Back to HomeStrategic Intelligence

The Death of the Coordination Tax: Orchestrating the Post-Management Era

By PickingMay 6, 20266 min read
The Death of the Coordination Tax: Orchestrating the Post-Management Era

The Coinbase Inflection Point

In May 2026, Brian Armstrong, CEO of Coinbase, sent out a memo that may well be remembered as the "Magna Carta" of the AI-native corporate era. The headline was a 14% reduction in staff, but the real story was in the "why." Armstrong wasn't just reacting to crypto market volatility; he was formalising a structural shift that had been quietly reshaping Coinbase's most productive teams for over a year. He was declaring war on the Coordination Tax.

The Coordination Tax is the silent killer of innovation. It’s the friction that arises when an idea must travel through a hierarchy of human intermediaries—PMs, BAs, Team Leads, Scrum Masters, QA—before it materialises as value in production. For decades, in both the 20th-century corporate model and the early 21st-century "Agile" model, we accepted this tax as the cost of scaling. The assumption was simple: more people equals more output.

Armstrong's memo shattered this assumption. He revealed that Coinbase engineers, leveraging AI, were shipping in days what used to take teams weeks. The company flattened its organisation to five layers and embraced "one-person teams."

For those of us in Solutions Consulting and Product roles, this is a wake-up call. We’ve spent our careers as high-fidelity bridges between "Vision" (Sales/Executive) and "Reality" (Engineering/Delivery). We were the tax collectors, managing handoffs.

But if a single individual can now deliver a production-grade product autonomously, the "bridge" is no longer a job description. It’s a workflow.

The Solutions Consultant as "Architect-Builder"

In my day job as a Solutions Consultant at Gentrack, I operate at the intersection of energy technology and financial infrastructure. Historically, the SC role has been one of "Technical Sales Support." We demo the possible, architect the vision, and then toss it over the wall to a delivery team tasked with building what we promised.

AI has "decontaminated" this role.

The shift from "demoing the possible" to "building the actual" is happening with zero latency. Previously, if a client requested a bespoke data integration or a new visualisation for tariff modelling, I’d file a Jira ticket with a Product Manager. That PM would prioritise it in a sprint, and eventually, an engineering squad would build it.

Now, as an "Intelligence Architect," I don’t file tickets. I orchestrate agents.

In one weekend, I built EnergiaProsto—a fully functional energy price comparison engine for the Polish market. This wasn’t a prototype. It’s a live system with a Prisma-backed PostgreSQL database, daily automated URE tariff scrapers, and a Next.js 14 frontend. Using the Hermes + Claude Code stack, I bypassed traditional departmental boundaries.

The Solutions Consultant is no longer just a talker; they are the Builder-in-Chief. This shift isn’t just about efficiency; it’s about Strategic Responsiveness. When you can go from a client’s "What if?" to a working URL in 48 hours, the competitive landscape changes. You’re not just selling software; you’re selling the speed of thought.

The Coordination Tax is dead. The new currency is speed-to-value.

Job Decontamination and the "Pure Manager" Dead-End

[Section to be completed — Focus on "No Pure Managers" and the 15+ direct report mandate.]

The Transition — From Headcount to Orchestration

For decades, the "Manager Track" was the default path to seniority in corporate hierarchies. Your status was measured by the size of your budget and the number of desks in your department. In the post-management era, this metric isn’t just obsolete—it’s a liability.

The most powerful "Direct Reports" in my organisation aren’t on my payroll; they’re in my ~/.hermes/ directory.

Shifting from a Headcount mindset to an Orchestration mindset requires a fundamental rewiring of how we view work. If you’re an SC or a PM today, your path to the "future of work" involves three specific transformations:

1. From Requirements to Context Engineering

Traditionally, a PM writes a Product Requirement Document (PRD), detailing what to build. In an agentic workflow, you engineer the context. You don’t instruct the agent to "write a blog post." Instead, you feed it 7,000 bookmarks, your daily morning briefing, and the specific brand voice of Lumen Pro, then ask it to identify the signal in the noise. You’re not managing a person’s time; you’re managing a model’s world-state.

2. From Approval to Alignment

The "Pure Manager" acts as an approval gate: "Bring it to me when it’s done, and I’ll tell you if it’s okay." The "Intelligence Architect" operates in a continuous alignment loop. Tools like Claude Code and Hermes run in parallel. I don’t wait for a weekly sync; I’m the "human-at-the-edge," aligning the agent’s trajectory every few minutes. The latency between thought and execution has disappeared because the approval gate is now an integrated part of the build process.

3. From Coordination to Synthesis

If your primary value is "knowing who to talk to" to get something done, you’re in trouble. The agents already know who (or which API) to talk to. Your value now lies in Synthesis. Can you take the chaos of the UK energy market, the rise of DeFi security risks, and Coinbase’s organisational shift, and synthesise them into a singular, actionable narrative? That’s a human skill AI amplifies but cannot replace.

Building the "One-Person Revenue Stream"

Brian Armstrong’s memo isn’t just a corporate strategy; it’s a blueprint for the modern individual contributor.

By applying these principles, I’ve built what I call the "One-Person Revenue Stream." I have a full-time job at Gentrack, but I also operate two independent startups—Lumen Pro and ProstoEnergia.

I’m not "doing more with less." I’m doing everything with nothing.

I have no engineers, no marketing team, and no pure managers. My agentic stack handles the URE tariff scraping, the GB grid intelligence snapshots, and the content drafting. My role is to be the "Expert Individual Contributor" who holds the context and directs the fleet.

This is the promise of the Post-Management Era. It’s not about the "end of work." It’s about the end of the toil of work. It’s about a world where the only thing standing between an idea and a production-grade revenue stream is the quality of your context and the boldness of your orchestration.

The Coordination Tax has been abolished. It’s time to start shipping.

Tags:

aistrategymanagementagentic-workflowscoinbaseorganisational-design

Recommended Reading