Hacker News new | comments | ask | show | jobs | submit login

Nice essay. IMHO the contents are (1) Common, good advice. (2) Good data from Sam's excellent experience. (3) Some not so good advice and attitudes.


The essay seems aimed mostly at current Silicon Valley (SV) style, mostly consumer, information technology startups. Okay, but that's not all of business or even all of startups. Yes, YC is pursuing much more, e.g., some shoe company in Pakistan, still, the essay is SV style and there, mostly Web and mobile. Fine with me, because that's what I'm doing, be we should understand this point about scope. And, as below, we should understand this scope and style because IMHO it's time for SV style of consumer Internet to borrow from some of the rest of technology and business.

Broad Point:

As we all can see, all heard from Mark Andreessen, etc., and see from Sam,

" ... investors’ returns are dominated by the big successes, ... "

The Exceptional:

So, from this Broad Point, the goal is something exceptional. From that, we have to suspect that we won't always be following the common and ordinary, some extensive experience and observations from the past, or even "big successes" from the past, and, instead, should be willing to consider some exceptions in order to be exceptional.

Users' Love:

> "Your goal as a startup is to make something users love."


Now, can we, please, have some more guidance on just how the heck to do that? And, please, don't ask me to draw from Snapchat or Homejoy. And I'm concerned about



"However, these statistics also reveal a grimmer reality: 93 percent of the 511 companies accepted by Y-Combinator have failed. Even more alarming, only 3 to 5 percent of the companies that apply to Y-Combinator are even accepted, meaning that only one in every 200 companies that applies to Y.C. eventually succeeds."

And, I'm concerned about the low ROI of venture capital as in




Instead, for history with some good examples, there have been many amazing projects, some astoundingly innovative, that worked the first time with high probability, e.g., for a picture:


IMHO, here is a better approach:

(1) Find a problem that huge number of potential users/customers believe or can come to believe is really important to have solved. E.g., want, say, 1+ billion people with Internet access, if only from a smartphone.

(2) Find a solution that is much better than anything else and difficult to duplicate or equal.

(3) Make sure that for the target customers right away or soon the solution will be seen as a must have and not just a nice to have. Want no doubts here; do not want to have to depend on gossip and fads from notoriously flighty teenage girls. One of the best examples would be a single pill, safe, effective, cheap, to cure any cancer.

(4) Deliver the product, easy to use at a very attractive price.

(5) Make sure the revenue covers all expenses and yields a fantastic margin, e.g., pre-tax margin 90%.

(6) For that solution, do some good original research with powerful, valuable results in the STEM fields.

(7) Offer the solution as a Web site, i.e., exploit software, Moore's law, and the Internet.

(8) Be a solo founder until at least $10 million a year in after tax earnings.


> "A word of warning about choosing to start a startup: It sucks! One of the most consistent pieces of feedback we get from YC founders is it’s harder than they could have ever imagined, because they didn’t have a framework for the sort of work and intensity a startup entails."

There's something wrong here: All across the US, east to west, north to south, cross roads to the largest cities, millions of sole proprietors do startups and are successful enough to buy houses, support families, and get the children through college.

All the larger bodies of water in and around the US have boats and yachts, and nearly all the owners are just such entrepreneurs. Maybe they own 10 fast food restaurants, are big in asphalt paving, are a manufacturer's representative for some great lines, run five new car dealerships, own and manage 2000 units of rental property, have a private label line of industrial floor cleaning supplies in a mid-size Midwestern state, did a rollup of dry cleaning stores, are the main beer distributor for half of a state, are a leader in design and construction of custom tanks on truck frames for hauling liquids, etc.

But, a startup that exploits information technology, software, Moore's law, and the Internet should have some advantages and generally be less difficult.


> "Remember that at least a thousand people have every great idea."

Maybe true with SV style, but more generally, no, and a thousand times no.

Instead, since so many startups fail, we want some advantages and definitely can get a lot of advantage from having a genuinely new idea. People who wrote a Ph.D. dissertation that was supposed to be "an original contribution to knowledge worthy of publication" and "new, correct, and significant" will quickly appreciate the importance of a unique idea and a lot about how to construct such. Here SV style is seriously lacking and, as above, needs to borrow from outside.

As in Sam's

> "Remember that at least a thousand people have every great idea."

I believe that SV style nearly trivializes the idea and, to raise the success rate, very much needs to go much deeper into the idea and associated considerations of user need, market size, meeting the user need, and new, proprietary technology to meet the user need especially well with a product difficult to duplicate or equal, and protected intellectual property that supplies a barrier to entry. In some places such unique intellectual property is taken very seriously, with laws, contracts, national security classification, etc. For higher success rates, SV style needs to do better with such intellectual property.

Several good examples of a such intellectual property and its power are in the picture


This is another case of where SV style needs to borrow from outside.


Once again, over again, one more time, yet again, we come to the issue of team. Again we learn that it's tough to get a good co-founder; co-founder disputes are a major cause of startup failure; it's tough to hire good staff; it's difficult to keep the staff well involved; being a good leader and manager and learning to do so is a lot of work, and BoD members rarely know much about the details of the business, likely much less if some new, unique, powerful, valuable crucial, core technology is key to the business.

So, with all those clear dangers to the startup, we begin to conclude: Be a solo founder, get to earnings ASAP, grow organically, well into very good profitability hire no one, and from the start carefully plan never but never to accept equity funding or report to a BoD. Or, follow the example of Markus Frind and his romantic match making Web site Plenty of Fish, initially, just one guy, two old Dell servers, ads just from Google, $10 million a year in revenue, and recently sold for $575 million in cash.

Understand the Users:

> "... it’s critical you understand your users ..."

Right. And for making, say, really nice seat cushions for the driver's seats of Rolls Dropheads with owners in the Chablis and Brie set in the Hamptons, sure.

But in consumer Internet, for a big success, there will millions, maybe billions of users, and about all that can be said about those users is that they are a not very special cross section of humanity. So, really just have to understand the pair of the product and the ordinary man on the street.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact