Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

A Failed SaaS Postmortem

mattlayman.com

Post reply on HN
Print view View original

71–80 of 141 posts

Re: A Failed SaaS Postmortem

#71

Post by Alex3917 »

briandoll wrote: post #2

The main point here seems to be that the failure to build a SaaS product is blamed in part on the compounding technical debt incurred due to a fast-moving software ecosystem and the effort involved in keeping up. The point I think most people should consider instead, is that building an app is not the same thing as building a business. Building the app is comparably easy, once you've proven that you understand the ma…

> The app, it turns out, is the easy part.

I dunno, I don't know anything about self-driving cars, but when the people making self-driving car SaaS say the software part is difficult I tend to be inclined to believe them.

Even if these sorts of entrepreneurship cliches are true in most cases, I don't think it's a good idea to use them to write people off without doing any critical thinking.

Re: A Failed SaaS Postmortem

#72

Post by phonon »

powowow wrote: post #60
streetcat1 wrote: post #35

Earlier quoted context omitted.

Hard to develop tech is a huge moat. The issue here is timing. Hard tech buys you time. Most of the M&A today is not done due to hard to replicate, but due to time. Even if I can replicate the tech, I do not have time to do that.

Can you name, say, three notable software businesses whose advantage was hard-to-replicate tech?

Ksplice, kdb+, Google search, Mathematica

Re: A Failed SaaS Postmortem

#73

Post by jjice »

spookthesunset wrote: post #65
madamelic wrote: post #16

Earlier quoted context omitted.

Or a barebones box from Vultr or Linode then log errors out to stdout, slap an email address on the page and there's your logging and error system. AWS is massively overcomplicated for an indie SaaS business, and it costs way, way too much. You can have a box with plenty of room for $5 - $10 / month and it takes about 25 seconds to launch it. I had 20k concurrent users with a $5 server and Cloudflare. Other people ar…

> Or a barebones box from Vultr or Linode then log errors out to stdout So instead of talking with customers and growing your business you are busy wasting your time installing a web server, database server, configuring build tools, etc? Getting root access to a barebones linux install is conceptually "not complex". But that is only because it pushed all the actual complex stuff up the food chain and into your head.…

I don't know if I've been doing something wrong now. For my personal web apps, I've always just installed Nginx, Let's Encrypt, and then a docker image of my application/DB. Spinning this up on a fresh Ubuntu Server install takes less than a half hour if I had to guess. Nginx to force TLS and proxy pass to my docket image.

I'm not trying to be rude, just genuinely curious if I'm missing something complex that could be making my services less secure.

Re: A Failed SaaS Postmortem

#74

Post by tnolet »

jjice wrote: post #73
spookthesunset wrote: post #65

Earlier quoted context omitted.

> Or a barebones box from Vultr or Linode then log errors out to stdout So instead of talking with customers and growing your business you are busy wasting your time installing a web server, database server, configuring build tools, etc? Getting root access to a barebones linux install is conceptually "not complex". But that is only because it pushed all the actual complex stuff up the food chain and into your head.…

I don't know if I've been doing something wrong now. For my personal web apps, I've always just installed Nginx, Let's Encrypt, and then a docker image of my application/DB. Spinning this up on a fresh Ubuntu Server install takes less than a half hour if I had to guess. Nginx to force TLS and proxy pass to my docket image. I'm not trying to be rude, just genuinely curious if I'm missing something complex that could b…

Heroku does that in 3 minutes. And maintains & updates that stack for you. And gives you deployment automation etc. For $7 a month.

Re: A Failed SaaS Postmortem

#75

Post by ericb »

jbob2000 wrote: post #67
ericb wrote: post #58

Earlier quoted context omitted.

You're not wrong about the size of the dev team and turnover. At the same time, I'm not sure if taking 2 months to get productive is a win? Wasn't one of the angular "upgrades" a total rewrite? What is the overhead of maintaining a second tightly coupled front-end app. I found doubling the surface area seemed to square the effort, not double it. How much of that productivity is illusory and could be done with fewer p…

I work for an enterprise company with more than 70,000 employees. Standing up a team and getting them productive in 2 months is basically unheard of at this level. This community is heavily biased to small companies, startups, and students. I'd be hesitant to judge their approval as affirmations that you are doing the "right thing".

> affirmations that you are doing the "right thing."

I'm not really looking for approbations--I'm happy to play this business game with a speed optimized team and see how my strategy fairs on merits against competitors maintaining two apps to my one. I'll make that bet all day long.

I'd say hold up and question your assumption that there is a "right thing" at all.

In terms of the perception of what is the "right thing", and the startup bias of this place, that could be true. I'd add though, that there's a reason houses are not made of rebar reinforced concrete. It's too expensive to work and rework. From that, I'd generalize that certain architectural patterns work better in some cases, and in others they are overkill. Wood works fine, and it is faster and cheaper to build with. And if you have to knock it down, you can rebuild the next one for cheap too. Neither is "right."

Re: A Failed SaaS Postmortem

#76

Post by spookthesunset »

tnolet wrote: post #69
spookthesunset wrote: post #62

Earlier quoted context omitted.

I was guilty of this. Spent far too much time optimizing database indexes and far too little time actually talking with customers and validating new ideas. Technology choice is the least important part of a startup. Pick what you know and roll with it. Make sure the first developers you hire "get" startups and aren't in it to tinker with their pet projects or push some bozo agenda (eg: don't use js frameworks because…

The "bozo agenda" is a great point about early tech hires. I bumped into a few over the years and they can totally kill your dev team and young business.

Tell me about it. It is even worse when they are protected by the founder, who is either non-technical and considers the bozo a "rockstar" or is technical and brought the bozo in because of past relationships (friend, coworker, etc).

The bozo gets protected by the founders and all future developers have no career path, no say in the technology, and get to swim in the sea of an opinionated hot garbage codebase.

People who leave big companies for startups to escape the politics eventually get a rude awakening. Startups, especially very small startups, have a whole different set of frustrating politics. Not even sure which is worse to be honest.

Re: A Failed SaaS Postmortem

#77

Post by planetzero »

pbreit wrote: post #18
throwno wrote: post #7

Earlier quoted context omitted.

The only problem with not caring about the technology, is if your solution is trivial to replicate then anyone with a bunch of cash can come in with hired gun coders and out advertise you. Hard to replicate technology gives you some kind of a moat to at least slow down copy cats. edit: Oh, I see, the guy in the article was just throwing together random open source stuff not actually developing anything of value. Neve…

Never, ever, ever, ever worry about anyone copying your stuff. Depend on your unique insight. No one can copy that. If in the extremely rare circumstance that someone copies you, be thrilled! That means you are building something valuable (which is extremely rare).

This is bad advice. If you have a really good idea and a company in the same space with the money and resources to beat you to market sees it, it's very possible they will bring it to market before you.

Large companies like Google and Facebook will regularly feign interest in someone's app or idea (that they might want to purchase), only to roll it out themselves.

Re: A Failed SaaS Postmortem

#78

Post by 3pt14159 »

ericb wrote: post #8

We just did something crazy. We dropped the Ember app that was 3/4th's done in favor of plain JQuery and Rails 6. I'm just done maintaining two apps instead of one and all the hell that involves. And the Javascript ecosystem can keep rolling that Sisyphean upgrade treadmill with deploy dependencies and promises everywhere and I promise to use it as little as possible. Because it is just a giant time suck. Tut tut if…

I have a couple of questions for you:

1. Was your Rails app solely an API or did it also serve HTML/CSS on first load?

2. Did you build a reusable component framework for your Ember app?

3. Did you follow something like JSON API[0] and use EmberData to hook into it?

I've found that Ember apps with a Rails backend API actually fit together pretty nicely. Don't get me wrong, I like simple. I don't even have a single line of JS on my personal site nor do I use any libraries to generate it (I wrote my own static site compiler), but I've seen JQuery in practice and it makes me unhappy as the app grows because so often things get inconsistent in terms of UI. Plus there seems to be less overall structure of the frontend.

[0] https://jsonapi.org

Re: A Failed SaaS Postmortem

#79

Post by spookthesunset »

tnolet wrote: post #74
jjice wrote: post #73

Earlier quoted context omitted.

I don't know if I've been doing something wrong now. For my personal web apps, I've always just installed Nginx, Let's Encrypt, and then a docker image of my application/DB. Spinning this up on a fresh Ubuntu Server install takes less than a half hour if I had to guess. Nginx to force TLS and proxy pass to my docket image. I'm not trying to be rude, just genuinely curious if I'm missing something complex that could b…

Heroku does that in 3 minutes. And maintains & updates that stack for you. And gives you deployment automation etc. For $7 a month.

Exactly. Deployment is merely a "git push" away. It is amazing how simple it is to get started building a new product.

Every second you spend configuring web servers and stuff is a second you should be spending adding value to your business.

Re: A Failed SaaS Postmortem

#80

Post by streetcat1 »

powowow wrote: post #60
streetcat1 wrote: post #35

Earlier quoted context omitted.

Hard to develop tech is a huge moat. The issue here is timing. Hard tech buys you time. Most of the M&A today is not done due to hard to replicate, but due to time. Even if I can replicate the tech, I do not have time to do that.

Can you name, say, three notable software businesses whose advantage was hard-to-replicate tech?

Yes. Any tech acquisition over 1B is never done for customers. For example deep mind from google.

Just this week Intel bought Havana labs for 2B due to their AI chip hardware.

Also, look at what AMD is doing to Intel. This is pure tech play (and pure moat).

The underlying rule in play, is that the tech business is bound to "leapfrogging", I.e. a better technology can destroy companies almost overnight (see the case of Nokia). So hard tech is more important than customers.

Look what Tesla is doing to GM, etc.

Post reply on HN