Lessons of Y Combinator: Things I’d do differently after 2 startups | Hacker News Reader

Lessons of Y Combinator: Things I’d do differently after 2 startups

foundread.com

81 pointsmarcus32 comments

Open articleView on HN

"they aren’t that much better than many of the great coders I’ve worked with. They’re just bolder."

Interesting way to put it. Sounds like an critical ingredient for success.

Yes, it's a critical ingredient for life in general.

Success is usually achieved by combining ability and boldness

Agreed!

I thought it was determination and luck.

It's also a critical ingredient for failure.

I'm a huge fan of "necessary but not sufficient" qualities like this.

A startup needs:

1) Boldness 2) A crazy idea that most people think is stupid. 3) Persistence

All of these things seem to be absolutely required for success. But none of them is sufficient. Combine 'em with good "product/market fit" (a la Andreessen) and I think you win.

Startups don't need "A crazy idea that most people think is stupid". In most cases, an idea that most people think is stupid is exactly that.

Startups need to offer something of perceived value, that's pretty much it. In some cases, it's doing a typical task in a non-typical way, or doing something non-typical altogether (your crazy idea), or marginally improving your typical task in a very typical way.

In the end, it's all about value.

Startups - like any market - are a parimutuel system. Your payout is determined not just by whether you're right or not, but by how many other people are also right. If you bet on social networks in 2004, you'd be right, but unless you're Mark Zuckerburg or Tom Anderson, you wouldn't have gotten rich off it.

So really, you need "a crazy idea that most people think is stupid, but is less crazy for customers than for competitors." Either that, or there has to be some barrier to entry that lets you execute but prevents everyone else from executing, even though they can see the same opportunity that you can.

Agreed, with one caveat: I don't define a successful startup as one that makes you wildly rich. (i.e. given Zuckerburg's "paper" value, I don't consider facebook as being successful... yet)

I disagree. If everyone knows that an area is an amazing opportunity, then it's probably not an amazing opportunity. Every single runaway success at some point had a whole host of people saying, "That's ridiculous-- it'll never work" on some level.

examples:

1) When google was on its way up, it was a well-understood fact that no one could make real money on search. Oops.

2) When Excite was getting funding, everyone told them that search was dumb. People would do 1 search and bookmark what they cared about (and would never search again).

3) The very idea of youtube (giving away tremendous piles of storage and bandwidth) didn't seem very smart to a lot of people. Plenty of online video startups had failed.

4) Blogger couldn't get more funding and had to lay off their entire team.

5) If I described twitter to you 3 years ago, would you have been blown away by the idea?

6) Salesforce.com was not exactly warmly embraced by the enterprise when they first launched.

7) Marc Andreessen said he was constantly stonewalled by businesses when selling them stuff at Netscape. "The Internet? Inside our network?! MADNESS!"

Hindsight make these ideas seem obvious-- but at the time, they were pretty ridiculous.

As you say, most ridiculous ideas ARE really ridiculous. Thus the "necessary but not sufficient" part.

My point wasn't that seemingly ridiculous ideas never succeed, it was that they are not a requirement for startup success.

3. We focused entirely on product/market fit.

I'm curious. In what way did you focus on product market fit? What signs did you look for to point you towards it? Seems to be a very important aspect of success (besides boldness).

This might be a too-fancy way of saying "we focused on what people wanted" (to put it in PG-speak rather than Andreessen-speak).

We started with a permission marketing campaign (before we wrote a line of code)-- showing a few screenshots and allowing people to sign up to hear about the launch. That helped us understand how much people wanted it (measured by traffic and conversion of traffic to sign ups). We were immedietely barraged by lots of people who had ideas about the product, and we listened very carefully.

We ignored lots of things that it might've been bad to ignore in favor of building features that people wanted and talking with users. We didn't incorporate. Didn't get a bank account. Didn't talk to a lawyer. Didn't worry about SEO or any other marketing (even though I'm a bit of an SEO nut). Didn't talk to investors.

Lots and lots and lots of focus on reducing frustration, reducing friction, and increasing value-- and lots and lots of communication with users (to understand that stuff better).

Boiled down-- lots of focus on understanding what the market wants (both individual lifehackers and biz teams that want to be more productive) and then building/iterating. Rinse, repeat.

Well done, and thanks for sharing that here.

Thanks for the reply. Definitely something to keep in mind.

If you are giving a product or service away I don’t think you can assert that you have “product/market fit.” You have a very compelling application, but until you figure out who is going to pay you for it, and how much it’s worth to them, I don’t think you can put a check in the box next to “product/market fit” since you don’t pass this test from the same Marc Andreessen post you reference ( http://blog.pmarca.com/2007/06/the-pmarca-gu-2.html ):

And you can always feel product/market fit when it’s happening. The customers are buying the product just as fast as you can make it — or usage is growing just as fast as you can add more servers. Money from customers is piling up in your company checking account.

Well, revenue is a beautiful thing, certainly. You can define product/market fit however you like-- but by that definition, most consumer apps don't have it. A customer can "buy" a product with money or their time/attention. I'd say both YouTube and Facebook have found a great product/market fit. And Gmail. Andreessen found a great product/market fit with the Netscape browser, but monetization of it really only came in the form of a few bizdev deals (which didn't cover costs, I don't think). Money, in many of these iconic companies, was NOT piling up.

Regarding my company, we focused initially on a non-paying market (individual lifehackers), as that was easiest to build for and launch a rough beta for. They "pay" in blog posts, tweets, word of mouth-- they've driven every single business lead we've gotten. Now we've got a new market (teams) that we have to find a "fit" for. Wish us luck!

It's unfortunate, but a polished presentation making a strong case for a large market share is more likely to get you angel funding than a working product with relatively few users. I say this after seeing a bunch of angel forums / VC workshops in the bay area. Many of the "ideal" presentations had no product, no engineers but a "strong management team," a large market, and a path to bigness. You are unlikely to get funded without a grand slam story.

This just convinced me to apply, less than 12 hours to go. What is it going to be. Great read especially if you still hesitate.

If you hesitate a lot, try climbing or ski mountaineering ;-)

Either one is easier than having a kid, though. Don't do that until after you've got a few businesses under your belt, if you can help it. Execution is a habit, IMHO. Hard to acquire but much easier the 2nd, 3rd, 4th times. Make it your habit and you will be better off regardless.

Oh, if I'd known back then... fuck, if I looked back on the opportunities missed (instead of the ones we capitalized upon) I'd slit my throat. But that gets back to the three important qualities for a (founder|mountaineer|parent):

1) good pain tolerance, 2) shitty memory,

...and I can't remember the other one. ;-)

3) Balls?

From the article: I'd rather own 94% of a watermelon than 100% of a grape.

That's a very good metaphor and a very good explanation of why you'd want funding.

would love to see a bit more depth here, most points seemed rather surface - "we focused entirely on product/mark fit" - how?, what are some methods you found effective? maybe best left for a follow up post but after reading it I wanted more on their successful/unsuccessful tactics.

RscueTime seems like one of the more credible YC startups, with an actual value proposition and a pretty killer initial user experience.

After all the YC help, weekly dinners, and demo day, this article says they're still looking for seed funding. Not an A-round; just seed money.

Do you have the screenshot's you used for the first version to get feedback?

surprising at 7%/week growth with Tony being featured on so many blogs

Surprisingly low or surprisingly high? If you think that's low, consider that 7% weekly growth is roughly 3300% annual growth, something that most companies would kill for. If that can be sustained for 3 years, the company will be roughly 25,000 times larger than when it started.

Yeah, we're pretty damn tickled with the growth (and stickiness-- abandonment seems unusually low for web software, from what I've read).

FWIW, the 7% week/week was well before any of the press coverage around YC demo day (and associated splash). We've been 12-14% for the last two weeks. I expect it to settle back to the standard word of mouth level of 7%. Of course, I think we have some clever SEO/viral things in the pipeline, which might pick things up a bit.

congrats -- wasn't dissing the 7% was just surprised. 12-14% ain't bad :)

With respect, it's a compelling app but it's easy to give something away. The real question is how to wrap a revenue model around your offering. If you base advertising on my on-line behavior it may be viewed as too invasive. Another key attraction is that it doesn't interrupt me when I am in flow, so you are probably limited to presenting ads during the analytics so that I am not interrupted when I am actually working. You've done a good job of addressing privacy concerns so far, but a local app would be even more secure: why not offer a version that competes with http://www.timesprite.com/ the attraction for me is that their revenue model is clear and I can rely on them being around for a while. And I would be glad to pay $40/yr for rescuetime as is if I knew that would allow you to persist in your current privacy model. Your situation is unstable until you get to at least cash flow positive. It's a great app, but I would worry a lot more about adding some kind of paying customers instead of more free users.

Heh-- if it's easy to give something away, why is it that so many free web apps/sites don't have growth? Or happy users?

It's damn hard to give something away online, as it turns out.

But I do agree-- while free accounts will always exist at RescueTime, our future is NOT as an ad-supported biz. :-)

Reply on news.ycombinator.com