The New Economics of Work: What Changes When Agents Join the Team

Alex Moore — Sun, 11 Oct 2026 19:57:49 GMT

For most of my career, the math of growth was simple. If you wanted more output, you added people. More clients meant more account managers. More content meant more writers. Capacity grew in the same increments as payroll, and every owner I know has felt the weight of that equation when deciding whether to hire.

That equation is changing, and faster than most leadership teams have planned for.

The cost of a task is falling

The price of machine intelligence has collapsed. Stanford's 2025 AI Index found that the cost of running a model at GPT-3.5's level of performance fell more than 280-fold between November 2022 and October 2024.

At the same time, AI has moved from answering questions to doing work. Agents can take a goal, break it into steps, use your systems and hand back a finished draft, a cleaned-up CRM record or a rebuilt web page for a person to approve. Microsoft's 2026 Work Trend Index reports that active agents in Microsoft 365 grew 15 times year over year, and 58% of AI users say they are producing work they couldn't have produced a year ago.

The cost of completing a routine unit of work is heading toward zero. So the real question for every leader is what holds its value when that happens.

What I saw in the room

Last week we hosted a group of agency owners and operators at our Charlotte office for two days of building, testing and debating exactly this.

One owner had been quoted 87 development hours for a single feature on a website rebuild. Instead of waiting, that owner pointed an agent at the migration, and we watched it work through the site and generate the theme, templates, metadata and page content while we talked. Another participant turned a design file into editable CMS modules in about five minutes and had a client landing page live in roughly fifteen.

Everyone at the table felt it. The price of production work is resetting. And in the same two days, the conversation kept coming back to the two things agents can't supply on their own.

The value moves to thought and guidance

Someone once told me a story that has stuck with me ever since. A plumber walks into a house with a problem in the pipes, taps the pipe twice and fixes it in five minutes. Then he hands the homeowner a bill for $1,500. The homeowner objects that he was only there for five minutes. The plumber says the five minutes were the easy part. The bill is for knowing exactly where to tap.

For decades, agencies and professional services firms have been valued for how long it took us to figure something out. That is no longer the measure. When an agent can produce the first draft in minutes, what clients pay for is knowing what to build, what to ask for, how to judge the output and what to do next.

The demonstrations that worked best last week had one thing in common: someone did the planning and the architecture before the agent touched anything. One owner showed a job-scheduling application built for a furniture refinishing client. The client had tried to build it themselves with AI and failed. The app worked because someone understood how that business scheduled crews, costed jobs and tracked capacity. The code was the easy part.

That is what a human in the loop really means. It is more than a person clicking approve at the end. It is a person framing the problem, setting the standard, reviewing the work and feeding every correction back so the system gets better next time. The tools change every few weeks. The mindset is what makes them useful, and getting a whole team to think that way is the real work of leadership right now. Microsoft's research points the same direction: only 26% of AI users say their leadership is clearly aligned on AI.

Your data and your way of working are the asset

Most agencies don't own physical assets. Neither do most consulting firms, nonprofits or service businesses. So we have been asking ourselves a harder question: what is our data, infrastructure and process asset?

Every organization now has access to the same models. The tool is no longer the advantage. What no one else has is your context: your client history, your playbooks, your pricing logic, the decisions you made and why, the lessons from projects that went sideways. For a team of 15 to 25 people, capturing that knowledge used to be nearly impossible. Today it shouldn't be.

No two organizations work the same way, and that difference is now the value proposition. An agent connected to your own processes, data and standards does work that reflects how you operate. An agent without that context produces the same generic output it produces for everyone else.

This changes how we talk about what we offer. Our value is our experience, encoded into systems that carry the way we work. The best version of a modern agency, or any professional team, looks a lot like a good business analyst sitting next to the client, understanding the process and the data, and engineering something around it that lasts. Clients will increasingly expect the same from their own organizations.

One caution from the room: in one participant's review of 200 businesses, 90% had no defined sales process. Agents amplify whatever they are given. If the process and the data are a mess, the agent will scale the mess.

Reading the shift through the 12 P's

In my work on P12, I describe twelve decisions that build, scale, and sustain a business. Four are changing fastest.

People. Roles are shifting from doing every step to framing, directing, and checking the work. Train for judgment and systems thinking, and expect managers to model the tools themselves.
Processes. An agent can only follow a process that exists. Writing down how you work is now the first step to scaling it.
Price. Hourly billing stops making sense when a week of work takes an afternoon. Price on outcomes, experience and judgment, and watch the shift in your own software costs from per-seat licenses to usage-based credits.
Profit. Margins open up for whoever reorganizes first. A team of eight can serve a client base that once required twelve, or add a service line without adding a department. Where that capacity goes belongs at the leadership table.

Where to start this quarter

Fix one foundation first. If your sales process, CRM data or reporting is shaky, start there.
Capture your context. Document one core process, the standards behind it and the lessons your team has learned. That becomes the knowledge your agents work from.
Pick one workflow with a clear cost and measure it in hours or response time today.
Name a human in the loop. Someone frames the work, reviews the output and feeds corrections back into the system.
Measure the capacity you gain and decide where it goes. Put it on your scorecard.

The opportunity in front of us

Every major shift in the economics of work has rewarded the leaders who redesigned around it early. The cost of doing has dropped. The value of thinking well, guiding the work and knowing your own business deeply has gone up.

If you are working through what this means for your team, it is exactly what I cover in my AI for Leaders and Teams sessions and in one-on-one coaching.

Sources: Stanford HAI, 2025 AI Index Report; Microsoft, 2026 Work Trend Index.

Insights