More than half of YC startups don't have a dot com | Hacker News Reader

More than half of YC startups don't have a dot com

reproof.app

106 pointsduvander107 comments

I think this is honestly less about the notoriety of .com, and more about the lack of affordable and available domains for startups to purchase. Got a five letter named startup? That’ll cost you tens of thousands of dollars, IF you can even get it. Meanwhile there’s hundreds of gTLDs that you can buy a domain on for $10-$30/yr.

Having a dotcom is a signal of the organization's wealth. If your company is worth $3B, and your cash position is good, spending 0.01% of your valuation on a dotcom ($300k) is a no brainer. It might add to your legitimacy and prestige (even if marginally) and it's an asset that will most likely increase in value over time.

> an asset that will most likely increase in value over time.

This is actually a myth. See the changing of hands of sex.com. Also, Domain names are being less important as users stay within apps / content platforms like FB, Instagram and TikTok. Also Google Search and soon ChatGPT.

Reminds me of nissan.com, which is now nearly worthless because Nissan Motors doesn't care anymore.

I thought that was what Nissan (the domain owner,not the motor company) wanted, that motor company leaves him alone, & don't bother him while he uses his lastname domain.

If I remember right he owned a computer company he used the domain for, but he pretty much devoted his domain to his anti-nissan campaign until he passed away.

Or steam.com

The future value of domain names aren't certain, but I think they are a fairly safe bet. They are viewed in to email addresses, search results, web address bars, and hyperlinks. I don't think a platform is going to usurp these. Walled gardens have been attempting to do this since AOL keywords.

The domain name system predates the Internet. Just food for thought.

Makes sense. Nobody just blindly navigates to a domain name. You either know it for sure or you search.

> Having a dotcom is a signal of the organization's wealth.

The pricing of .com domains is extremely uneven, depending on your name. It could be $20.

> legitimacy and prestige

Since it’s semantically equivalent to other TLDs, it lies entirely in the eye of the beholder. This non-.com ickiness seems to be an outdated US-centric perspective, never heard of it elsewhere. The rest of the world are used to other TLDs, and some (say ai, io, app) indicate other positive business values.

> If your company is worth $3B, and your cash position is good, spending 0.01% of your valuation on a dotcom ($300k) is a no brainer.

So an established company should switch domain name and you’re not factoring in the cost of that? Or they should just have the .com on the side (but then how will the majority of your customers know about your state of the art .com domain)? Or should startups buy a $300k domain upfront?

> This non-.com ickiness seems to be an outdated US-centric perspective, never heard of it elsewhere.

Not exactly true, for example for most German users not having a .de is odd. .com and .net already feel like a stretch for many. Beyond that other TLDs/gTLDs are not really trusted by most from what I can tell.

Speaking of (and confirming your point), ai.com redirects to chatgpt as of ~1 week ago...!

https://mashable.com/article/chatgpt-ai-dot-com-domain-name-...

If my company is every worth $3b then I'll think about buying the .com for $300k.

However right now $300k is much better spent on staff, hosting and other business things.

Moreso if you're a $3bn company, you certainly don't want _someone else_ owning the .com for the same reason — it lends legitimacy.

Even if they don't use the .com as their primary domain, they'll still buy it and redirect so that no-one else squats on it / phishes from it.

Why would having .com domain add any prestige?

Because there's only one. It's like a bunch of Chinese restaurants fighting over the name 'CN Wall', but there's only one that'll be able to hold 'cnwall.com'.

(~$5k).

this, .com domains are an ego thing more than anything at this point. Why waste valuable money on a vanity thing rather than trying to fuel growth? I guess in the worst case somebody could try to trick potential customers but otherwise as an early startup you have bigger things to worry about than a domain name

They are artificially necessary in some cases.

I have a personal .com and a gTLD based domain. THis is mostly to retain control of mailflow (I use workspace but can always move it, and have in the past)

The number of services/online forms that wont accept a .email domain is still very high.

I would imagine at a business level this would be equally difficult, even if its just for B2B type transactions/relationships.

Same. I have a .family domain which is great, but about 1/3rd of the registration flows claim it’s an “invalid email address”. I have an older .org to fall back on, but it would be nice to use one domain for everything.

Imagine spending hundreds of thousands of dollars to get your own TLD and then discovering that "username@tld" emails just won't be accepted by basically anything. RIP username@google :(

That is indeed true for gTLDs (I've experienced it with .online), but from my experience never the case for ccTLDs. Maybe .tk that is (or was) handed out for free, but I never had any problem with ccTLDs I've owned (.me, .dj, .sh).

So, .com isn't the only silver bullet, and from what I can tell from that chart, only .app and .tech may be somewhat affected.

interesting. Just registered a couple ccTLD's.

I have a .com. But its super long and I dont really like it. Its also the primary domain in workspace since i have had it since gsuite was free but before they closed the free tier registration ~2012.

Now that I'm a paid user (by forced attrition) im looking to honestly just migrate the TLD of the workspace to something less cumbersome to type.

.email is really the primary at this point.

.in is rejected by discover for email address, its a cctld. .us was fine.

Can confirm. My .email is regularly reject even by my own country administration.

This. Discover bank literally said email ending with .in is not a valid email. As if no Indian moves to US & still keeps his own domain.in email address. It happily accepted .us email.

getting a .com means there s less chance of it being trademarked also. and it's not that hard to come up with a decent name that's not taken (eg. airbnb)

Why choose a name that's unavailable? pretty much every tech company has a weird .com name, and that's fine

No, it's not ego. Normal non-tech people are untrusting of anything not .com

I also believe that changing your domain after you build an user base is challenging from a UX perspective. Many users have saved their credentials on their browser's password manager, and if the domain changes, the auto fill didn't work anymore.

Courious about why Notion owns the .com but still using the .so.

To prove your point I got borg.games for $20/yr for my p2p cloud gaming startup yesterday. Although lots of other good names were "premium" and costed upwards of $5000.

borggames.com is available and will cost you $10/yr.

Meanwhile my four letter domain, lwei.com, gets no offers and is appraised at like $1000.

There’s still stuff available for reasonable prices but it’s just a lot of work so it’s easier to get some random tld.

Well sure you can also run a random string generator and get a short .com TLD, but that's not the point.

No 3 letter.com domain is available to register (all combinations are registered, can only be bought). 4 letter mixed numbers letters are sure available. I am not sure of 4 letters only is available. 5 letters only sure is available.

I bought a three letter domain at one of the original TLDs for $3k or so, they can be surprisingly reasonable depending on your use case.

Pretty sure all 4 and 5 letter .com domains have been taken since at least a decade.

4 letter ones are all taken, but there are still 5 letter domains available. I wanted a short domain for personal use and bought xlmpq.com

A fellow ascii art enjoyer :)

And very few of them will ever serve anything useful.

So many random 5 letter, letters only domains are available. If you add numbers & hypen to mix, it baloons to thousands.

Today I realized how xkcd prob got it's name

There’s a faq on this on the site (find it yourself). It was created as an unpronounceable “acronym” (although it doen’t stand for anything) and the domain came later when Randall was still an unknown. It’s old enough that a lot of “normal” domain names were still available, but the name itself was one of the jokes.

Possibly, although the official explanation is that it's "just a word with no phonetic pronunciation". There's a "What does XKCD stand for?" section on the About page. https://xkcd.com/about/

Randall wanted a unique short string associated with all his work. That way he could Google "xkcd" and find everything he ever did online. I believe he had a script that generated all possible 4-letter combinations and filtered out ones with any search results, and settled on "xkcd". But he became a victim of his own success when the comic became so popular and "xkcd" is now synonymous with the comic.

Source: talk by Randall, might be this one but I'm not sure: https://youtu.be/zJOS0sV2a24.

That's Iwei robbery!

"our company is called Iwei, and you can reach us at Lwei dot com"

Appraised by whom? What's a domain's value if no one wants to pay for it?

When I started my YC company, we used the .io (which we found on HN, actually! https://news.ycombinator.com/item?id=6397526). A few years later, once we could afford it, we bought the .com for $170k.

I imagine that if you filtered based on only companies that make it to (for example) a Series B, the % goes way up.

I'm honestly shocked you decided to pay that for a domain name. I guess I can kinda understand as a percentage of your company revenue, but morally that is so hard for me to justify feeding a squatter, let alone in a sum of that amount.

I don't love it either, but we didn't really have a choice. All simple .coms are already owned at this point, and I think with few exceptions pg's article is dead on:

http://www.paulgraham.com/name.html

> A few years later, once we could afford it, we bought the .com for $170k.

How are these bigger ones sold? Do you sell it back or to someone else later?

I might be misinterpreting the question, but the way to buy an expensive domain is to go through a broker.

We bought it outright (sent money to escrow), but a lot of people will do rent-to-own. For example, we would have paid $10k/mo for it, and then when we hit $170k it would be ours. But in the meantime, if we missed a payment, the rights would revert back to the original owner.

From my perspective as a user, maybe someone who would visit your domain, I’m indifferent to the specific TLD used. You might be able to compose a clever word that grabs my attention, which I’ve seen a lot from .ly domains.

However, from an email marketing perspective, TLDs are not interchangeable. Some TLDs (think .xyz) see a lot of abuse, tanking the reputation and deliverability for good actors using them too.

So my advice is to stick with TLDs that most good actors are using (.com, .io), giving you better baseline deliverability, because email marketing remains incredibly valuable for growth.

As a personal user of .xyz I couldn't agree more. People often get a chuckle at my web/email address when I give it out. Sometimes I have to repeat it because (especially non technical) people aren't expecting it to end in .xyz but often it works to my advantage. The repetition and uniqueness makes it memorable. Also .xyz is generic enough to be used for business and for personal sites. It's also not a ccTLD so I don't have to worry about registration restrictions.

But... Even with using a correctly configured major provider to deliver my mail, it gets sent to spam all the time. Which for personal email, fine. Most of my recipients know to check spam, and mark me as safe. But I know of at least one destination that flat out rejects my mail and if I were a business that'd be nightmare. Also on the web front a lot of proxy and filter providers just block .xyz all together.

When I purchased it I hoped Alphabet having abc.xyz would've helped with .xyz's reputation, but it hasn't. Oh well. Mine is for personal use so I'm willing to tolerate it. Businesses be warned though.

My company uses xyz for staging and one of my colleagues couldn't figure out why he couldn't get the app working locally (some local services reach out to staging) and we found his ISP was blocking xyz

The problem with .xyz is that it’s so cheap, it’s a common choice for spammers/scammers to the point where wikipedia has a blanket ban on links to any .xyz domains unless explicitly whitelisted, even on a user’s personal page.

It may make sense to put bulk email on a completely separate domain anyway, so then you could do something like:

    example.tech
    exampletech.com

That has it's own downsides if it confuses your users, but most email clients barely show sender addresses any more, so a lot of users may not even realize you're emailing from a different domain than you use for your website.

Doing support, if customer has subscribed for example.com, we will not associate support requests from @example.tech as legitimate…

IMO the biggest problem with anything that's not a .com is the registry owners. As a registrant I want low (or at least consistent) prices, minimal price increases, and an assurance the TLD is going to be around long term.

The registries constantly adjust pricing and the new registrant agreements aren't as good as the old ones (ex: no price caps). The registry operators probably think they're geniuses that are maximizing profits via price discrimination, but I think they're fools that don't understand their product or the target audience.

I don't care too much about premium pricing. It's the *fluctuating* pricing without any limits that turns me off. If shortsighted management ruins a TLD and it collapses, I don't want to be collateral damage. I know there's a bond to cover operations for a period of time, but that's not forever.

For example, see the .audio domain or any of the other Uniregistry domains that went through a massive as much as 1400% price increase in 2017 after people had already bought domains. https://blog.dnsimple.com/2017/08/uniregistry-tld-price-incr...

It caused quite a stir. I luckily was able to get grandfathered pricing for a .audio domain through my registrar (namecheap). Although their system did fail at one point and charged me the non-grandfathered price at one point, and I had to go to support for a refund. What a headache

This is a good thing IMO. There are hundreds of great TLDs [1] just sitting there that can help ease the gold rush and squatting problem with .com domains, but are not favored because historically non-.com websites have been associated with scams and low effort content rather than established brands. Hopefully more and more legit startups out there with a .horse their name will finally normalize such domains.

1. https://en.wikipedia.org/wiki/List_of_Internet_top-level_dom...

I completely agree.

Back when the first wave of quirky TLDs was created, I remember a fair amount of HN users insisting there was no use case for them [0]. But I always thought our sense of normal would shift as time went on and new sites and apps became popular.

0. https://news.ycombinator.com/item?id=7492226

This .com=serious shtick is laughable, your fomo tricks have no power here. So why would I get a .com? Nobody even types the full domain anymore. People search and click. But the number one reason is I refuse to feed the squatters.

People used to spend hours to customize their phone backgrounds and desktop backgrounds and myspace profiles, then came the big guys and now everyone uses the defaults and all webpages look alike.

Getting a non-.com domain is like that, it seems amateurish and somewhat unprofessional. There are .com domains if one is willing to be creative and settle for something that is not a simple dictionary word . The length doesnt matter as long as the word is reasonably memorable. 'facebook' and 'instagram' and 'snapchat' did not have a problem with their name. If they 'd gone for face.book, instagr.am snap.chat etc people would be confused "how do i access this site". The dot is not part of our mental dictionary.

Some of them look even dangerous. for example MS onedrive links are from "1drv.ms " which makes my emails look like phishing scams.

That first line got me, everyone used to have ringtones and one day I just left my phone on vibrate and never turned it back on.

Funny fact about me is that a certain Motorola ringtone from the mid 2000's will trigger a panic attack, old TV shows especially. And that's why I have a rule to pass the on-call to the next person in rotation if someone had a really bad night.

Related, PG’s “Change Your Name”

http://www.paulgraham.com/name.html

That post said, in 2015:

> 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.

Compared to the article:

> Today, only 42% of Y Combinator startups in the winter 2023 batch use a .com domain, while only 40% of the most popular Product Hunt startups used a .com in 2022.

which supports the assertions in the article. It's still true that there are only 3 YC companies in the top 50 that don't have a .com, so maybe the lesson is "just plan to buy the .com later"!

.com domain availability has dried up and alternate TLDs dropped the price of squatted .coms. .io was too popular for a while and now everyone seems to be going with a .ai even if their product is far from actual AI. For my startup, I went with .dev. Don’t know if that was a good idea, but I was able to readily get a simple three-letter TLD there.

I don't know much about the marketing potential for .dev vs others, but I think .dev is one of the safest TLDs in terms of registry management and longevity. It makes sense for Google to keep their TLDs priced reasonably to keep the web accessible and there are some big anchor domains on .dev; web.dev, pages,dev, etc..

Google's also the most likely to suddenly decide they don't want to be a registrar anymore and just graveyard the whole TLD. They've done similar to bigger things.

In addition to the price issue folks have mentioned, I also think it's less important to have a .com if your product isn't consumer end-user focused. E.g. if you're selling to developers or other businesses things like .io or .ai are fine. My guess is that .com is still important if you are selling something direct to consumer.

More than half of YC startups don’t have a dot com for the same reason people aren’t building new homes on Manhattan island - there’s no empty space and buying existing desirable space is unnecessarily expensive.

that happened decades ago. squatting is not a recent thing

It didn’t say it was squatting. Just like property in Manhattan isn’t being squatted on. It’s being used.

I think that the fact that every possible .com is probably already taken has played a big role in this. On top of taht, who types in the domain name? You either go to search engine, click the link somewhere, scan QR code.

"A rose by any other name would smell as sweet" as William Shakespeare has written could be said today "A startup can have any domain name and it would still be the same".

Entrepreneur Pieter Levels posted a before / after screenshot of his traffic after switching from .io to .com. It's quite the jump!

https://mobile.twitter.com/levelsio/status/14722048099817594...

There's an old game fan site that had an unpronounceable five letter dot com. I made a lowball offer and since then Unregistry has periodically tried to sell me this expensive domain to me, including by setting up a fake site temporarily to make it look like it was in use.

I picked up/resold a lot of expiring domains in 2008 when people couldn't hold on to them/stopped speculating. Made around $30K off some scripts that took expiring domains and parsed english words, grabbed the number of return results for those words in Google as well as cost per click and some other mojo to determine potential worth, pick up the decent ones and send emails to relevant companies. I'm no longer that scummy but I wonder if there will be an upturn of lapsed domain renewals this downturn? I'm guessing there's a ton of competition now (even from the Registrars) as even then I think a lot passed through Godaddy auctions or some such before dropping of.

I'm in the domain industry (created a tool called Newsy - https://www.newsy.co) and this is where the domainers take the opportunistic sales to established companies who can buy their .com domains they've been holding on for so long.

You may not like domain parking industry, but for people who have decent set of .com domain names and many of them generate thousands of dollars through just ad-clicking, it's an absolute no-brainer to continue on with this trade and they will always say "it's an investment" and "it's like real estate".

Just a heads up, the url you provided doesn't hit a website. Gives me the error "www.newsy.co redirected you too many times" on Chrome.

What users even look at a domain, much less worry if it's a .COM, anymore? In the age of apps and abhorrently priced .COM domains, it seems to me that a new company with a relatively low amount of startup cash shouldn't choose to do something as idiotic as "investing" tons of cash in a domain name. Just buy a .gg or .co/ etc and get over it. If you're successful, and when you have the cash, then go wasting money on stuff like that. Not all of us have billions donated from large companies to spend on domains.. cough cough

We bought our .com after we became "successful" and had significant revenue, however it just redirects to our non-dot-com, because it's too much of a hassle and SEO risk to change at this point.

I wonder what the updated figures would be.

[1] http://www.paulgraham.com/name.html

YC also mentioned more of the cohort being from outside the United States, which would mean a higher proportion of regional domain names.

I found it to be really a helpful piece of info. If I only had access to it a year ago when I AskHN-ed [0] regarding this same matter.

In the end, I ended up purchasing the corresponding names in .io and .app when still available.

[0]. https://news.ycombinator.com/item?id=29758152

I wonder if it would have been seen as a sign of prestige for startups if yc had become a domain provider for .yc

Two-letter TLDs are reserved for countries. IANA has a slightly more generous definition of country than the UN, but they're still not generally available. Definitely a nerd status symbol if you can get your hands on one!

Honestly that was a bad decision and it's time to end it.

Nobody uses those for their intended "national country" use. .io and .ai are tech-focused, .tv pretty much only has tv stations or video providers on it and so on.

Most countries except the USA have their local domain in general usage.

Even the "hacks".

Search "site:.ai Anguilla" for examples. You obviously be little reason to visit websites of Anguillan restaurants, government, telecoms, tourism etc but they exist. Within Anguilla it's a small signal of a local business.

> Nobody uses those for their intended "national country" use.

Most of them do? I'm sure most .fr or .jp properties are operating in the French or Japanese markets. A handful of them (mostly the ones you list) have become trendy in other niches or provide opportunities for puns.

Two letter tlds are reserved for countries

AKA .com's are too expensive. Next.

AKA .com's are squatted. Next.

What's the difference? It's too expensive because of squatters.

Aren't some of the newer TLDN's pretty good alternatives, such as ".shop", ".store", ".social", ".insurance", ".save", ".software", ".search", ".game", ".travel" etc?

Manhattan's overpriced. People are moving to Brooklyn.

Brooklyn is overpriced, people are moving to Queens.

Queens‽ Nobody lives there anymore. It's too crowded.

My concern would be email deliverability for less popular TLDs, although I imagine it's not an issue for the more popular alternatives like .io and .ai

What % are actively operating but don't have any website... not counting a placeholder "coming soon" page?

I suspect the answer would be surprising.

> I suspect the answer would be surprising.

Seems like a paradox.

Plotting libraries should use larger fonts as default. The text in the plots in the post is so small, it's not readable.

They should just get a .chatgpt TLD because that's what 95% of them are.

yeah .com's aren't going anywhere. they're just expensive now

“Nobody goes there anymore. It’s too crowded.”

— https://quoteinvestigator.com/2014/08/29/too-crowded/

I have a .co but I've been to and love Colombia.

.com is very important

Especially for email deliverability

Using other TLDs is a mistake

Reply on news.ycombinator.com