MMarketing Against The Grain
← All frameworks
Strategy

The Zero Human-Cost Thought Experiment

Redesign customer workflows as AI drives task costs toward zero

Difficulty
Moderate
Time to result
~weeks to results
Steps
6
Confidence
94%

The Zero Human-Cost Thought Experiment asks what a company would do if AI could perform a defined task as well as a human for almost no marginal cost. Leaders first identify activities whose economics currently limit coverage, such as supporting every customer or engaging low-intent contacts. They then imagine removing the labor-cost constraint and redesign the experience rather than merely cutting headcount. The model also requires examining second-order effects: if every competitor gains inexpensive outreach, inboxes saturate and attention becomes scarcer. The useful output is therefore a prioritized pilot that expands service or productivity while preserving human oversight and preparing for market-wide behavioral changes.

Origin

Extracted from Marketing Against the Grain as a strategic mental model for identifying AI-enabled operating opportunities and their second-order consequences.

Core principles

  • 01Model AI as near-zero marginal labor for suitable tasks
  • 02Look beyond headcount reduction to newly viable service levels
  • 03Apply capacity where traditional unit economics prevented human coverage
  • 04Consider second-order effects such as channel saturation
  • 05Reserve humans for judgment, creativity, and exceptional cases

How to run it

  1. 1

    Map costly tasks

    List recurring tasks where labor cost currently restricts frequency, speed, or customer coverage.

    Pro tip Start with high-volume workflows that already have clear inputs and outputs.

    Watch out Do not assume every human activity can be reduced to a stable task.

  2. 2

    Test the capability assumption

    Determine whether current AI can perform each task to an acceptable standard under human supervision.

    Pro tip Use real examples and edge cases rather than a polished demo.

    Watch out Near-zero cost is irrelevant if error rates create greater downstream costs.

  3. 3

    Remove the cost constraint

    Ask how the workflow would change if each acceptable AI execution cost almost nothing.

    Pro tip Look for customers, leads, or service moments that were previously uneconomic to address.

    Watch out Avoid limiting the exercise to replacing existing employees.

  4. 4

    Choose the value allocation

    Decide whether to convert the gain into lower cost, faster growth, better service, or reassignment of people to higher-value work.

    Pro tip State the intended benefit before deployment so savings do not disappear into general operations.

    Watch out Efficiency gains do not automatically create customer or business value.

  5. 5

    Model second-order effects

    Consider how competitors adopting the same capability will alter attention, channel quality, expectations, and differentiation.

    Pro tip Ask what becomes scarce when execution becomes abundant.

    Watch out A tactic that works early may become table stakes or harmful after widespread adoption.

  6. 6

    Pilot and monitor

    Run a bounded deployment, measure economics and experience quality, and escalate unusual cases to humans.

    Pro tip Track both direct savings and customer outcomes.

    Watch out Do not scale before validating safety, accuracy, and escalation paths.

In the wild

Serving a cold database segment

A company historically ignores contacts who have shown no activity for 90 days because human outreach would not recover enough revenue to justify its cost. AI makes individualized re-engagement inexpensive enough to test, with humans handling positive or unusual responses.

The company can explore previously uneconomic demand while keeping human effort focused on qualified opportunities.

Universal support availability

A business trains an AI support system on its knowledge base and internal data so customers can receive immediate answers through chat, email, and eventually voice. Human agents manage exceptions and sensitive cases.

More customers receive timely support without requiring a proportional increase in staffing.

Common mistakes

Equating the model with layoffs

The thought experiment can support workforce reduction, but it can also redirect people toward higher-value work or improve customer experience.

Ignoring attention saturation

When every company can generate inexpensive outreach, message volume rises and customer attention becomes more valuable and difficult to earn.

Assuming capability before testing

The framework only applies where AI can perform the task at an acceptable quality level; many complex or creative tasks are not there yet.

Is it for you?

Best for

Leaders redesigning service, support, sales, or marketing workflows around dramatically lower marginal task costs.

Not ideal for

Tasks requiring empathy, accountability, nuanced judgment, or reliability beyond current AI capabilities.

From the transcript

one of the ways to think about Ai and how you appli to your company we're going to get into the examples um is what…

Kieran · 10:30

when a AI can do a task in a similar way to a human how would you apply AI to your business

Kieran · 11:30

you have unlimited human capital through AI why not try to improve the experience of your customers

Kieran · 12:00

From the episode

Watch These 33 Minutes To Become An AI Expert