swedonym

I'm not usually one for soapbox moments, but this matters. The future of innovation depends on new ideas from new players, not just whatever suits a few giant corporations.

This post is an engineer’s guide (in plain English and a bit of math) to the startup metrics that VCs and founders obsess over. We’ll cover the essentials: how fast you’re growing, whether users stick around, how long your cash will last, and if your business model makes any money. By the end, you’ll see why 100 delighted daily users beat 1000 drive-by signups, and why these metrics have become my new “performance specs” in evaluating startups.

Thanks for the kind words! Would love to read your piece too if you want to link.

I recently found myself this close to trading in my role at the firm for an early seat on a shiny pre-Series A Sequoia-backed AI startup in the consumer space. The opportunity seemed like a rocketship - exciting tech, star investors, that intoxicating “get in on the ground floor” energy. Why on earth would I hesitate? Well, spoiler alert: I didn’t go through with it. In this post, I’ll candidly share why I nearly jumped - and the reality check that made me pump the brakes.

Google is an interesting example.

I'd say brand awareness and being the default entry point to information is probably their strongest "moat" for search. I'd argue that this moat is being eroded in the form of competition from ChatGPT, Claude, Perplexity. Their primary revenue generator (Ads supported search) is being chipped away at by competition and they've had to pivot (LLM based search) to keep up.

I agree, unique data assets allow them to stay ahead of the competition. They are in a much more tenuous position these days though.

For new entrants in these markets, the distance between idea and implementation is getting smaller and smaller at the same time that the big players are doing more land grabs. It's quite an interesting dynamic!

We’re in an era where spinning up a new app is easier than ever. With AI-assisted coding, no-code tools, and cloud infra, the barriers are down. If your moat is “we write complex code,” there’s an AI and ten hungry devs who can match you by next Friday.

The only real constant is motion. In 2025, software moats are not ditches. They’re treadmills. Keep running.

Leverage is earned.

It doesn’t come from being early to a trend. It comes from going deep where others are shallow. Look for the boring edge cases. Listen for the expensive mistakes. And build tools that compound - not just scale.