Human-to-Logic Ratio and The 80 Percent Threshold

The Human-to-Logic Ratio and the 80 Percent Threshold are not alternative criteria — they are sequential ones. The ratio identifies markets where the economic gap is large enough to justify building. The threshold confirms that the revenue loop can be executed above the autonomous line. A market that passes one and fails the other is not an Arco market.

The dominant approach to market selection for agentic deployment is capability-first: identify what agents can do, find tasks that match, estimate the time saved. The selection criterion is technological compatibility, not market cost structure. A market with a favourable Human-to-Logic Ratio passes the capability test easily — it has high-volume, scripted, repetitive work that agents execute well. So does a market where that work represents twenty percent of gross margin rather than seventy. In the first market, displacing the labour cost with an agentic stack removes the primary cost driver and restructures the economics of the operation. In the second, it produces an efficiency improvement at the margin while the primary cost driver — capital, IP, network effects — remains unchanged. Without the ratio, both markets look equally viable to a capability-first evaluator. The Operational Arbitrage appears similar. The structural moat available is not. A Breakable Market is not any market where agents can work. It is a market where the Human-to-Logic Ratio confirms that capturing the labour cost creates a structural advantage, not a marginal one.

A market that passes the Human-to-Logic Ratio but fails The 80 Percent Threshold produces a specific and frustrating operational signature. The Workforce Arbitrage looks real at the task level — individual workflows execute faster and cheaper with agents than with human operators. The overall cost structure of the business does not shift. The reason is architectural: the Revenue Loop still routes through human decision points at enough handoffs that adding volume adds human cost proportionally. The Coordination Trap from the legacy market structure reproduces itself inside the autonomous competitor because the task architecture feeding the Revenue Loop was not redesigned — it was automated in place. Administrative Density persists because the coordination overhead required at each human decision point does not disappear when the tasks around it execute faster. The business competes on efficiency rather than on structural cost advantage. Those are different competitions with different durability.

The conventional agentic market entry move — identify a high-labour-cost market and deploy agents to the most automatable tasks — recovers a fraction of the available arbitrage without capturing it structurally. The 80 Percent Threshold is the line that separates structural capture from marginal recovery. A business operating at 65 percent autonomous handoffs still carries a coordination layer at 35 percent of its cross-departmental handoffs. That layer funds the Operational Drag that prevents the Revenue to Headcount Advantage from compounding with scale. The Task Tiers (T1/T2/T3) framework explains the structural cause: handoffs that fall below the 80 percent autonomous line are dominated by T2 tasks — tasks requiring contextual interpretation or human oversight that the current LLM capability cannot eliminate. The market failed the threshold because its revenue loop contains a T2 concentration the Human-to-Logic Ratio did not surface. The ratio measures the economic incentive. The threshold measures the architectural feasibility. Neither is sufficient alone. Together they are the Autonomous Business entry test.

Together the two instruments define what Arco names a Breakable Market in its fullest sense: a market where the Human-to-Logic Ratio confirms the economic incentive and the 80 Percent Threshold confirms architectural feasibility. The first autonomous entrant in such a market captures a cost structure the incumbent cannot replicate without paying the Rebuild Tax and the Legacy Liability that the incumbent’s accumulated architecture has produced. Most incumbents do not pay either. Memo #05 develops the full market selection framework: what the two conditions look like in specific legacy service markets, how the ratio and threshold are measured in practice, and what the structural moat looks like once both are confirmed. What Not to Build develops the failure modes: markets that pass the Human-to-Logic Ratio comfortably but fail the 80 Percent Threshold structurally — where the task composition or regulatory environment makes the threshold unachievable regardless of LLM capability improvement — and the specific structural reasons each market fails.

A business that has confirmed both conditions operates in a market where Architectural Certainty is a design property, not an engineering achievement. The task composition of the market guarantees that the system can reach the 72-hour MTTI threshold without T2 escalation as a structural feature of its Revenue Loop. The Stewardship Model operates in this context as the human role above the 80 percent line: one operator governs the exceptions the logic cannot process, while the system executes 80 percent of handoffs autonomously. Headcount Decoupling follows by design: volume adds compute requirements, not headcount requirements. The Labor-to-Compute Substitution that generates the full Workforce Arbitrage is structurally total in the Execution Layer, and the 20 percent of handoffs that remain human are handled at fixed steward cost that does not scale with volume. Memo #01 closes the argument precisely: the distinction between an Automated Business and an Autonomous Business is not a technology distinction. It is a market selection distinction, confirmed before the architecture is drawn by the two measurements this episode defines.

Technology changes what is possible. Market structure determines what compounds.