THE FUTURE OF SOFTWARE M&A I’ve looked at buying more than 100 software businesses doing $3M ARR or less in the last 3 months. And every time I evaluated one, I asked myself the same question: how fast could a competent founder rebuild this with Claude Code? This made me realize there are now only 2 kinds of software businesses. Type 1: Utility businesses. These are feature-first products with value in the code or interface. They can be rebuilt QUICKLY, depend on ads or creator posts for distribution, and have no STRONG forces (community, network, data moats etc) that strengthen over time.. If they were worth 3x ARR before AI, many might now be worth 1–2x because the core product can be rebuilt faster than ever. If someone can clone it in 48 hours with Claude Code, Cursor, and a few automations, the defensibility collapses. Doesn’t it warrant a LOWER multiple? Type 2: Compounding businesses. These are products where the value sits OUTSIDE the code. Distribution loops, workflow lock-in, proprietary data, community density, and network effects. A rebuild does not replace these dynamics, and they gain strength as they grow. These businesses will be worth MORE going forward because they survive the collapse in creation cost. If they were worth 3x ARR before AI, some maybe worth 6x because if you can really add features with Claude Code/Cursor etc faster then ever before/cut costs faster than ever before/grow with social faster than ever before Doesn't it warrant a HIGHER multiple? Type 1 companies will struggle to sell at meaningful multiples because they can be recreated faster than they can be acquired. ...