354 comments

[ 5.7 ms ] story [ 263 ms ] thread
100% of the top 20 by valuation have their .com

What proportion had it at their founding or shortly after? Surely a well funded and successful company had the means and motivation to purchase their .com

Cheaper still, would be to get a new name that has an available .com
A great example: how long did it take Square to get their hands on square.com? Even now it's just a redirect to squareup.com.
Another example: Dropbox originally had getdropbox.com, and didn't obtain dropbox.com until more than 2 years after founding.

https://en.wikipedia.org/wiki/Dropbox_(service)#2007-2011

(comment deleted)
(comment deleted)
Actually, paypal owns x.com, not Stripe. :)
True, but x is a variable in this case.
Should've use ^([^\.]+)\.com$ for clarity.
(comment deleted)
If your numbers are going up and to the right, what does it mean to "signal weakness" with your name?
I think the essay is missing crucial elaboration on the central point. That is, the idea that .com is associated with strength, and lack of .com with weakness.
.com is supreme

.com is love

.com is life

Surrender your soul unto VeriSign

Exactly! This why I commended Casey Neistat for waiting to announce his startup name to the public until he secured the domain name.
>Don't believe a domain is for sale unless the owner has already told you an asking price.

One of my favorite examples: http://nissan.com

or tesla.com
I've always wondered how long will domain squatters wait before they negotiate. Surely Tesla Motors has approached tesla.com and apparently been turned down. Telsa Motors has shown they don't need that domain name to succeed. So when will the squatter, if ever, cave to Tesla Motors' offer?
Maybe never. The Homo Economicus model of human behavior as optimizing for expected economic return is only a first approximation. In some cases, human behavior is more accurately modeled as "Mine! My precious!"
Even if Tesla is not willing to pay the amount the squatter wants, he may be hoping that Tesla's competitor will.

By the way, the coolest thing Tesla could do, that would buy them a lot of love and good press, would be to buy tesla.com and donate it to Nicola Tesla Science Center.

Shouldn't that be tesla.edu rather than .com? Their current .org would also be more appropriate than a .com...
It can't be, unless time travel...

From Wikipedia: "Since 2001, new registrants to the [.edu top level] domain have been required to be United States-affiliated institutions of higher education, though before then non-U.S.-affiliated—and even non-educational institutions—registered, with some retaining their registrations to the present."

Yeah. I imagine if the offer and counteroffer are rebuffed, the squatter enters "well, now you're never gonna get it" mode.
Nissan.com is a great example. I am sure Nissan offered them a lot of cash for the domain name before going hostile on them.
You should probably read up on the case then.
must be killing it with that huge insurance ad
Similar thing happened here in Minneapolis back in the early 2000's with our local Amazon Bookstore (great bookstore, btw) and amazon.com
> "Whereas (as Stripe shows) having x.com signals strength even if it has no relation to what you do."

I would contest that it has no relation, I could be wrong since I have no knowledge. That said I always saw it as a relation to the cards themselves and the magnetic-strip on the back of the card.

Yep, I agree, that's what I thought it referred to the entire time (and even commend them on a smart name because of it)
> How do you find them? One answer is the default way to solve problems you're bad at: find someone else who can think of names.

Imagine a website where startups could describe what they do and users could propose and vote on ideas for names. It's been on my list of potentially decent project ideas.

and regularly scraped by bots to squat domains? You can hold a lot of domains for two years when you can sell a few for 5k
Very cool! This makes sense to me. Specialization and trade.
Oh, wow, that's a throwback to the past! I remember when Dmitry Davydov first announced it, kinda surprised that it's still around :-)
Great names are made not found.

I think Paul for once have it the wrong way around.

If you become big enough and happen to have enough of a market you might need to .com name to optimize your numbers.

If you happen to find a good name you can afford that you still think represents you and is .com you should most probably change your name.

But for most people just like their logo it's it's not that important until you become big and you can become big without .com name. Whether you can stay big is another discussion.

Far too often we fool ourselves thinking the wrong things are important.

But just as you most probably can easily find another name because your current one isn't as great as you might think, you can probably also wait a little until it becomes an actual issue.

and even when you stay big, your brand name is big enough that your domain doesn't matter (e.g., teslamotors.com)
Correct.I am trying to think of one instance where someone would refrain from working with Tesla, buying from them, or doing any sort of business with them because: Teslamotors instead of Tesla.
Little details are always important as long as they don't require big effort.
Agree. Problem is changing your name, to most people, is a big effort :)
There are different strategies and it certainly can be an advantage if you have a single word .com domain that encompasses a core value or offering.
A lot of things can be advantageous. But lets keep in mind that 99.99% of all companies are not going to be unicorns with budgets to buy +100K domain names anytime soon.

That doesn't mean they can't become a hugely successful company.

My general point is just that you shouldn't prematurely optimize because you end up spending a lot of time on it. Unless you stumble on a great name that you can afford it's most of the times simply not worth the effort.

How many jokes have been told about 1990s-era dot-coms spending $200k on domain names and then going under 6 months later.
Definitely, prioritizing is important - $100k on development will bring your startup much further along and having a great domain may be an artificial crutch.
For entrepreneurs, the domains name should be the least concern.
Couldn't agree more.

Google, Yahoo, Twitter.

Three global brands that aren't dictionary names, and don't have to be. You don't need a dictionary word .com to survive either.

All companies that have their .com, and have had since the beginning as far as I know.
'Yahoo' and 'twitter' actually are words that existed in the dictionary before the respective companies.
All combinations of 6 letters .com are taken. All 16k words of the dictionary are taken in English. It wasn't even the case 2 years ago. I wonder whether pg's impression that if is possible to get a .com is outdated.
I agree - he even signals to that by noting that only 2/3 of the current batch owns their dot com. I predict this number will go down for the next batch, and the next one, and ...
> All combinations of 6 letters .com are taken.

Do you mean this literally for any 6 random characters? It rings true for 4, but I believe you can easily find a 5 character domain name even now. In fact, I just registered three five character and a six character domains without much trouble.

we registered a new 6 letter .com 8 months ago for 20 dollars.
I found an available 6 letter domain name ( http://TapRun.com ) about a year and a half ago. I've found others since then...
this is nonsense. I registered normally a good speakable 5 letter .com domain just 3 weeks ago.
BoxGal.com is currently available. I'm sure others are as well.
This sounds implausible to me. I have acquired a couple five letter domains in the past year or three. They kind of suck and are hard to pronounce but they exist.
Sorry but that's complete nonsense. That's over 2.5 billion domains. You probably mean all 6 letter English dictionary words -- which might be true.
Twitter and Yahoo are dictionary words, and Google is just a misspelling (intentional or accidental, it's unclear) of a dictionary word.

All three did own their .com domain names from the very beginning.

And Twitter started as twttr, if I recall.
And Facebook as thefacebook.com
> Great names are made not found.

> But for most people just like their logo it's it's not that important until you become big

Exactly this. If you look at the most successful companies in all sectors out there, most of them have logos that range from plain to just ugly, and many of them feel like they were initially created as a formatted company name on top of some document, back when the company's secretary first discovered Microsoft Word.

The more I see things and think about them, the more I believe that this whole talk of logos being "very important since it's what your customer looks at all the time" is a big scam perpetuated by graphics designers, because they make (a lot of) money on you believing you need the slick, $expensive new brand image.

> The more I see things and think about them, the more I believe that this whole talk of logos being "very important since it's what your customer looks at all the time" is a big scam perpetuated by graphics designers, because they make (a lot of) money on you believing you need the slick, $expensive new brand image.

Patio11 has actually A/B tested two different logos. His conclusion was that "Your logo could potentially add or subtract 10% from enterprise value"[0].

[0] http://www.kalzumeus.com/2010/08/04/does-your-product-logo-a...

I wont even get into the amount of issues with testing a logo via A/B testing. Patios A/B test is anecdotal not by any metrics statistical.
> "Your logo could potentially add or subtract 10% from enterprise value"[0].

...for a very small enterprise with 0 previous brand recognition, choosing logos at random.

BCC site's main goal is to not look like a scam, so people understand it's a legitimate product, not a "steal $10 and disappear" site. That's important, but not as important if your product is already more substantial and credible than "pictures you could download free from the internet, for a small price."

Check out this short video by a well known graphic designer (Michael Beirut). He explicitly talks about how logos are inherently NOT important and are overemphasized in the world. Rather it is the consistent actions of a company that give a logo meaning and builds a "brand": https://www.youtube.com/watch?v=I0jw-Q7r-ng
Exactly. It's the same reason there is no such thing as branding per se. Everything you do is branding. From your packaging to your customer service to your employees.
That's exactly how I see it. Your logo has an important function - of an anchor. It's a symbol that can bring up all the feelings and memories people have associated with your operation. But it isn't magic, it has no power if people's feelings aren't there.
I don't think the logo is important but that's a different topic. If you at most successful companies in all sectors out there, they have their .com.
Yes but thats exactly the wrong point to make. The question is whether the name made them big or whether they got a good namer after they had success.
Great names are made not found.

If you didn't know Tide was a brand of laundry detergent, what would it mean to you?

It’s also strange to me since more then once when discussing startup naming I’ve quoted this earlier article Paul wrote where he makes nearly the opposite argument: http://aux.messymatters.com/pgnames.html
I think it's still relevant. That was before most of the TLD expansion so .com was just about the only option. I the gist of the article is that you have a lot of options in naming and many/most of them will be fine (as long as you get the .com).
I guess to me the main point of the new article seemed to be that the prestige and “signaling of strength” associated with the .com is what makes it important, while the older article argues having a traditionally prestigious name can actually be a negative, in signaling that maybe the founders “have more money then brains”. (I understand he's referring here to names that are obviously bought from squatters, but still the main point seems to be don't worry if a name fits into the traditional mold or seems 'prestigious' as long as its unique, memorable and ideally communicates something about your business. I'd also never expect a newer startup to be able to find "stripe.com" or "parse.com" without paying significantly for them today.)

I’d also say the proliferation of new TLD domains has made them more, not less acceptable for new companies, especially ones like .io and .ly that have come to be pretty closely associated with the startups in general. Like you say, in 2006 nearly everything still was .com so a non .com might have stood out more as a negative, and even then pg says he still had no problems with del.icio.us

Also the last section stood out for me as a contrast with the title of this one:

"Whatever name you choose, be careful. Names stick. You need a way to refer to things, and whatever you call something rapidly becomes its name."

Since pretty much every huge, successful company has their .com, I suppose the question is when they acquired it and the compensation.
(comment deleted)
Best advice I've ever seen on naming a business is from this old Seth Godin gem:

http://sethgodin.typepad.com/seths_blog/2003/06/naming_a_bus...

I read that in 2003, and have never forgotten the "Lemon Pie" example since.

I've done some scuba and I don't get the Lemon Pie name in that context. Care to explain?
It's the whole point of the linked article.

> The LESS it has to do with your category, the better.

Apparently I'm still really, really bad at this.
:). Just think to yourself what apples have to do with computers and phones.
The Apple name did have a somewhat contrived connection with computers when the company was first started.

Back in 1976, Steve Jobs explained the name to me: "Take a byte of the Apple, get it?" And that's where the famous shape of the logo came from.

(I almost ended up working for him at the time - but that is a whole other story!)

To me I don't think it has anything to do with Scuba per se.

More like it's a simple and easy name that you can build your scuba tour business around.

In other words - pick a good and easy English name or two and go from there. Avoid hard to pronounce words or smash two words together to make up a new word that means nothing.

At least that's how I see it.

easy as lemon pie
I was developing websites in 2003, so I feel like I should be able to relate here, but he has this comment at the bottom that I can't figure out...

"Use a stock photo CD and find cool pictures that match your name BEFORE you pick the name. If you can find a bunch of $30 images that work with a name, grab the pictures, then the name."

Huh?

I think he's saying that once you have a set of potential names, get some stock photos/pictures that match up to those names, and then pick a name that you like AND has cool stock photography. That will keep you from picking a name and then having to spend even more time picking photos that go along with your name.
The name "lemon pie" makes a clear image in your head. This makes it easier to remember. This is why "peg" mnemonic systems work.

It's how I came up with the name "Big Blue Saw".

"Nearly all your attachment to it comes from it being attached to you".Interesting.Very interesting as this is truth in the real sense.
i thought pg was finally changing his name to butt
I knew I needed to change our company's name but this article really sealed the deal. The problem is I don't have the skill of naming. If you do, I'd love your help (email me at lalalee at hotmail.com).
> 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. Which suggests there are lessons ahead for most of the rest, one way or another.

I wonder how much of that is due to the fact that landscape has changed in past 8 years and top 20 YC companies are all at least a few years (5-ish?) old. I appreciate the rest of the sentiment of the post, but I don't think the numbers he chose to share clearly support the assertions. Correlation isn't causation.

(comment deleted)
yes, didn't dropbox used to be trydropbox.com? And, tho non-YC, delicious.com was del.icio.us (or something). Successful companies can buy their domain.

But I agree with pg; a startup by any other name would succeed as sweet, so why not choose one you can get.

del.icio.us is an awesome domain. I'd be inclined to believe someone came up with the domain before the name.
I couldn't always remember where the dots went. Not a great property for a domain name.
Which is why I had it on the bookmarks bar. A single bookmark to all of my bookmarks :-). I haven't used them in a long time, what are they up to these days?
I'm thinking that joshu did, IIRC
Dropbox was getdropbox.com and airbnb was airbedandbreakfast.com
I made the same mental mistake as you just did and which I now bet PG thought about and carefully dismissed. Correlation isn't causation, but the fact is the top 20 companies either started with or changed their name. Which is his thesis.
What also bugs me is that there are probably things that:

(1) successful companies do often,

(2) marginal companies do less often, and

(3) aren't actually in any way instrumental to success.

It could just be that there's a mindset among successful founders that produces a particular vector of decisions, not all of which are all that important.

Graham would probably respond: whether every component of that vector is important or not, investors are scraping for every bit of evidence they can find, and your name might be a "tell" that other components of your decision vector are hinky.

And that might be true enough, but cargo culting every decision Airbnb makes in order to slip through an investor filter is no way to live either.

You forgot (4) require significant effort. Yes, those things bug me as well. As he points out though, naming is not particularly difficult, expensive, or time consuming when done right.
>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. Which suggests there are lessons ahead for most of the rest, one way or another.

How much of that is due to the pressure that he admittedly puts on them to change names? (And that higher valued ones are likely to be older and have had this pressure for longer?)

In addition to having had the pressure for longer, they have the money from having raised multiple rounds of financing.
Well, yeah, that's what "higher valued ones are likely to be older" implies.
correlation is not causation...
We'll buy it when we raise series A. We'll use our hipster version until then. Thanks.
Is this signaling as simple as having a .com ?

Or are there other considerations such as SEO uniqueness, easy-to-pronounce (especially critical for enterprise software), length of url.

Is kuickkoder.com a better domain than quickcoder.io ?

It's a sliding scale.

If the .com is kuicksuperkodderdudes.com, then quickcoder.io is superior. Yes to length of url - if your only option is a .com that is 25 letters long, or a .io domain that is six letters, then the .io wins.

For most of the cases in which a .com ends up being inferior to a .io, it's due to a naming exploration failure. You can still easily find good two word .com addresses today. Unless you've got coder.io or similar.

> 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.

That's not a straight comparison. A better stat would be, what percentage of the top 20 YC companies had the .com while they were in the active YC batch?

No, because his thesis is "change your name", not "start out with .com"
But he's implying that having a .com correlates to success - so % starting with a .com would be a more apt comparison.
Change your name isn't enough you have to first find a name that you can afford. And so while it's true your current name isn't that important it doesn't mean that you can find any "good" .com names that are affordable.
I think a lot of the comments here are missing the point.

If somebody is already using the <yourname>.com domain, then that name is already being used for something and creating a new business with the same name is a bad idea. I guess in the worst case scenario they could file a lawsuit claiming that you're infringing their trademark.

People can try to justify it by saying everybody just searches anyway, but if the .com name is already established it'll be a never-ending battle to keep your name in the top search results.

I mean, nobody would try to use ford.io or yahoo.tk for their startup, doing the same thing but with a smaller company is an equally bad idea.

Yea, I would agree that this is really the key point. If it's a fallow .com name (that you can't afford), then maybe you'll be lucky with an alternative TLD. But it's probably a poorer choice than changing your name.
In principle, agreed, but you are neglecting the huge role played by domain squatters.
The correlation might be backward: the top 20 YC companies by valuation probably have enough cash to buy the domain name that matches their company name.
It's also wrong because it implies that having the .com today is as important as when those companies got their .com. Anecdotally, many (most?) developer tools companies go for .io, so much so that a .com sounds strange these days.
Exactly. Maybe it's because the top 20 stuck around in YC long enough to be badgered by Paul Graham into getting a .com?

That said, I don't disagree with him on .com, I just don't think this particular factoid adds anything to the argument.

And for at least some these, it's definitely come post-success.

Twitch.com redirects to twitch.tv, but was a music site as late as the end of 2014.

Dropbox was GetDropbox.com < at least 1 counter example.
Not sure the domain name really matters much now. Reasons:

(1) Do users really pay attention to the extension, COM or anything else?

(2) IMHO, now users mostly just click on links while paying little or no attention to the actual URL or domain name. Of course, the HTML link element (tag) doesn't have to expose anything about the extension.

So, why do 1+ billion users know or care about COM?

For PG's point about the most successful YC companies use COM, that was then, when COM didn't cost so much, not now.

I'm not sure.

I'd rather have a memorable name with a sub-par TLD, than I would have a confusing / hard to remember name with a .com.

Obviously, the ideal is memorable + .com, but most of those will cost 50-100k+, which is not feasible for seed-stage budget.

Sounds like a series A problem.

According to the article, there's a third option:

> It turns out almost any word or word pair that is not an obviously bad name is a sufficiently good one, and the number of such domains is so large that you can find plenty that are cheap or even untaken. So make a list and try to buy some. That's what Stripe did. (Their search also turned up parse.com, which their friends at Parse took.)

I have a related question. When you have one product, which name is different from the company's name, is it better to use comapany.com or product.com?
Perhaps both? Company.com can describe the company, show your products, etc. Product.com could be either a landing page advertising your product or, if it's a service, perhaps that's the login / primary access for your product.

Just my thoughts.

I couldn't disagree more with this article and am kinda surprised Paul wrote about this.

Domain names are transitioning into becoming useless. When you open any web browser or mobile phone and go to look for a company it does a search. The most relevant results come back at the top. This type of trend, as we abstract away from using TLDs, is only going to continue. Owning a .com isn't nearly as meaningful today as it was 5 years ago.

What I think will happen is ICANN will continue adding more, useless TLDs as quick cash grabs and the next 10 years you're going to see people stop referring to their domain name at all. I mean, why would you? You're Uber, you look for "Uber" in the app stores, you open your browser and type in "Uber". All of these take you exactly where you need to go.

> 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. Which suggests there are lessons ahead for most of the rest, one way or another.

Correlation is not necessarily causation; this statistic should be compared with general domain name availability with the time periods mentioned because an incredible amount are purchased every year making it impossible to even compare in this way.

Yeah. I heard this 13 years ago too. http://shirky.com/writings/domain_names.html
> Yeah. I heard this 13 years ago too. http://shirky.com/writings/domain_names.html

So because someone suggested it 13 years ago makes it not true today? Look at how the new TLDs are working out; if you want to find a company using a funny TLD you just search for their name and unless the .com is higher ranked the company comes right up to the top.

Look at it from a user's perspective; do you know anyone who actually types in a domain name anymore that isn't a technie? If a user wants to find a specific company they just simply use voice and ask their phone or open a browser and just type it in.

Typing in a TLD adds friction especially since ICANN added all of these stupid TLDs that no one is even used to. The lower the friction the better, right?

The absolute lowest friction is to have a .com domain. And yes, it's still true.
I'm not convinced; if you look at Google Trends there are many, many company names in the top 100 searches. Why are people only searching the company name instead of going directly to their website? Typically their website is the first several results depending on the size of the company.

Also, only anecdotal because I can't find any studies one way or another, but I can't find anyone in the younger generation who actually types in any domain names. I even had a frustrating experience telling someone who's used the internet all his life what .com even was as he had never typed it before. Plenty of people in my generation still type it though.

Also many of the older generation folks I've worked with who had their internet start with AOL don't seem to type in domain TLDs either.

So until I see some solid, current research regarding the behaviors of the different generations I'm not convinced a .com is lower friction; you type less when you just type the company's name and most browsers nowadays will give you the option to autocomplete directly to their website.

> do you know anyone who actually types in a domain name anymore that isn't a technie?

That's a problem. It means we rely more and more on search engines, thus centralising the web (and with it, the internet) more than it needs to be. Centralizing the network means centralizing the power. Not a good idea. It never was.

DNS is already centralized.
Yes, and that is a problem. But that's not nearly as bad as having Google and Bing have the same role. DNS goes through a highly decentralised network of name servers, many of which aren't at all controlled by the ICANN. Each country based top level domain has its own set of servers, and there name requests are generally cached on other servers still.

But when you type "company name" in a search engine, you give away your search to an ad supported private company in real time. Juicy profiling, no privacy. Much worse than using DNS.

Besides, when you click on a link from your favourite search engine, you're still using DNS. Now you have two single points of failure.

I don't see the problem with G and B in this role; at least they answer to the end user. Who does ICANN answer to? What is your alternative if you get crappy service from the .com operator? If what PG says is correct, you don't really have an alternative.
> at least they answer to the end user.

Yes they do. Just don't forget that the end user is the advertiser. We poor schmucks who type text in the search box are the product.

That's a different kind of centralized.

Everything going thru search engines, means you depend on someone who has no responsibility towards the people they have power over. That's bad; you have to waste time understanding their incentives and staying on their good side.

Err, I've just read this article, and it seems its main thesis is, domain names are probably not going away. Because people will likely always need memorable, globally unique identifiers. I didn't see any talk about .com going away either.
Exactly! Referring to your domain name is something you do so rarely it's not even worth mentioning. Most of the time people get redirected to your website, it's via bitly or a number of other url shorteners.
It seems like we are going back to the AOL keyword days... not a good sign.
> It seems like we are going back to the AOL keyword days... not a good sign.

Not really comparable in my opinion. In AOL, if I remember correctly, you had to register or pay for your keyword much like a domain name. So all keywords were owned by a central authority, could only be used within their ecosystem and had the same type of squatter issues as domain names.

What's happening today are "keywords" being associated with specific websites when, and only when, the users are actively looking for and using that website. This isn't centralized (sure most people using Google but people forget Bing is actually really good and used by many!). You can't simply pay for the top spot anymore.

You can "simply pay for the top spot". This is why the results at the very top of the Google search are paid ad placements.
I was referring to search results not advertisements. You can pay one fee, like a domain TLD or AOL Keyword, and get that spot. With those advertisements you have to constantly put money in to stay at that spot.
Domain names are also a constant, on-going cost.
While there is an on-going cost (i.e. domain renewals), it is very small compared to the initial payment one might make to acquire a premiere domain.
Sadly my parents don't know that the top yellow box is an ad, they think it's the top search result. I expect most people are the same.
Does it really matter, though? If they find what they're looking for, what difference is it if the result was paid or organic?
In most cases the ad is not what they are looking for.
> You can't simply pay for the top spot anymore.

There's ton of money being made (sometimes even by honest businesses) on the opposite assumption. Instead of paying the gatekeeper to put you higher on the list you pay some third parties to game the sorting algorithm.

You're assuming most people search first; many users don't even understand the difference between entering a search term versus a URL in a browser address bar. Yet, non-techy people know ".com".
I dont have the data but I would argue such minority is very less and even if they are unable to find what they are looking for in first attempt (by using .com), they do check search engines result afterwards.
I think it is the opposite. Among most of the "non tech" people I know, they enter a name at a google search box rather than a URL. Even if they know the URL, they enter it at a search field!
Correct. I have seen this with my own eyes. Me: Go to Facebook. Them: Google, Search Box, Enter Facebook.com in search box, Click.
Actually many non-techy do this:

1) Open browser, 2) Defaults to whatever search engine that add-on they have switch them to, 3) Type facebook, or facebook.com, or whatever they are looking for, 4) Click first result. If fails, go back, look around, click other result.

It sounds (unsurprisingly, of course) like the perspective of an investor or other tech industry insider. Notice the phrases like "it signals weakness" and "suggests you're a marginal company." Those don't seem like phrases most people are ever going to think about a company, and especially not based only on their domain name.
I disagree.

My parents are typical web users. Fairly tech savvy, but not tech insiders. I recall telling them about bit.ly. It took a while to explain that it wasn't bitly.com, bit.ly.com, etc. They ended up just searching for it instead of typing in the domain name.

Think of the added friction this causes... and as we all know, every added interaction detracts from the conversion funnel. Taken in aggregate, many startups cannot afford the attrition of a shitty domain name.

Isn't that a slightly different point? You're talking about ability of a user to discover or find the site, not the user's perception of the strength of the business.
...but the point is that to some non-technical users a "new" TLD won't even look like a full URL.
That is beambot's point, and it's valid, but it's different than the point in this essay.
Don't typical web users these days always search instead of typing in the domain name?
They ended up just searching for it instead of typing in the domain name.

And they found it. That's how a large, large number of people find web sites these days, and it works fine.

In B2B weakness in the form of undercapitalization is important because companies don't want to make middle term or long term bets on a technology with uncertain prospects for support, maintenance and further development. Signaling "this product might be legacy in 18 months" is worth avoiding.

A rose by any other name still smells sweet, but thornbush isn't the way to market it.

Insiders usually have ungainly phrases for things that outsiders sense but cannot/do not put into words themselves.

The public are hugely influenced by branding, advertising, look and feel, and experiences. They might not put the feelings they have into words like insiders do, but those things absolutely exist and insiders to the retail, advertising, and similar industries have all sorts of jargon to analyze the behavior of the public.

I agree with you.

Name of the company and domain name has very little to do with company's success. There so many things you need to worry when company is small and getting domain name or renaming company is really very dangerous distraction.

I'm not saying domain name does not play some role but spending time on it is a very dangerous distraction.

So, does it show as weakness? Yes. Is that biggest weakness of your company in relationship to how hard to fix it? No - but if answer is Yes then you are already made it and all this discussion is moot.

An easy thing for a well funded start up to spend money on is their domain name. A good question to ask is when did the top 20 YC companies by valuation buy theirs. Did they pass over really talented devs just so they could drop $500k on a dot com?

Things are changing. Talk to 18-20 year olds about how they are using the internet. Throw out an address using one of ICANN's new extensions. Do they even recognize its a website? EV Certs don't even show URLs in some browsers. Do mobile users look for you on Google, or in the app/play/other store? Personally I'm uncomfortable with the positions ICANN has taken.I don't trust them to protect domain's owner rights. We need better alternatives.

(comment deleted)
Google's search results page is pretty much a DNS Server.
You completely ignored PG's argument about it signaling weakness. This is one of those perception versus reality things, where people's perceptions (those who are judging a startup) ARE the reality, whether those perceptions are theoretically refutable or not.

In other words, let's say you're right, and owning the .com domain name is becoming less meaningful (an argument that PG already addressed, by the way). That doesn't mean that .com domains are becoming significantly less meaningful to people's perceptions. And perceptions matter, when going after customers, funders, employees, business partners, etc.

I wouldn't say that the perception that having the .com domain name is important is an incorrect one. Maybe you would disagree and say it is an incorrect perception. But even if you think the perception is incorrect, the perception is still part of the reality that the startup lives in.

This classic "perception IS reality" pattern is something that a lot of people miss.

signaling weakness is an important factor for a fraction of companies for most others that is the very least of their worries.

PG talks from a perspective that does not apply to most companies in this world.

Again if you get a .com name that is great. But whether it's worth the effort of finding one it if you don't have it already is another discussion all together.

> But whether it's worth the effort of finding one it if you don't have it already is another discussion all together.

Whether or not you agree with pg, finding a decent .com is pretty cheap in terms of time and money when you have free domain name checking services that give you available .com variants of desired keywords. Even though I know this, I'm still always surprised when I see the 2-3 syllable .com domain names that I wanted, being available without auction or private seller inquiries.

Thats just not true. It's not some universal law you can apply. You have to take into considering the amount of time and what kind of name you have access to.
> You have to take into considering the amount of time and what kind of name you have access to.

I'm not sure I understand. Free name checking services do take into account what names are available to you and they give you variants of your desired keywords. Results are returned in seconds. It takes a few hours at most to go through a bunch of keywords.

(comment deleted)
But changing perception is in our hands. If everybody starts not giving a F for the .com, then people will start not caring about that... If we keep fighting for the .com we are just supporting that ugly business of buying nice domain names in order to sell them for a ridiculous price to people that actually want create something.
When you're trying to get your company off the ground, fighting the good fight for the betterment of other companies is not something your can afford.
re:Signaling weakness.

True, but to whom. Most people don't know the difference between getDropbox.com and Dropbox.com.

Investors, is I think the implication.
I agree here with Natch, compare point about signaling weakness with sending your CV from "fancy_kid_45@gmail.com" to "firstname.lastname@gmail.com" or even "firstname@lastname.com" what looks most professional? Can you imagine reading CV of a person with fancy kid mail?
> Can you imagine reading CV of a person with fancy kid mail?

Yes. In fact I did quite a bit of hiring at some of my previous companies and I never cared about email (why would I care about email? This isn't a signal, this is noise). Do people, at least in tech, really care about such things? At least in my deals I never found one. I'd imagine in stuffy jobs it might be something to look down upon but in tech? That seems silly considering how progressive our industry typically is.

If you want to hire someone detail oriented will you look inside his CV if he has "detail oriented" listed or you just see he has typos?
Wouldn't appearing weak by not shelling out potentially tens of thousands of dollars for a domain name be, well, accurate? I mean...these are startups, not established companies. I think showing weakness could be to the advantage of a startup, which PG seems to ignore.
How is showing weakness an advantage?
It can signal value, though you might end up chasing worse customers.
How does it signal value?
I think it can signal breaking the mold. "We are new, we are not constrained to use a .com domain, come and find cool stuff here"
(comment deleted)
It can signal a prioritization of values. Startups tend move fast and be flexible, and tend to have less customers so each one is more valuable. Established companies tend to move slowly and often provide worse customer service. By being honest about being a startup (or being an established company) expectations can be set between company and customer. Pretending to be established when in reality you are not seems disingenuous, and I think it's pretty easy to argue that being genuine is good/can be an advantage.
I don't really see how .com or not translates into being genuine or not.

Having been a founder for four years now, I have learned that the word "startup" usually scares potential customers. It is actually a good thing to seem more established than you are. However that doesn't mean you are disingenuous either. It is not black or white. You can look more established: .com domain, decent website, proper terms and privacy policy, no beta or "coming soon" wording on website, etc. However if someone asks directly how established you are, you are truthful.

What the potential customer wants to know is if they can trust you, and you won't be gone tomorrow. Therefore if you are truthful about the company's state, but seem like you are established then more people will look past the word "startup" since it will be obvious to them that you are serious and have some idea what you are doing.

In other words, you dont need to go around yelling at the top of your lungs that X is a startup. That strategy will most likely fail. But you certainly shouldn't deceive people either. There is a line.

I suspect that your parent is conjecturing that the fact of not having X.com will eventually stop being perceived as a sign of weakness.

I do not know to which extent I agree. I find it sad but likely that people may eventually stop paying attention to URLs altogether; the main counter-argument I would have would be phishing, because, no matter what, you'll still have to check that you are on the right website at some point. That said I do believe that if it may happen for the general public, it may not happen before a long time for more knowledgeable people (prospective funders, employees, partners, etc., as you mentioned).

I didn't ignore PG's argument at all. My point is how can there be any perceived weakness when the thing in which we're calling a weakness doesn't exist to most users?

When I, admittedly anecdotally, poll people I've met who were practically born after the Internet started, almost none of them even type in .com and a few didn't even know what that was. I'm convinced this sentiment is going to become more common especially with so many squatters on .com domains it's almost impossible to have a good company domain name anymore without spending stupid money to get it. Yet you can get the same name with a different TLD for cheap and you show up at the top of search results.

I'd love to see more search engines or maybe even a decentralized search infrastructure to help facilitate this but that's pretty difficult.

The preponderance of the evidence suggests (proves?) you're wrong.
Remember this is in the context of startup businesses who may not have much on the web. Your position works out okay if you are confident in google's "I'm feeling lucky" kind of search. For everybody else, having a solid .com name still reigns supreme.
> Domain names are transitioning into becoming useless

That's an astute observation...however, they are not useless YET. Perhaps in 5 years .com's will be obsolete and every browser will have powerful company lookup utilities that everyone can use. The issue is that startups don't have the luxury of losing that sort of time.

The latest "new TLD" thing was a money-making scheme for domain registrars. Nobody goes to those domains. Even the previous round of TLDs (".museum", ".aero") are not used much; ".aero" has entries for all the major airport codes, but they're redirects.

Amusingly, even if you have your very own TLD, it doesn't help you. Most browsers send single-word lookups to a search engine before they send them to DNS. Try "ca" in a browser. There is a web site at "ca". But to get there, you'll have to use "ca.", to force a DNS lookup from the root. Nobody knows to do that unless they're into how DNS works. There are some low-level bugs embedded in very common C libraries which cause problems with single-word domains, and they probably won't be fixed.

As for the original poster's advice, the price of the domain you want in .com is probably more than YCombinator's initial funding. With all of today's domain hoarding, the ".com" domain usually comes later. Facebook started as "thefacebook.com".

There is no website at "ca". Did you mean www.ca? That redirects to circa.ca
(comment deleted)
For a clearer example, try "http://ai." No redirection; there's a web site there.
There's no DNS record for "ai.". It's actually the browser prepending www in search for a site and "www.ai." happens to exist.
No, there is a DNS record for "ai":

    $ nslookup ai
    Server:	127.0.1.1
    Address:	127.0.1.1#53

    Non-authoritative answer:
    Name: ai
    Address: 209.59.119.34
Rechecking using Google's DNS server:

    $ nslookup
    > server 8.8.8.8
    Default server: 8.8.8.8
    Address: 8.8.8.8#53
    > ai
    Server:	8.8.8.8
    Address:	8.8.8.8#53

    Non-authoritative answer:
    Name: ai
    Address: 209.59.119.34
It's pingable:

    $ ping ai
    PING ai (209.59.119.34) 56(84) bytes of data.
    64 bytes from offshore.ai (209.59.119.34): icmp_seq=1 ttl=46 time=133 ms
    64 bytes from offshore.ai (209.59.119.34): icmp_seq=2 ttl=46 time=132 ms
    64 bytes from offshore.ai (209.59.119.34): icmp_seq=3 ttl=46 time=132 ms
    ...
And it speaks HTTP:

   $ wget ai
   --2015-08-09 14:26:26--  http://ai/
   Resolving ai (ai)... 209.59.119.34
   Connecting to ai (ai)|209.59.119.34|:80... connected.
   HTTP request sent, awaiting response... 200 OK
   Length: 1257 (1.2K) [text/html]
   Saving to: ‘index.html’
   100%[======================================>] 1,257       --.-K/s   in 0.001s  
   2015-08-09 14:26:26 (842 KB/s) - ‘index.html’ saved [1257/1257]
It doesn't respond to HTTPS, though. So we don't get to see a cert issued for a TLD.

A base level TLD site is rare; most TLDs don't have one. But it's supported in DNS.

Well, at least prg.aero is actual website and not redirect (which may have something to do with the fact that ownership structure of the airport had changed around the time .aero was introduced).
The advice seems like a bit of an anachronism in the mobile world.

Nice, short .com domain names were probably more important for desktop web browsing -- and I suspect are less relevant to startups that are focused on the iOS App Store, Google Play Store, third-party app stores, Twitter, Facebook, etc. for discovery.

exactly ..people go to the app store. Few ever go to snapchat.com They hear about it on Facebook, on twitter, the press, through word of mouth
majority of people just enter the name in their address bar and click the search result. domain name is not as important as it was.