
TL;DR: „A website scores SaaS products on whether Claude can replace them with a .md file. Sounds absurd – but it hits a nerve."
— Till FreitagIn 30 Seconds
deathbyclawd.com is the "SaaSpocalypse Survival Scanner": enter a SaaS URL and get a "Death Score" – how likely it is that Claude replaces your product with a Markdown file. Scores range from IMMORTAL (Google, Cisco) to SWEATING (Intuit, Shopify, Adobe). Satire? Yes. Completely off? No.
What Death by Clawd Actually Measures
The site analyzes SaaS products across several axes:
| Score Range | Status | Meaning |
|---|---|---|
| 0–10 | IMMORTAL | Physical infrastructure, hardware, irreplaceable moat |
| 11–20 | SAFE | Complex integration ecosystems, regulatory depth |
| 21–30 | SWEATING | Core function fundamentally replicable by AI |
| 31+ | DEAD | Pure UI-wrapper SaaS – Claude can do this as a Skill |
The best quotes illustrate the logic:
"Intuit spent 40 years building a tax empire only to watch Claude read the entire IRS codex in 0.3 seconds."
"You can't replace Microsoft with a .md file, but Microsoft is absolutely trying to replace YOU with one."
Why This Matters
The SaaS Disruption Is Real
Behind the satire lies a serious observation: Many SaaS products are essentially UI wrappers around logic that an LLM can replicate. This particularly affects:
- Form builders – Claude can generate and validate forms
- Reporting tools – data analysis is a core LLM competency
- Content management – Markdown + AI = functioning CMS
- Simple CRMs – contact data + automations as a Skill
What's Not Replaceable
The "IMMORTAL" scores show where the line is:
- Physical infrastructure – Google (data centers), Cisco (cables), Marriott (hotels)
- Regulatory depth – ADP (payroll), CrowdStrike (cybersecurity)
- Network effects – Booking.com (hotel inventory), Netflix (content licenses)
- Hardware lock-in – Palantir (government systems), Palo Alto (firewalls)
The Vibe Coding Connection
Death by Clawd perfectly illustrates what we observe in Vibe Coding: The barrier to building software is approaching zero. If an LLM can generate a working tool as a Markdown file, the question is no longer "Can we build this?" but "Why are we paying for it?"
This doesn't mean all SaaS products die. It means SaaS products whose only value is a pretty UI have a problem. Value must come from elsewhere:
- Data network effects – more users = better product
- Compliance & certifications – regulatory moats
- Integration ecosystems – deep enterprise IT entanglement
- Proprietary algorithms – non-trivial logic an LLM can't simply replicate
What This Means for Tool Selection
When evaluating a SaaS tool today, add a new question to your evaluation framework:
Could an LLM replicate this tool's core function as a Skill?
If yes, consider:
- Are you paying for the logic or for the UI?
- Is there a data moat that makes the tool irreplaceable?
- How deep is the integration into existing systems?
Conclusion
Death by Clawd is satire – but the best kind: the kind that hits a true nerve. The SaaSpocalypse isn't "AI replaces all software." It's: AI lowers the barrier so much that SaaS products must prove their value beyond the UI.
Three takeaways:
- UI-wrapper SaaS is threatened – if the core logic fits in a prompt, the product is replaceable
- Moats matter more than ever – data, compliance, network effects are the new defense
- Tool selection needs an AI question – "Can an LLM do this?" belongs in every evaluation
→ What is Vibe Coding? → Tool Selection: Needle in a Haystack → Get in touch
What's Your Death Score?
5 questions – and you'll know if Claude can replace your SaaS.








