We Didn't Need More Work. We Needed To See Our Work Differently.

The profit-sharing bonuses used to be real money.

At the 18-person digital agency where I was working, those bonuses had been ten to twenty percent of total compensation. Twice a year, for years, they landed. They were part of what made working there feel like the place was yours, not somebody else's.

Then, over a couple of years, they shrank. Then they disappeared.

The team was as strong as it had ever been. The work proved it — clients renewed, projects launched, awards came in. But the margin was gone. And with it went something harder to replace: the quiet confidence that when the team did good work, the team would share in it.

Then came the one-on-one.

An employee asked me, straight out, when they were getting their bonuses back. There's no question a leader wants to answer less. You can shrug. You can placate. Neither is what the person asking deserves. What they deserve is to hear that the company is getting back to healthy, and that someone is working on it.

I didn't have that answer yet. The question was the moment I knew I needed one.

The problem was that nobody was looking at the whole picture. We were working in silos. The account manager was looking at a pipeline that kept coming up thin. The founder was looking at a top line that wouldn't grow. Production was looking at a backlog they couldn't keep up with. Each person was holding a piece of the same picture, and every piece kept pointing to the same answer.

We need more work.

I disagreed.

The Disagreement Was the Data

The argument was simple. Revenue wasn't the problem. We were already stretched. We were already hiring contractors to cover overflow. People were being shifted mid-project to meet deadlines. Other projects went short-handed. More shifting followed. The pattern was visible to anyone looking for it, but we weren't. We were watching the top of the funnel.

I had a different theory. The issue wasn't how much work was coming in. It was what was happening to the work once it got here. The delivery model was broken, and every additional project ran through it. More work wouldn't fix margins. It would scale the problem. Declining profitability doesn't end with a team being upset. It ends with the company falling apart. And we weren't going to stop that by staying in our own lanes.

Leadership was divided. We had opinions. We didn't have data. That kind of disagreement isn't noise. It's the crux. I stepped in to break the stalemate.

I partnered with the Operations Manager and ran an evidence-based audit. Two roles holding the same picture instead of their own. Small move at the time. The first crack in the silo, in retrospect.

We looked at how often projects were going over budget, how much production time we were actually using, where our estimates diverged from what we delivered, and what the real opportunity cost was on the projects that had lost money.

The numbers, once we had them, realigned the room. Large projects — the ones over a hundred thousand dollars — made up a significant share of our portfolio. A majority of them lost money. The total opportunity cost, if we had captured even twenty percent of it, would have made that one of our most profitable years on record.

Project overages became the top priority. Not because I convinced anyone. Because the data had.

What Actually Changed

Over the next two years, net profit margin went from 0.3 percent to 33 percent. A 32.7 percentage point swing. Everyone got their bonuses back.

Here's what I want you to notice.

No AI was involved. This was 2015 and 2016, before any of the current AI conversation existed. There was no new technology driving the change. There was no platform, no tool, no framework we bought. What changed was not the work the team did.

What changed was what each person focused on.

We moved from siloed delivery to shared visibility. Everyone could see capacity, including their own, and make smarter decisions about what to commit to.

We moved from individual deadlines to collective flow. Project managers could push back on scope changes without breaking relationships, because pushback was the culture now, not a personal stance.

We moved from "estimating is hope" to "estimating includes project management time, and we bill for it." What had been tens of thousands of unbilled dollars per manager every year became revenue.

We moved from "scope creep is inevitable" to "scope changes become future features." That protected budgets without shutting down ideas.

I said out loud what nobody was saying: the skill gaps, the process flaws, the broken assumptions. I brought in guest speakers. The Operations Manager and I designed internal tools that gave everyone on the team — not just leadership — a clear view of project health, capacity, and cashflow. I trained managers to respond to what the data was telling them.

The shift was in what people valued and what they spent their time on. It was a shift in operating rhythm.

By year two, I was stepping back. The Operations Manager was carrying the momentum forward. The team was making the system better without me in the room. That's the part I was proudest of — not that I fixed the margin, but that I built a team that could lead itself once the pattern was in place.

The bonuses came back. Twice a year, like before.

The Factory Era of Management

I've been turning that experience over for years. Here's where I've landed.

Most of what we call modern management is still Taylorism with better software.

In 1911, an engineer named Frederick Taylor published the book that taught American factories how to ten-times their output. The formula was simple: break every job into its smallest measurable steps, separate planning from doing, put managers in charge of the whole system, and optimize relentlessly for throughput. It worked. It built the century of American manufacturing.

It also set the frame that almost every modern operating system quietly inherits. Agile fragments work into smaller pieces. Lean cuts waste from the pieces. OKRs and KPIs measure the output. Performance reviews rank individuals against standardized metrics. Even the flattest orgs organize around roles defined by their throughput. The "10x engineer," the "10x operator," the individual who ships more than everyone else — same idea, scaled down to the person. The software has changed. The frame hasn't.

These are sophisticated optimizations of a model built around one constraint: human processing capacity. When one person could only hold so much context, only make so many decisions, only track so many variables, the work had to be routed through hierarchy, fragmented into roles, and measured by throughput. That's what the factory era required, and it worked for its time.

But that constraint is ending.

AI doesn't accelerate the factory. That's the part of the current conversation I think most people are getting wrong. The productivity framing — AI helps us do the same things faster — treats AI as the next spreadsheet. It isn't. The genuinely new capability is different in kind, not degree. AI holds context across scales. It detects patterns across domains. It operates as a thinking partner, not just a tool. That creates a new organizational primitive, not a faster version of the old one.

The teams that will thrive in the next decade won't be the ones that automated the most. They'll be the ones that already knew what to do with capacity once they got it back. Because that's what the removal of a capacity constraint does — it returns capacity. The question is where that capacity goes. Into doing the old work faster? Or into work that couldn't exist before?

What we're really asking is the difference between efficiency and effectiveness. Efficiency is doing the same thing faster. Effectiveness is doing the right thing — and efficiency is a subset of it, not a synonym for it. The factory era ran on efficiency and treated it as the whole game. The next era runs on effectiveness, with efficiency still inside it, still useful, just no longer the point.

The orgs that win redirect their returned capacity toward the work that actually creates value. The ones that don't, concentrate it, over-automate, and end up run over by the orgs that did.

We Didn't Get Faster. We Got More Effective.

That's what that two-year stretch proved.

The transformation at that agency wasn't an efficiency gain. We didn't get faster. We got more effective — at every layer. The tools we used. The technology we depended on. The culture we built. The relationships between roles. The programs that shaped our work. The work product itself, and the processes that moved it through the team. Effectiveness across all of it.

We changed what each person focused on, what we valued, and the rhythms we worked in. The margin improvement was a lagging indicator. The real shift was upstream, in how the team thought together and decided together.

That's not an AI story. That's a leadership story. The people who came through that transition learned what it feels like to work on a team that reasons together, not one that hands off deliverables between roles. They learned what it looks like when estimates reflect reality instead of aspiration. They learned that pushing back on scope is a skill, not an attitude problem. They built new defaults together.

And that was before any of us had tools that could hold context across the whole organization at once.

What's possible now, with those tools, is the same shift at a different scale. AI doesn't make us faster at the old work. It makes us more effective at the work that matters. And effectiveness is what the next era runs on.

Transformations like that stay rare because efficiency and effectiveness get treated as the same thing. They aren't.

What Caldwell Leadership Is For

I'm writing this as the first piece on the Caldwell Leadership site because it's the argument that sits underneath everything else here.

The work I do is with founders and leaders whose teams have outgrown running every decision through the top of the org. The pattern shows up early. Growth breaks something that used to feel natural. The leader becomes the bottleneck for decisions that nobody else is equipped to make, because nobody else has the full picture. The work increases, the margin gets worse, and the conversation becomes "we need more work" when the real issue is "we need to see our work differently."

What changes is where decisions get made. They distribute through the organization, to the people at the right level and right resolution to see the problems and work them together. Leadership's role shifts too. It becomes less about making every call and more about shaping the space around the work — what's visible, what matters, how the team decides.

I've been in that room. I've been the one raising my hand. I've done the audit. I've watched what changes when teams start reasoning together instead of handing off to each other. I've seen it work before any of us had the tools we have now, and I'm designing for what's possible now that we do.

Caldwell Leadership is the practice of building teams that think clearly, reason together, and move in the same direction. The capacity constraint that shaped the last era of management is ending. The leaders who recognize that shift get to design what comes next.