Let’s talk about the Sufganyot economy: Eating half a Sufganya instead of a whole one is the biggest lie ever. You just end up eating 3-4 halves anyway. Do yourself a favor—eat a whole one, enjoy it, and move on. Stop lying to yourself. 🤣 Happy Hanukkah! 🕎

‘No, I don’t need more work. Just 1-2 hours will do.’ This is literally you asking for more work. 🤦‍♂️

Hardcoding your revenue forecast is useless. It shows nothing and it adds no value. The same is true for using TAM for revenue forecasting. The hard truth is only bottom-up analysis is helpful for you and investors. Might take more time, but it adds value to the discussion.

Numbers matter in AI. 2025 multiples analysis shows: • AI companies average a 23.4x revenue multiple (3x that of public SaaS). • Infrastructure tops the charts at 47.7x. • Cybersecurity & Health Tech are surging. These are clear signals of where innovation grows.

With all the cool new AI agents, the one that brings me the most value isn’t some shiny, futuristic tool—it’s Zoom’s AI companion. Used to juggle notes, keep up with calls, and make sense of everything. Now, Zoom’s AI summary does it for me. Worth every penny.

‘How is Finro better than other firms? Why should we work with you?’ Honestly, I’m not even sure I want to work with you. These questions are a red flag—usually a sign someone’s just window shopping. From my experience, they almost never close.

Also, the number of AI agents focused on managing or building other AI agents is mindblowing. At least half of the industry is trying to sell these internal tools to other players in the industry. Consolidation is inevitable

How do I spot a founder who came from Big Tech? They forecast their staff like project allocations: 0.3 customer support, 0.45 frontend… Hate to be the one to say it, but that’s not how hiring works in the real world.

If there’s one thing I learned this month is that I need to sign a contract before EVERY job. Even if a client I know for 10 years referred his best friend to me. I still need to sign a contract. Ugh.

The best part about valuing AI startups in 2025? Half the comps work is already done. Between LinkedIn humblebrags and TechCrunch features, everyone’s shouting their valuation, revenue, and user numbers from the rooftops. BUT, the interesting stuff is what they don’t post.

Apparently, AI tools for COBOL are a real niche—and in surprisingly high demand. Feels like future Marty McFly met past Marty McFly and launched a startup together.

Forget the M&A work, financial analysis, and endless research. I just reduced a client’s Excel file from 2MB to 700KB without breaking a single formula. Feeling absolutely almighty.

Every client before a board meeting: 3 weeks out: “Let’s build a solid 5-year model for meaningful, data-driven discussion.” 1 week out: “Let’s boost user growth and lower COGS.” 2 hours out: “Just put $10M in revenue and move on.”

My 6yo: Daddy, why aren’t you working? Me: I am working. 6yo: No, you’re not. You just talk to people and play on the computer. 🤦‍♂️