Seth Levine: Sick of start-up BS
finance.fortune.cnn.com
finance.fortune.cnn.com
When I give a talk and my bio says J2EE I'm a bit sheepish, and all the twenty-somethings are like LOL you use java, but that's silly - I work for a better, stronger business than all the little startups[1] emailing me on careers.stackoverflow. And the money is better. And we dodge the whole class of issues that come with taking money[2], like being told by your VC to collect and sell personal information.
That's why I'm still here. I believe in this business, and when interviewees ask the founder about the business model, the answer isn't "freemium" - its how we've been paid to contract a product for a huge client but retain IP rights so we can reinvest the profit to productize something that the entire industry wants. and we have assets (domain experts, software, client relationships, functional sales team) that a startup doesn't have - there are tremendous barriers for competition to overcome. we're halfway through this, and it's working, and when you look at it you're like, duh, of course its working.
[1] i don't want to pick on anyone, and there are a few awesome startups out there that I'd love to work for, and maybe yours is in this set, and maybe i'm not good enough for the ones in this set. What I do see is that the startups emailing me seem to need a decent rails guy to build their commodity webapp. rails apps are important, and i'm sure they have all sorts of challenges which i'm not aware of, but they aren't particularly interesting, to me. J2EE on a good team is about building out < http://c2.com/cgi/wiki?WhyIsPayrollHard >, which is in a whole new class of scale/complexity that is, well, the type of hard which causes languages like Clojure and Scala to exist. [2] http://37signals.com/svn/posts/3107-99-problems-but-money-ai...
edit: i'm getting lots of voting activity in both directions, it surprises me that this is controversial, i'd love to hear criticism.
You shouldn't feel sheepish at all. With how the client side JS frameworks are shaping up I think for many apps the optimal architecture is going to end up being JS in the browser calling web services in the backend written in Java/.Net ecosystems.
First product was using GWT to communicate with a special type of Servlet (GWTServlet) via XML-RPC. We definitely can substitute the back-end to use plain Servlet or JAX-RS or SpringMVC and tweak the front-end a little bit with no problem.
Second was an enterprise project utilizing IBM WebSphere Portal (JSR-286, or... if you're not familiar, take a peek at Liferay) in which we implemented 90% of it using AJAX to talk to a Portlet (like Servlet) that returns JSON.
One of the components was implemented using JAX-RS that can easily switch between XML or JSON or ATOM without having to write a single line of logic other than annotating the method to return any or all of the above format (JAX-RS uses content-negotiation).
You'll be surprised how easy it is to implement such set-up in Java. It is built with modularity in mind but you can always use SpringMVC if you'd like to have something more like Rails as well :).
Java is an overall great language. While it does have it flaws, it helps protect you from lots of stupid programming mistakes and it has pretty good performance.
The best part of GWT is the testability aspect via MVP pattern if you compare it against weiting in pure JS which is usually hard to write headless test automation.
We struggled for a long time, and our business came perilously close to the end a few times. But you know what got us out? Funding. VCs. Amazing investors that were thoughtful, supportive and value accretive. The marketing that we could invest in with our funding. We're doing millions of revenue now, and it is all because we were able to light the rocketship with our VC fuel. And did I say our VCs became great friends and mentors, even through the tough times?
It's not for everyone, and my experience may be an outlier. But let's stop with the snarky ant-VC comments.
I agree with the rest of your points and good luck with the pivot!
> ... equate raising VC money with selling your soul... one opinion... Raising VC is good for some businesses, bad for others
one interesting datapoint is that most* YC startups accept the Yuri Milner offer (auto 150k angel round). not that all startups are like YC startups, and that all businesses should take money, and that taking money/aiming for the stars is what defines a startup, but I do feel a bit of a stigma of "if you aren't a startup you and your company suck". As I get better acquainted with Dunning-Kruger I wonder if all those YC founders are mortals after all, just like you and me. or maybe i'm on the wrong end of it again ;)
*no source, read it 6 months ago
edit: this doesn't even make any sense, sorry haha
Really? I find the opposite. Every "we're hiring" thread brags about the company being "VC funded" and people seem to fall all over themselves figuring out the best pitching strategies or slide deck appearances.
I always find it odd, because being VC funded isn't good or bad in and of itself. There we agree.
As said before: The fact that people "celebrate" a funding event is really bass-ackwards when you think about it.
It says that the company couldn't figure out how to grow without bringing in a bunch of financiers, who have a low hit rate (3 go north, 3 go south, 4 turn into the living dead), who provided negative 10 yr returns even with Google in the portfolio, and who take 2-3% + 20% of exit from their own investors.
Definitely an excuse to drink, but not for reasons of celebration.
> And did I say our VCs became great friends and mentors, even through the tough times?
If you consider them among your friends, you have not hit the tough times.
They are your business partners. And when it comes time to sacrifice you in order to make some money you will be sacrificed.
I've been using JavaEE 6 (with Spring MVC) for a while and I believe it has improved a lot (and a lot simpler) compared to the previous versions and perfectly suitable for modern software systems: MVC (Spring), API json/xml (JAX-RS supports this easily), superb Testing tools (can mock almost everything), Asynchronous like Node via Servlet 3.0, WebService via JAX-WS easily (less than 5 minutes to produce XSD/WSDL or to generate client stubs).
I'm also impressed with the tools such as Maven for dependency, build, and package management, FindBugs for code analysis, CheckStyle for code-style enforcer, Hudson for Continuous Integration, and JavaDoc for code documentation that Java ecosystems has provided in order to write high quality software.
Why do VCs love startups? Because you can create value where there is none. Somehow 1,000,000 users capable of generating $200,000 in ad revenue per year (never mind the startup is burning 2 million a year!) are able to create valuations of $50 million. But hey, hot investors X, Y, and Z are in on this deal so it must be good! And thus the creation of companies no one gives two shits about continues.
How many startups are actually profitable - that is, genuinely creating more cash than they are burning? I think I can count them on one hand.
Surprisingly less true than you might think. I am friends with many founders that are cash flow positive!
Also, I haven't yet seen many of these crazy valuations. I mean, I see a few in TechCrunch every now and then, but the consistency of the valuations I'm seeing seems to be rare compared to others' experiences.
There may be "hard work" going on in the start-up world, but you'll have to seek it out. For every "serious" start-up you'll find 5-10 Zynga/foursquare/etc... clones.
Ah, but then they wouldn't be your founders would they?
Problem is, it's unavoidable. These guys have to get their name out there as quickly and loudly as possible. It's unfortunate, but until everyone in your target market has heard of you, you can't stop screaming "Look at me!"
It is a complete separate thing when you either haven't produced anything, or have produced, but it is far from remarkable, and you are being noticed by running around screaming.
It is the latter that gets tiresome and old. Not to mention the former is so much more honerable.
But pragmatically this is how small companies make a lot of money now. Run around screaming, "pivot" your non-business until it's the lowest common denominator buzzword compliant non-product and get an exit.
It's like the Hollywood sitcom model. It's basically focus group driven and hit based. Your agent and connections are more important than acting ability. The market created around start-ups cares way way more about "connectors" than introverted technical talent. And you can't blame them, web development tools have gotten very good and there are all kinds of new niches to fill on the web. You run the numbers and they are worth more.
In many other career paths there are various different cultural norms that push back against this kind of pandering, with varying degrees of success.
If you think that the engineer personality type is well suited to push back against extroverted suits promising popularity, wealth and calling them "rockstar programmers" because "you geeks are cool now" in order to keep working on big problems or doing something honorable then you have more faith than I.
So I hope that we find a way to correct the perverse incentives in place but the rest of the world would also love to hear a solution to this problem.
He's worried about this now?! 10-year returns are negative (and that includes funds lucky enough to have google!)
It's been that way at least since the 90s internet bubble.
Not that I disagree with anything said, but it is really easy to put your head down and build good products that eventually fail because no one knew about them. Just a thought.
This is so insanely untrue. There are tons of great products that don't see the light of day because the people in charge of said product don't have the skill set to find the people that would be interested in them.
These days the difficulty in building a great product vastly outweighs the difficulty in finding an audience to validate it. Both skills can - and must - be learned, but these days, someone who is able to learn how to build a great product will find it comparatively easy to learn how to find an early-adopter audience for it.
This is what YC is talking about with their motto "make something people want".
Hype can just as easily destroy a product by setting expectations too high and the product doesn't deliver. Think of how many movies have fallen into this trap?
You have to strike the perfect balance of sparking just enough interest while still delivering something of interest.
To find the quiet, experienced, skilled, well-adjusted people that aren't out drinking/partying once a week you have to actively seek them out while working. If you meet anyone like this, stay in touch. They are, by definition, not out there at events.
There's probably a fair amount of "introverted, seasoned professionals" that stop by HN from time to time. But then they reply with something that pisses the clique off, get down voted and never return.
Now in all seriousness, of course the focus should be on product. But having a ton of relationships to leverage doesn't hurt if the end goal is the success of the business. (and I love that video!)
Money is sloshing around, and nobody wants to earn 2% on long term junk government paper.
$500,000 isn't what it used to be, literally, but it buys a lot in the way of a web start-up now. If that $500k is now really $250k in 1999 terms, it's buying more like $5 or 10 million in 1999 terms (when it comes to the infrastructure needed to get a web start-up off the ground).
I'd normally say the start-up bubble would continue until the next economic melt down dries up the funding, but the funding levels driving the start-up bubble are so low, it's hard to see what will slow it down short of government regulations (they're working on that).
That's the most concise expression of this thing I've ever seen!
I'd normally say the start-up bubble would continue until the next economic melt down dries up the funding, but the funding levels driving the start-up bubble are so low, it's hard to see what will slow it down short of government regulations (they're working on that).
So, arc you thinking the Bay Area is in something like a perpetual bubble? Sort of like Helium starting to move and "roil" weirdly near absolute zero, once things have cooled enough to let quantum effects come into play?