Frequently Asked Questions About Greg Isenberg AI Opportunity Scanner
21 answers covering everything from basics to advanced usage.
// Basics
What are Boring Gold Mine Verticals in Greg Isenberg's framework?
Boring Gold Mine Verticals are industries with legacy workflows — phone calls, faxes, 30-year-old processes — that are ripe for Vertical AI disruption. Examples include insurance, legal, logistics, eldercare, government, accounting, and construction. The more boring and niche the industry, the better the opportunity. Isenberg advises picking a wedge in a subniche rather than tackling an entire category, because well-funded startups will own the top-level categories.
What is the Scarcity Flip and how does it affect my business positioning?
The Scarcity Flip is the reversal of what's valuable as AI commoditizes execution. Generic content, basic design, data entry, and routine analysis become cheap and abundant. Creative judgment, human-made crafts, physical experiences, original thinking, and proprietary data become scarce and premium. For positioning, Isenberg's Premium Stack ranks: (1) human-made with no AI, (2) AI-assisted but human-led, (3) fully AI commodity, (4) race to zero. New businesses should aim for tiers one or two to avoid the commodity trap.
How does the 100 True Fans concept differ from 1,000 True Fans?
Kevin Kelly's original concept required 1,000 fans paying enough to sustain a creator. Isenberg's AI-era revision drops that to 100 fans because agents cut operational costs so dramatically. 100 people paying $500–$1,000/month is a real, high-margin business for one person. Combined with Micro Monopoly Math, you only need a 5,000-person niche audience to extract 100 customers at $50/month for roughly $60K profit — then replicate across similar niches.
Is the Greg Isenberg framework only for tech founders?
No. The framework is designed for anyone evaluating business opportunities in the AI era — consultants, agency owners, solo operators, creators with audiences, and existing SaaS founders. The 1-Hour Company Stack uses vibe coding and no-code tools, so deep technical skills aren't required. What matters more is founder-agent fit: the ability to orchestrate agents toward a goal. Domain expertise in a boring niche is often more valuable than technical ability.
What is the Agent Economy and when does it start?
The Agent Economy is Isenberg's term for the 2025–2030 era, following the App Store era (2009–2015) and API Economy (2015–2024). In it, agents discover and hire other agents on the fly, fixed teams dissolve, and Gartner projects 20% of commerce will be agent-to-agent by 2030. The shift means building for agent consumers, not just human consumers. Products need to be discoverable and usable by other agents, not just people with web browsers.
// How To
How do I price an AI agent product using outcome-based pricing?
Identify the specific result your customer cares about — a resolved support ticket, a qualified lead, a reviewed contract, a confirmed appointment. Price that result directly. For example, $1.50 per resolved ticket or $5 per qualified lead. Only use seat-based pricing if the product is genuinely a human-operated tool. Usage-based pricing works for infrastructure plays. The key question: what outcome does the customer actually pay humans to produce today? Price that, and you capture labor budget instead of IT budget.
What is the Ghost Team Org Chart and how do I create one?
The Ghost Team Org Chart is a company structure with a few humans and named AI agents handling sales, content, customer support, analytics, and other functions. To create one, map every business function, determine which can be handled by agents, assign named agent roles, identify where human judgment is required (creative direction, relationship escalation, ethical review), and structure the business so you check in every few days. This is the organizational model for ambient businesses.
How do I evaluate founder-agent fit for myself?
Ask: Can I orchestrate a fleet of agents toward this specific business goal? Do I understand the niche deeply enough to direct agent output the way a film director directs actors? Test this by giving agents real tasks in your niche and evaluating whether you can spot errors, improve output, and maintain quality. If you consistently improve agent output through better direction, you have founder-agent fit. If you can't tell good output from bad, either deepen your niche knowledge or pick a different niche.
How do I run a quarterly agent cleanse?
Review every agent's permissions across the four axes of the Agent Permission Stack: access, memory, actions, and sharing. Revoke any permissions that aren't actively needed — treat it like reviewing app permissions on your phone. Check whether agents have accumulated access to sensitive data or systems over time. Review logs for unusual behavior that could indicate prompt injection or context poisoning. Update agent instructions to reflect current business needs. This should take 1–2 hours per quarter and prevents permission creep.
How do I decide between human-made, AI-assisted, and fully AI positioning?
Use Isenberg's Premium Stack. If your niche values craft, authenticity, and exclusivity — like luxury goods or artisan services — position as human-made with no AI, treating it like organic certification. If your niche values quality and speed — like professional services or content production — position as AI-assisted but human-led. If your niche is price-sensitive and commodity-oriented, go fully AI. Avoid race-to-zero positioning where you're competing purely on undifferentiated AI output with no moat.
// Troubleshooting
Can I use the AI Opportunity Scanner if I have no audience at all?
Yes, but you need to acknowledge that no distribution is your primary constraint. The 1-Hour Company Stack compresses to hours only if you have someone to sell to on day one. Without an audience, either build distribution first through content, newsletters, or social media, or accept that paid acquisition will compress your Micro Monopoly Math margins. The Scanner still works — it just flags distribution-building as step zero before anything else.
What happens if my idea fails the Micro Monopoly Math check?
If the math doesn't produce meaningful profit for one person, you have three levers to adjust: increase the price point by targeting a higher-value vertical where the outcome you deliver is worth more, increase the reachable niche audience size to improve conversion volume, or reduce agent operating costs. If none of these adjustments work, the idea may not be viable as a micro-monopoly and you should test a different niche. The 1-Hour Company Stack is designed for running multiple experiments, not committing to one.
How do I know if I'm building Vertical SaaS when I should be building Vertical AI?
Ask one question: are humans operating your tool, or are agents doing the work humans used to do? If customers log in, click buttons, and use your software as a tool, you're Vertical SaaS — capturing IT budget with a capped TAM. If agents perform the labor and customers receive results, you're Vertical AI — capturing labor P&L with a 10x larger TAM. The telltale sign you're in the wrong category: you're pricing per seat for a product where the agent does all the work.
What should I do if my niche is too small for even 100 customers?
Raise the price per customer. If your niche is only 200 people, you need 50% conversion at a higher price point to reach the same revenue as 100 customers at $50/month. Alternatively, expand to adjacent sub-niches and stack multiple micro-monopolies inside a holding company. Isenberg's model is designed for replication — one $60K micro-monopoly alone isn't the goal. Five or ten stacked across related niches is the architecture. A tiny niche is fine if the outcome you deliver commands premium pricing.
What's the biggest mistake people make when applying Greg Isenberg's AI frameworks?
Building without distribution first. The entire speed advantage of the 1-Hour Company Stack and the New Timeline depends on having someone to sell to on day one. Without an audience — even a small one of 500–5,000 — the framework's time compression evaporates and you're back to the old timeline of hunting for customers. The second biggest mistake is building in a top-level vertical category instead of a subniche, where well-funded competitors will dominate.
// Comparisons
What is the difference between the Greg Isenberg AI Opportunity Scanner and a regular business model canvas?
The AI Opportunity Scanner is purpose-built for the agent economy era, while a business model canvas is era-agnostic. The Scanner specifically evaluates whether your idea should be Vertical AI or Vertical SaaS, whether pricing should be outcome-based, whether agents can run operations, and whether you're in the SaaS Graveyard. A business model canvas doesn't address agent attack surfaces, the scarcity flip, or micro-monopoly math — all of which are central to AI-era business design.
How does the Greg Isenberg framework compare to the Lean Startup methodology?
Lean Startup focuses on build-measure-learn cycles over weeks or months. Isenberg's framework compresses that into hours using the 1-Hour Company Stack and assumes AI agents handle most execution. Both value rapid validation, but Isenberg's approach is more opinionated: it prescribes outcome-based pricing, vertical AI over vertical SaaS, micro-monopoly sizing, and agent-powered operations. Lean Startup is methodology-agnostic about pricing and team structure. Isenberg's framework also explicitly addresses the asymmetric time window — urgency that Lean Startup doesn't encode.
// Advanced
How do I protect my AI agents from prompt injection attacks?
Audit the Agent Permission Stack across four axes: what can the agent access (files, email, calendars, bank accounts), remember (conversations, personal and business data), do (send emails, make purchases, modify code, delete data), and share (with other agents or third parties). Apply minimum necessary permissions. Schedule quarterly agent cleanses to review and revoke permissions. Be especially cautious with agents that browse the web or accept external inputs, as hidden instructions in context windows are the primary injection vector.
Should I build in public if competitors can copy my idea?
Build in public only when your audience is your customer base. In that case, the co-builder flywheel — where users vote on features, you ship in days, and trust compounds — creates a moat that's harder to fork than the code itself. If your audience is not your customer, be selective about what you share to limit competitive signal while still building distribution. Community co-creation is a durable moat because even if someone forks your product, they can't fork your community's trust and involvement.
Can I apply the Micro Monopoly Math to a B2B enterprise product?
The Micro Monopoly Math is designed for niche, high-margin, one-person businesses — not enterprise sales with long cycles. However, the principles translate: outcome-based pricing works in enterprise, and the Vertical AI filter applies to enterprise verticals. For enterprise, adjust the math: fewer customers at higher price points. Instead of 100 customers at $50/month, think 10 customers at $2,000/month. The holding company structure still works — run multiple vertical enterprise micro-products with ghost teams.
How does building in public actually create a moat against copycats?
When your audience co-builds the product — voting on features, providing feedback, feeling ownership — they become invested in your specific version even if someone forks the code. Trust and community can't be cloned. The co-builder flywheel compounds: users become distribution, their input shapes a product competitors can't replicate without the same community, and switching costs increase with each iteration they've influenced. In a world where code is easily forked, community loyalty is the real defensible asset.