[booruguru]

If the .com name you want is already registered, you can always buy it later once you business is successful enough to justify the investment (much like betali.st and getdropbox.com).

There's no point in buying an expensive domain name for a company that may not survive past the startup stage.

Note: a large swath of web users don't even understand the concept of a domain name or address bar...instead they google terms like "Gmail" and "YouTube" thinking this is how they are supposed to visit web sites.

What's more, tech-savvy users wouldn't think twice about visiting domain name ending in ".io", ".ly" and so on. (Hence their proliferation.)

*

I can't help but wonder if this article is some kind of prank--it's far too ill-considered compared to Graham's usual standards.

[jccc]

You're thinking of customers looking at your company. Could it be that pg is actually talking about investors looking at your startup?

That seems to be more of what he's getting at here:

The problem with not having the .com of your name is that it signals weakness. Unless you're so big that your reputation precedes you, a marginal domain suggests you're a marginal company.

[brudgers]

My reading was that at an early stage it's just better to come up with another name and be done with it...if airmatressbreadandbreakfast.com is taken, just buy airBnB.com. Consumers won't care. In cases like Stripe, where the target is B2B and the businesses are mostly small businesses, then a good name matters because it's marketing and its cost is just a cost of doing business.

To push it further a .io domain is signalling and that's what makes it attractive.

[GuiA]

That's a bad example though. If airbnb.com is taken, should you take airbnb.io or airmatressandbreakfast.com? Paul's argument is: neither, get a different short name for which you can get the .com.

[brudgers]

I take it that the article is aimed at early stage startups and is premised on the idea that choosing a name is worth about 20 minutes of time by a founding team and an experienced investor. I find Airbnb useful because "air" has nothing to do with the current company. Any connotations to air mattresses are probably non-positive at this point and the company's success came after pivoting away from low quality low cost accommodations.

[nostromo]

Airbnb.com actually did come later.

It was originally airbedandbreakfast.com.

[pbreit]

"There's no point in buying an expensive domain name for a company that may not survive past the startup stage."

This is exactly why the .com domain wins. They do not have this attitude. Their attitude is to be in it to win it and do what it necessary to be successful. That's the .com signal.

[nostromo]

It also depends on your market.

If ProductHunt disappeared tomorrow, I'd be sad, but that's it.

Stripe and Parse on the other hand both needed to signal strength to their customers from day one. There are many customers that would not entrust their payment details with getstripe.ly.

[corobo]

I've never really thought about it until now but I don't think I'd even pay for a service on .ly et al even if it was using stripe/PayPal

Maybe I'm just getting old but the .com wins it for me, at an absolute push for online services I'll trust a .net

[pbreit]

.ly is particularly bad because Lybia's maintenance is a real risk.

[CmonDev]

What about https:\\ though?

[corobo]

It's trivial to get a cert for HTTPS. That doesn't tell me if the people behind the site are trustworthy, security conscious or going to be around in a year's time

I know that the dotcom doesn't do that either, it's not a rational thing - just a thing

[booruguru]

Not everybody finances a startup with a mountain of cash. If you're funding your business with your life savings and credit cards, it's difficult to place a premium on a domain name while contemplating how long you can pay your mortgage and put food on the table.

Also, blind optimism is not required for success. And no domain name is an antidote for a shitty idea that nobody cares about.

[marincounty]

You are right about the chitty idea, but decent .com names are dropped every day. (You need to scout for names every day though, but will become a nasty habit in no time.) I have found that people, outside of tech, will remember weird, odd .com domains. I got people to remember physibles(and how to spell it.), and these were senior citizens. (I guarantee these people have never been on the mother site either!)

I tried a different domain with a .io domain; and it just wasen't worth it the effort.(they knew about .net. ,org, but that's it.). There's almost a anger that comes up when a domain ext. gets to specific, like .photography?

[pbreit]

You basically make the case for .com. It signals funding and optimism (whether you wrongly think it's bind or not).

[taytus]

Maybe that was 2 years ago. Most people google the product/service nowadays, so if you have a better product/service, your company might be on top of some .com domain doing the same thing. I think that the .com fever is over.

[pbreit]

It has nothing to do with how you find the website.

[ThomPete]

Most people don't know what .com is. So sure if you are talking about a bunch of VCs your might be right but for most people it doesn't have that value.

[pbreit]

How do you explain that pretty much every very large successful company has its straightforward .com? And why do you not place any significance in that?

[ThomPete]

It's .com for historical reasons. Today most startups wont be able to afford a proper .com name, but as they grow more and more successful they will.

So PG is right that a .com is important but not for the reasons he seem to indicate.

That you can acquire a proper .com is a manifestation of your success it is not the base of your success. And that is how his essay comes of at least to me and many others.

[paulsutter]

Exactly - his statistics are missing the important number, how many of those late stage YC startups had purchased the .com after they had raised substantial funding? (ie, could it be the pattern rather than the antipattern)

[snowmaker]

It's a good point, but I think empirically not many. Dropbox is the only one I can think of; they were originally getdropbox.com.

[bhartzer]

I read this and honestly thought that it's kind of bizarre that he seemingly "randomly" came out with this post. My first impression was that it was a publicity stunt by Verisign.

Nonetheless, Paul's point about naming is a good one. You have to have a good name. But, I see no evidence (and he didn't provide any) why the name has to be associated with a .COM domain. His only 'evidence' that he mentions is the fact that "100% of the top 20 YC companies by valuation have the .com of their name. 94% of the top 50 do. But only 66% of companies in the current batch have the .com of their name." This doesn't mean that 94% of the top 50 are right (in owning the .com).

[ghufran_syed]

Yes, you would think it would be easy to analyze two yc portfolios, one with .com, the other with every other domain, controlling for the time since formation, and compare results, both survival and valuation. That would probably be a lot more meaningful, since the most highly valued companies are almost certainly also the oldest (surviving) companies.

However, I do think the point he is making is correct, that there are a number of constraints, but that the optimal decision involves a different weighting of them to that of many founders (my interpretation). I would argue that the constraints for startup founders naming companies are: 1) short and easy to say & remember 2) available (cheaply) 3) domain (.com vs others) 4) intrinsically meaningful 5) related to the business

Some founders compromise on 3 and give more weight to 4 and 5: what I think he's saying is that 1,2,3 are much more important than 4 and 5, to the point that "almost any word or word pair that is not an obviously bad name is a sufficiently good one".

[guywithquestion]

> There's no point in buying an expensive domain name for a company that may not survive past the startup stage.

The point (and title) of the essay is "change your name," not "make sure you get the .com."

[RowanH]

So I started with a non .com address, after watching a client trying to find my homepage with .com my heart sank. The first thing anyone does is write in abc.com or their local country tld (in my case .co.nz).

3 weeks ago I purchased the .com after the original owner let it expire, cost me more than I wanted to pay bet less than I would have paid. I was ecstatic, feels so much better to say abc.com ... !

[rl3]

This is what Product Hunt did.

Originally it was producthunt.co, and once there was sufficient traction, producthunt.com was acquired for a few thousand dollars (if memory serves).

I don't think anyone would argue that Product Hunt should have chosen a different name because the corresponding .com was taken.

[stephengoodwin]

Facebook originally operated under "thefacebook.com" and bought "facebook.com" for $200,000 in August 2005.[1]

Looking at Crunchbase, they had raised around $13M at that point, so I'm sure spending $200,000 seemed like a lot at the time.[2]

[1] http://mashable.com/2006/08/25/facebook-profile/

[2] https://www.crunchbase.com/organization/facebook/funding-rou...

[CmonDev]

Well, that's 13M that you don't need to give back personally (unlike some bank loan). And .com is pretty important for a site like that. Fb.com is even better though.

[jscheel]

.co is pushed hard in the startup world. I can't stand it though. It's way too close to .com, and results in a lot of messed up traffic.

[smoyer]

I wondered about the tone too ... perhaps we've got a fake-paul-graham in our midst? The thought experiment referenced by [1] is pretty weak. I don't think it's the attachment to a name but more commonly the lack of a better name - and it's especially hard to find a name you like that still has a .com domain available. Having a .com name like stripe indicates a position of strength because (at this point in time), someone paid extra for the privilege.

[dweekly]

The good news is that if paulg was testing to see if HN was just an echo-chamber that applauds everything he says even if he gives terrible advice, HN just passed the test with flying colors. :-) Hooray for critical thinking!

[ricree]

thefacebook.com also comes to mind.