Strategic analysis · Agentic era

When software development costs go to zero, what market position is actually defensible?

AI agents can now build a functional CRM, ecommerce platform, or data warehouse for around $200 in less than a month. When the cost of software collapses to near zero, the software itself stops being the asset!

A Shopify-equivalent can be rebuilt in a week. Plug in Stripe's hosted checkout and you inherit every payment provider, fraud detection, and PCI compliance in one move — arguably better than what Shopify offers natively. The technology moat was never really the technology. It was the cost and friction of rebuilding. Remove that friction, and most SaaS businesses are left with one thing: the customers they already have.

Company Appears to sell Actually owns
Amazon Products Distribution + trust + attention
Shopify Ecommerce software Merchant base
Salesforce CRM Your custom workflows
Google Search Attention + trust

In every instance above, the software layer is collapsible. Which means every company in that middle column is exposed unless they've built something in the right column.

The 50 companies below are mapped against that question: when software costs go to zero, what do they actually own?

50
Companies shown
7
Genuinely defensible
12
Highly exposed
Data flywheel
Most common moat
#
Company
Health
Score
Defensive
Moat
Time to
Zero

Moat types

Data flywheel more users → better data → better product → more users
Regulatory / trust compliance and certification barriers that take years to earn
Network effects each new user makes the product more valuable for everyone
Physical infra real-world assets like data centres, cables, or hardware
Workflow lock-in deeply embedded in daily operations, painful to rip out
Distribution only no structural moat — just existing customer relationships

Fail or Die Index

🟢 80–100 Survive — structurally defensible
🟡 60–79 Adapt or Fall — clock is ticking
🟠 40–59 Failing — moat eroding
🔴 20–39 Dying — standing still
💀 0–19 Dead — most likely to die