Comment by xwdv | Hacker News Reader

Comment by xwdv | Hacker News Reader

I think Lean Startup has fallen out of favor in recent times and hasn’t kept up with modern trends.

Lean Startup can be boiled down to: don’t waste time and resources on things that are not necessary or that people don’t actually want. Wow, such wisdom.

In old days, a lean startup looked much different, more rough and scrappy. Nowadays that doesn’t really fly, and everything comes with polished facades straight out the gate, or else you have no chance.

Lean startup was meant for a time when building big things upfront was difficult and time consuming. Nowadays if you know how to leverage technology, you can build impressive things very quickly, and you must, because if you don’t your competitors will and they will win because today’s consumer is too burnout from cheap looking startups that disappear overnight because they decide they had the “wrong assumptions” about their customers.

In a world where building things gets increasingly cheaper and easier, Lean Startup is too cautious and slow and has no value.

> if you know how to leverage technology, you can build impressive things very quickly

I worked at a start-up for 18 months at the start of 2020. I look at it as the best experience of "flow" I'm ever likely to have. I come from an infrastructure/devops background, the two other engineers were developers (one with significant management experience) and the rest were data scientists. Within a crazy short time we had CI/CD, separate environments for dev, testing and production, extremely scalable infrastructure in multiple regions. We set up kanban, daily standups, weekly design meetings. We considered logging and metrics from the start. Product development was insanely quick, and there were essentially no infrastructure blockers because I was able to produce IaC and deploy it as it was required. We put heavy focus on minimizing cost to increase runway (ie: scale to zero, free services like gitlab & Cloudflare). In under a year we had an MVP that we could deploy to any region from scratch with two commands (terraform apply and a script to helm install the rest).

I'll add one thing: while you can spin up impressive things quickly, it's easy to screw up. Like deciding you can ignore environments, regions and scalability "for now". You think you can do it later? You can't. Not when you've got a shitload of customers running in your one big janky data center.

I know. Right now I'm working for a company who did this. There are no naming conventions. Deploying a new data center is a nightmare. I want to throw out all of their shitty code and start over, but I can't. The company is doing well but the VCs want to see rapid expansion into different regions and they're simply unable to do it.

Build good foundations.

Thanks for sharing. What happened to the startup you worked on/at?

They're still doing well. I got to a point where I needed to take time out, spend time with family, do gardening, lie out in the lawn looking at the sky. I did that for 8 blissful months before a guy I know pulled me into this dumpster fire.

Agree very much with this sentiment. Lean Startup advise is quite outdated these days. Shipping a lean product... was a total disaster for us. Nobody wanted to adopt us since we were hardly competitive. We solved one problem but customers wanted the product to also do 100 things. We did 50 things and only then could we get customers to stick and even give constructive feedback.

You’re right.

Inevitably startups these days must be willing to waste some resource, maybe time or money, in order to truly iterate toward solutions that customers want. Anything else is too slow.

Preach. I agree with both comments. The reality is that we need at least something to differentiate ourselves from the competition and that cost money/resources/time.

The era of "ugly but functional to a certain extent" is dead. Customers want good solutions right out the bat.

It's also worth keeping in mind that the 'lean startup' methodology was mostly proven out by Eric in the games space, which is VERY different from trying to sell a product to businesses or doctors or whatnot. The way customers make purchasing decisions is different.

Lean startup is not about building something bad and ugly. Although the core process is named Build-Measure-Learn-Loop, it really starts with "Learn". You've got to define hypotheses about your business model (e.g. customer pain points) and build experiments to validate. A MVP is the product with the narrowest set of features that predictably converts a target into a paying customer. From the customers perspective, you think of a MVP as a minimal lovable product (MLP).

I think more people need to talk about "how lean should you be while still being competitive". I think that's really hard and that's very different depending on what the startup is doing. I think there's a lot of nuance / opportunity there to figure out "how lean is lean enough"

The core takeaway from the Lean Startup is about learning. Everything should be aimed at learning more about your business and how to make it successful. That point really has very little to do with building products; it's precisely the "wrong assumptions" bit that the approach targets.

Reply on news.ycombinator.com