I generally spent most of my time managing the velocity of engineering development and the quantity of work produced- basically simply getting written code. Today, the majority of my unanticipated time is now spent in AI governance and the oversight of automated decision-making processes. Velocity management alone is no longer sufficient; now I also need to verify that guard rails are effective in holding up during use.
As a result, I've transitioned from being a 'check-the-box' manager to being a 'check-the-guardrail' type of leader. Rather than asking 'did you build it?' I now ask 'how did you verify the integrity of the data?' My leadership style has shifted increasingly toward verifying the system as a whole rather than focusing on performance of individual team members. The roles that I originally assumed as COO have evolved from overseeing the production pipelines to overseeing the integrity of automated decision-making processes.
Transitioning my leadership style from speed-first to trust-first is uncomfortable due largely to the perception that it's a slower pace than what we previously operated at. However, since we now heavily rely on AI in our operational processes, that slower paced friction is the sole factor that keeps the organization functioning flawlessly.
Champion Culture and Candor
The area where I found myself spending significantly more time, and where I see every strong second-in-command spending more time today, is at the intersection of people and culture.
Five years ago, the COO role was understood primarily as an operational function. Systems, processes, execution, efficiency. The assumption was that if the structure was right and the metrics were visible, performance would follow. That assumption still holds, but it is no longer sufficient on its own. The human layer beneath the structure has become just as important as the structure itself, and the COOs who have not adjusted to that reality are leading organizations that look organized on paper but feel disconnected in practice.
The shift happened gradually and then all at once. Remote and hybrid work removed the ambient signals that once made culture visible. A five-minute hallway conversation, the energy in a room before a meeting, the informal feedback loop that happened naturally when people shared physical space. When those disappeared, the COO could no longer rely on proximity to know how the organization actually felt. You had to build intentional systems to surface what used to surface on its own.
That meant investing real time in one-on-one conversations that went beyond status updates. It meant redesigning meeting rhythms to create space for honest communication rather than just progress reporting. It meant paying close attention to what wasn't being said in group settings and following up with people who went quiet.
The impact on leadership style has been significant. I became more deliberate about creating the conditions for candor. I stopped treating culture as something that existed alongside the operating model and started treating it as part of it. When people feel genuinely seen and heard by their leadership, they perform differently. Not because they are more talented, but because the environment is no longer working against them.
The COOs who are most effective today are those who understand that execution is a people problem as much as a systems problem. The org chart matters. The meeting rhythm matters. The metrics matter. And none of it works the way it should if the people inside the structure do not trust the people leading it.
Derek Fredrickson
Founder & CEO, The COO Solution
derek@thecoosolution.com | thecoosolution.com
LinkedIn: https://www.linkedin.com/in/derekfredrickson
Teach Teams to Direct Machines
Five years ago, I managed humans. Today? I spend half my week managing how humans talk to machines.
Frankly, it's a completely different game. Back in my cross-border e-commerce days, fixing a bottleneck meant tweaking a spreadsheet. Now at TAOAPEX, scaling tools like TTprompt and TaoTalk, the real bottleneck is psychology.
Look, when we deployed our latest internal AI agents last month, productivity actually dropped 14% initially. My team treated the AI like a glorified search engine—expecting magic without providing context. I spent 30 hours that week deep in the weeds (literally rewriting over 200 system prompts myself—it was painful) just to show them how to communicate with the models.
And that completely flipped my leadership style. I used to be a classic COO—enforcing KPIs, building rigid processes. Now? I'm basically a context provider. I don't just assign tasks anymore; I have to teach my team how to actually delegate to MyOpenClaw and our other internal tools.
Here's the truth about building SaaS right now: The smartest algorithms won't fix a broken human workflow. The irony of artificial intelligence is that it demands twice as much human empathy to implement correctly.
Set Context for Constant Change
Five years ago, I spent most of my time on client strategy and delivery. The area I did not expect to be in today is internal communication, specifically, helping the team make sense of an industry that is changing faster than most people can process.
That shift surprised me. I assumed scaling would mean more time on operations, systems, process. Instead it meant spending significantly more time on what I can only call clarity work. When Google changes core algorithm logic, when a technique that worked six months ago stops working, when clients start asking what AI Overviews mean for their traffic, the real job is not just figuring out the answer. It is keeping a team confident and directional while the ground is actively moving underneath them.
The leadership style impact has been real. I became less of a decision maker and more of a context setter. That is a genuine distinction. A decision maker shows up with answers. A context setter makes sure the team understands the situation well enough to reach their own answers. The second one is harder and takes more time, but it builds a team that can think, not just execute.
The habit that came out of this is what we now call a weekly signal briefing. Every Monday, before client work starts, we spend twenty minutes on what shifted in the industry that week and what it means for how we operate. Not a lecture, a conversation. It has done more for team retention and output quality than any process document we have ever written.
The unexpected areas are almost always where the actual leadership work lives. Everything else is management.
Architect Lean Autonomous Systems
I'm Runbo Li, Co-founder & CEO at Magic Hour.
The honest answer is that I spend a dramatically larger percentage of my time on AI systems management than I ever would have predicted. Not building AI products for customers, but using AI internally to replace entire workflows that used to require teams of people. Five years ago, a company at our scale would have a head of marketing, a customer support team, a finance person, maybe a junior ops hire. We have none of that. It's me and my co-founder David, and we serve millions of users.
That shift forced me to think about leadership completely differently. I don't manage people. I manage systems. Every morning I'm evaluating which AI tools are handling customer support tickets, which ones are generating marketing content, which automations are breaking down and need to be re-architected. I spend real time designing prompt chains that handle tasks a mid-level employee would have owned three years ago.
Here's a concrete example. We had a spike in support volume after a product update. In a traditional company, you'd pull in extra support reps or ask the team to work overtime. I spent two hours rebuilding our AI support pipeline, tuning the responses, adding edge case handling, and by that evening it was resolved. No hiring. No overtime. No Slack fire drill.
That changes how you lead because it changes what "leadership" even means. I used to think a great operator was someone who could recruit, motivate, and retain excellent people. That's still true for companies at a certain scale. But for early-stage founders right now, the highest-leverage skill is knowing how to design, deploy, and maintain AI-powered systems that do the work of ten people. The founders who figure this out first will build faster, stay leaner, and outrun competitors who are still posting job listings.
My leadership style went from "how do I get the best out of people" to "how do I architect systems so good that two people can do what fifty used to." That's not a subtle shift. That's a completely different operating philosophy.
Curate a Focused Tool Stack
Five years ago most of my week was process design and cross-team alignment. Today, most of my week is AI and tooling decisions. Every department wants a new tool, and the silent cost of saying yes to all of them is tool sprawl that destroys focus. I now spend about 20% of my time just evaluating, piloting, and sunsetting software. The other shift is time with customers. Operational patterns hide in customer feedback, so I take two customer calls a week even though my title does not require it. The COO job used to be about systems. Now it is about systems plus stack plus signal.
Build Durable Decision Frameworks
The area that's consumed far more of my time than I anticipated: AI integration decisions. Five years ago, the technology and tooling choices at the operations level were relatively stable. You picked your stack, implemented it, and spent your time on execution. Now, the landscape shifts fast enough that a meaningful portion of operational leadership is continuous evaluation — what's worth adopting, what's not ready yet, what's worth the transition cost versus the performance gain.
At Dynaris, we build AI automation infrastructure, so this might seem unsurprising. But even outside our product work, the internal operations questions have multiplied: Which AI tools should our sales team use? What's the policy on AI-assisted writing in customer communications? How do we audit the accuracy of AI-generated outputs before they reach customers? These aren't pure technology questions — they're judgment calls about trust, risk tolerance, and team capability.
The shift this has created in my leadership style: I spend far more time on decision frameworks than on decisions. Five years ago I was making more direct calls. Now I'm more focused on helping my team develop the criteria to make those calls themselves, because the volume and velocity of decisions has outpaced what any single person can handle thoughtfully. The question I ask more often now is: "What principle would you use to make this call in a similar situation next time?" rather than just answering the question in front of us. That shift feels like the most durable change in how I operate.
