Hacker News Reader

Comment by jyrkesh | Hacker News Reader

I agree with the central thesis of the article (I didn't watch the video...), that PLG is basically inevitable for developer tools, particularly given the way that devs push for adoption of tools that they personally like from within an organization. Cloudflare has done a particularly good job of this, and as upstarts, I'm really optimstic about stuff like Tailscale and Fly.io

But I also think that startups and VCs alike are going to have be a LOT more reasonable when it comes to valuations. You can build very sustainable, profitable businesses with PLG, but it does usually require burning a lot of money up front, and won't likely end in a Cloudflare scale home run. And I've seen firsthand how just getting a _little_ bit wrong with your free vs. premium tiers can result in some significant and unforeseen infra charges.

View on HN

Great points and thanks for the tailscale mention. That is exactly the product we needed this week. One thing not mentioned is how to make product led growth accelerate by word of mouth and how to encourage that. It’s the key to acceleration imho. That is hire to get a network effect. From the last century but this was how AWS started our Blackberry if you go farther back. The value proposition is individual to team to enterprise but how do you get team success to multiply. The term for this was influential end user marketing. :-). Who you get is as important. The opals trope was are these people who give more advice than they receive. Those are really important folks :-)

>And I've seen firsthand how just getting a _little_ bit wrong with your free vs. premium tiers can result in some significant and unforeseen infra charges.

Can you elaborate?

I don't know if I got in me to write out a ton here--I probably owe this sort of thing a blog post or write up at some point--but I'll just say that marketing _really_ likes terms like "unlimited", sales loves to sell committed and contractual spend, and engineers love to automate (particularly in a PLG motion).

So you get things like...a lot of compute or bandwidth spend. (You know, I tried to think of other examples, but at some point it really does come back to the compute and bandwidth.)

And in a up-or-out kind of growth mentality, it's easy to handwave a lot of that away as just growth spending. But if the returns don't kick in big time at the end, you've just accelerated your runway.

>And I've seen firsthand how just getting a _little_ bit wrong with your free vs. premium tiers can result in some significant and unforeseen infra charges.

same question here - would appreciate some war stories here.

Reply on news.ycombinator.com