AI Green-Space Opportunity Test
Find newly viable opportunities by revisiting what was once impossible or too costly.
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 6
- Confidence
- 94%
The AI Green-Space Opportunity Test begins with activities a business or customer does not currently perform because they are too expensive, slow, specialized, or technically impractical. The team identifies the dominant constraint behind each abandoned idea, then evaluates whether AI has reduced that constraint enough to change the economics. Opportunities are ranked by newly created customer value, feasibility, niche specificity, and vulnerability to broad platform providers. The goal is not simply to add AI to an existing product. It is to uncover useful work that could not previously be delivered profitably, such as disposable applications, personalized services, or highly specialized tools. A small experiment then tests whether the theoretical cost reduction produces a real, repeatable outcome.
Origin
Extracted from Marketing Against The Grain during a discussion comparing AI's reduction of software-development costs with the printing press and industrial manufacturing.
Core principles
- 01Cost reductions create opportunities beyond simple efficiency gains.
- 02The strongest opportunities may be activities businesses do not perform today.
- 03Previously uneconomic niches become viable when production costs approach zero.
- 04A valuable use case starts with a real constraint, not with an AI feature.
- 05Incumbent workflows and new green-space opportunities require separate evaluation.
How to run it
- 1
Inventory Abandoned Possibilities
List customer experiences, internal capabilities, and product ideas that were rejected because they appeared impossible, too expensive, or too labor-intensive.
Pro tip Ask what the business would change if it had a magic wand.
Watch out Do not limit the inventory to improvements of activities already being performed.
- 2
Name the Binding Constraint
For each possibility, identify the specific cost, skill, time requirement, or technical dependency that prevented execution.
Pro tip Express the constraint in measurable terms whenever possible.
Watch out A vague claim that something was difficult is not enough to establish changed economics.
- 3
Test the AI Cost Shift
Determine whether current AI capabilities can automate or compress the binding constraint. Estimate the new marginal cost and required human oversight.
Pro tip Build a narrow prototype rather than relying on a feature demonstration.
Watch out Do not treat AI output as costless when review, integration, or error correction remains substantial.
- 4
Locate the Green Space
Favor opportunities that enable genuinely new work, underserved niches, or use cases that were previously unprofitable.
Pro tip Look for small markets that become attractive when a tiny team can serve them well.
Watch out Broad consumer features may be vulnerable to direct competition from platform providers.
- 5
Validate Customer Value
Test whether the newly viable capability solves a meaningful problem for a defined customer group. Measure willingness to use, pay, or change behavior.
Pro tip Start with one customer segment and one outcome.
Watch out Technical feasibility does not prove market demand.
- 6
Recalculate Defensibility
Evaluate whether customer relationships, niche expertise, proprietary context, or superior service can protect the opportunity as AI becomes widespread.
Pro tip Treat ownership of the customer experience as a potential moat.
Watch out Do not base defensibility solely on access to a general-purpose model.
In the wild
A small distributor repeatedly reconciles orders from two incompatible systems but cannot justify commissioning conventional software. A developer uses AI-assisted coding to create a narrow application for that single workflow, with human review for exceptions. The reduced build cost makes a previously uneconomic application practical.
→ The distributor eliminates repetitive reconciliation work without funding a broad software project.
Two founders identify a specialized local trade with only a few thousand potential customers. Instead of pursuing a mass-market platform, they build scheduling, quoting, and customer-support tools tailored to that trade and use AI to keep development and service costs low.
→ A small addressable market supports a profitable focused business because its cost base is dramatically lower.
Common mistakes
Adding AI to Existing Features Only
Incremental improvements can be useful, but they overlook work that has become possible only because the underlying economics changed.
Assuming Zero Total Cost
AI may reduce marginal production costs while leaving meaningful expenses for validation, integration, support, and risk management.
Choosing a Niche Without a Moat
A narrow audience alone is not defensible if a general platform can provide the same experience without specialized context or relationships.
Is it for you?
Best for
It is best for founders and business leaders searching for products, services, or capabilities that were previously uneconomic.
Not ideal for
It is not ideal for teams that have not identified a meaningful customer problem or operational constraint.
From the transcript
“Where the most opportunity is going to be, whether it be for a business using AI or for somebody building an AI business, it's gonna…”
“What are the things that like, if we had a magic wand, would change our business, because some of them might be possible today now.”
“You can have applications just for your own personal needs because the cost to actually create those things are zero.”
From the episode
OpenAI Launches The ChatGPT AppStore (#105)