I'll never forget this essay, because he first delivered it as a talk at Harvard that Steve and I trained up from Virginia to attend. After Steve had lent me Masters of Doom, I was already convinced we should start a startup together, but this talk sealed the deal. I know Chris Slowe (later our first 'hire' at reddit) was in the audience that evening and I'm sure a number of other future YCers were also there with no clue of where it would lead.
Thank you, Paul Graham, for making that comment about finding investment from rich people who'd done it before (but not him). If I recall, it was feeling guilty about hearing the dejection in the room that got him thinking about starting Y Combinator.
And what a difference that decision has made in so many of our lives.
So here's the back story: we're working on migrating the reddit blog off of blogger and have a test environment set up that we were playing around with. We have a semi-private IRC channel that your coworker, as a close friend of the reddit team and heavy duty open source developer also hangs around in.
I asked if anyone wanted login to the new blog interface to which I got a "Me! Me!" from the gentleman in question. Since he isn't strictly speaking a member of the dev team, I did the natural thing and said, "ok, but in exchange you're going to have to dance for my amusement." Suffice it to say he called my bluff.
>as a close friend of the reddit team and heavy duty open source developer also hangs around in.
I know the significance of what he contributes to Reddit, I've been on Reddit for a long time and ran into him because of IRC. He's the one that got me my job :P
I'm not surprised at all that he called your bluff, he's quite the playful fellow. I was amused when he shared it with the rest of our dev team (the video).
Any other reddit anecdotes I can share for you? I shared a few on my blog (http://reddit.blogspot.com) during the startup, but we lucked into Chris because he went back to his PHD after doing YC that first summer and had 2 spare bedrooms, which Steve & I promptly moved into when summer ended and PG talked us out of moving back to Virginia.
Chris generally just made us feel guilty for ever needing to sleep (since he was helping us part-time and PHD-ing at Harvard the rest of the time -- oh, and he met & dated his future wife then, too!)
So as an interesting anecdote, a Squidoo lens I have has for the last several years been the #2 Google result for the phrase 'start a startup,' after this essay. The page gets on average 45 visits per week, and has sold about 40 startup-related books total over the last 3.5 years.
I think it's interesting because
* The number of people looking to learn about starting businesses seems to be flat or even slightly declining, at least if search traffic is any indicator.
* Based on what I can extrapolate from my analytics and well as my experience playing around with Google / FB ads, it seems like the total number of people searching for information on starting new businesses is on the order of a a couple hundred per week. This is pretty disturbing for a nation of 300M people.
I'm not sure how much validity there is to this, but I've always been a little disturbed whenever I've done things like, say, comparing the number of parenting books sold on Amazon to the number of children born each year.
I think this just reflects how few people are part of the Valley culture. Outside of the Valley and people strongly affected by its culture, they're not called "startups" — they're called "businesses." The phrase I would guess most people would use to search for information on the topic is "start my own business." I don't think anyone I know would say "start a startup."
There is a big difference between a business and a startup. And at this point most people trying to start a startup would know the term. So I think the explanation is simply that "start a startup" is not the kind of phrase people would type into Google.
That said, whatever you are using to determine that "the total number of people searching for information on starting new businesses is... a couple hundred per week" is very very very far off the mark.
I'm sure that's true so some extent, but I'd bet the vast majority of 'home business' searches are people doing lead gen, e.g. 'home business' + 'palo alto'.
What is even more interesting is the countries and cities the searches are coming from. A lot of Asia as far as countries go but the top cities are in the US
Definitely a brilliant essay, but I might take issue with a possible interpretation of:
"It's very dangerous to let anyone fly under you. If you have the cheapest, easiest product, you'll own the low end. And if you don't, you're in the crosshairs of whoever does."
I definitely understand his point, but I am afraid others might not. This sort of mentality can lead you to under price your services greatly. I see it everyday. Our customers sell downloadables and just the other day some guy was trying to sell a brilliantly manicured and useful spreadsheet for $2. His competition charged $10-$15+ and looked 10x more confusing and horrible. He said he wanted to "sell the most possible and low price was the way to do it".
I think I may talked him into raising his price and testing a few different pricing schemes.
In short: "owning the low-end" isn't always the best way to go about it, you might be leaving a lot of money on the table and selecting a different set of customers.
When we started Virtualmin, we had one low-end competitor that charged slightly less (the software did quite a bit less) and three major high end competitors that were higher than ours (and have close enough to feature parity) prices by 25-50%. I figured we were right in the sweet spot for a new product: not the cheap option, since we can do a lot more than the cheap option, but a better buy than the big options.
Soon after we started, a new very low end competitor came onto the scene (about a third of our lowest initial price, and without recurring fees, while we have annual renewals). And, one of the big three high end competitors got acquired and discontinued. Same number of competitors (not counting a bunch of also ran products that aren't competitive for whatever reason), but now we're right in the middle on pricing and at the top on capability.
Since we're talking about people flying under you, I'll talk about the low end option. From what I can tell, that low end product sold about four times what we sold during the time the company existed, though revenue would have been about the same as ours during that time (because of recurring revenue). So, this company had four times the customers to support, with the same amount of resources. Of course, if you can get users helping users, this becomes less of an issue and more users is better. And, customers that are happy are a much better marketing value than buying ads or for trade show booths or whatever else.
Our situation is further complicated by the existence of Virtualmin GPL, which is the really low priced option in the market (free). Virtualmin GPL probably has about 30 times the market share of Virtualmin Professional, and is probably in third place in the market overall now (possibly fourth, but definitely gaining).
The company behind the low price option went out of business, and the low end product has since become free. Development has dropped off significantly, and its market share seems to be slowly shrinking.
In the end of all my contemplating on the subject, we've kept our pricing pretty much exactly the same: Middle of the pack, but closer to the high end than the low end. We figure the free product will mop up the really cheap folks, and when they start making money, we'll probably get some of it for a Virtualmin Professional license.
What I'm trying to say is: It's hard to know the right price. Flying under your competitors is great, if you have some significant technical advantage they don't have (like you have a web-based service vs. their installable boxed application, or you have a much simpler product that can answer the needs of a specific niche without the complexity of your competitors). But, if you're building a complicated product that has only marginal cost advantages over your competition, it's probably dangerous to significantly undercharge. Sometimes there are reasons for the prices you see.
Now, I don't want to sound like a hopeless groupie or somesuch, but I don't know how else to say this... I love this guy!
Too often my hope of ever launching my own startup one day falters after attending yet another pointless big company Dilbertian meeting or when I give in and watch a stupid action movie from Netflix that I had watched at least twice before, rather than learning Rails. After those moments, I read one of PG's essays and I feel it is possible, I can do it.
Maybe I'll never launch, but just for giving me that hope for a short while, I love him.
falters after attending yet another pointless big company Dilbertian meeting
Really? It's those mind-numbing, pointless time wasters that increase my drive to go out on my own! Bring on the meetings: it just increases motivation.
Maybe my startup should build a tool for better meetings :-)
You're right :-) But those meetings turn one's mind to mush, so it's hard (or at least I find it so) to concentrate on something at night.
If you can design an some software/hardware/toy that one can pass time discreetly during those meetings while pretending to listen, my friend, you'll have a huge market
I've been meaning to go through everything I've written and collect the mistakes. IIRC I was wrong about the low end of the age range here. It was also a mistake to advise people to work for an existing co for a couple years before starting their own.
Hopefully you can find the time to do this soon, it will be very helpful for all of us trying to become successful and change the world (or a piece of it) with our startups.
It would be interesting to know how many HN readers have first seen and/or read this essay just now, after seeing this thread. Or have we all read it years ago...
It's my first time reading it. I didn't get interested in starting a startup until a year ago, and only in the last week did I really start seriously doing it. I'm happy that old articles pop up on HN years after, since I've missed a bunch.
A classic, but it's interesting to read alternative ideas to:
>> During this time you'll do little but work, because when you're not working, your competitors will be.
such as from Jason Fried who recommends against working all the time.
Overall my experience has been: work as much as you can while not burning out and still enjoying your life overall. You never know how long you'll be at it, this could be your lifestyle indefinitely especially if you enjoy startups and may do another even after a big success.
For me so far this has meant incredibly long hours come in spurts, maybe a burst of 3 days every couple weeks, and otherwise, hours that approach "normal" (10-12 hour days, take at least one weekend day completely off).
This was written in 2005, but I can see at least 2-3 businesses that have cropped up since then that match the end of the "What customers want" section almost completely. He certainly knows how to find a niche.
Thank you, Paul Graham, for making that comment about finding investment from rich people who'd done it before (but not him). If I recall, it was feeling guilty about hearing the dejection in the room that got him thinking about starting Y Combinator.
And what a difference that decision has made in so many of our lives.