I get this question almost every week. Someone has an idea for a project, they know what it is called, and then they spend three days stuck on the last part: what should the domain name actually be?

It is a reasonable question to get stuck on, because the choice matters more than it looks. The name is what people type, what people remember, and what people say out loud when they tell a friend about your thing. Get it wrong and you will pay for it for years. Get it right and half your marketing does itself.

Since we register domains for a living, I thought I would write down what I actually see. Not theory. The patterns in the domains that people actually buy, the ones that work, and the ones that quietly cost more than they should.

Start with the extension, honestly

Here is the thing nobody tells you when you start out. The extension is not decoration. It is half the decision.

When I look at a registration, the extension tells me what the buyer thinks the thing is. .com means business, and it still means business. If you are building something you want to sell, or a company, or a service, .com is the default expectation and you will have to explain yourself otherwise. It is the shortest word in the conversation and it is understood everywhere.

.io is the developer favourite right now, and I understand why. It reads as technical, it is short, and it survives in a spoken sentence, which most two-letter endings do not. If you are building a tool, a product for people who work in tech, a SaaS of some kind, .io is a genuinely strong choice. I would still say it is a slightly more expensive habit than it needs to be, but it is not a bad habit.

.ai is the loud one. Every AI product that launched in the last two years seems to have taken it, and it now carries the association so strongly that using it for something unrelated to AI feels a bit like a costume. It is also expensive, and it is heavily regulated in a lot of countries, which matters if you are a normal company and not a research lab.

.se is the obvious one if your audience is in Sweden. It says where you are from, and for Swedish traffic that is worth a lot. The limitation is that it reads as local, and it makes a later move international feel a bit awkward. People do it anyway and it works fine.

And then there are the ones that are less about your audience and more about how you want to look. .agency, .studio, .media, .consulting, .design, .tech, .marketing. These are not bad. They are descriptive, they are readable, and they do a lot of the explaining for you before you have said a word.

The trade is always the same. A shorter, more general extension is more flexible and usually more expensive. A longer, descriptive extension is cheaper, clearer, and narrower.

The prices in the sidebar are real

I will not insult you by pretending the pricing here is a secret. The shop shows the current prices, and they change with promotions and with what the registry is charging that month. The rough shape of it, though, is worth knowing before you decide.

At the cheap end you have endings like .xyz, .site, .store, .shop, .space and .online. Some of these go for under a dollar a year in the first year. That is not a trick, the price really is that low, and for a side project or a test that is completely fine.

In the middle you find the descriptive endings. .agency at under four dollars, .media around ten, .pro in the high twenties, .group similar. Then you have the mid-priced national and industry endings, .se around twenty-nine dollars, .asia around fourteen, .group, .solutions.

And then the expensive ones, where the price is not about scarcity so much as it is about how short and how desirable the name is. .io at around sixty-five dollars a year. .design close to sixty. .tech around sixty-four. .ai at over a hundred dollars a year, which is the one that stops people, and rightly so.

Here is what I would tell a customer who asked me directly. If the name is the whole point, pay for the good one and eat the cost. A domain at sixty-five dollars a year is less than most businesses spend on coffee in a month, and you will keep it for a decade. A domain you regret at four dollars a year is still a regret, just a cheaper one.

What a good domain name actually looks like

After a few thousand registrations, some patterns are obvious. None of them are rules, and some of them I break myself, but they are a good starting point.

Short beats clever. Every time. A name that needs explaining at a party will be explained at a party and then forgotten. If you can say it once and the other person has it, you are done.

Say it out loud. This is the test I use. Say the name to someone who cannot see the screen. If they can write it down afterwards, it works. If they ask how it is spelled, you have found a problem. Every company with a hard name has stories about this and every company with a hard name is still paying for it.

No hyphens, no numbers. Hyphens read aloud as a pause. my-company is my dash company. Numbers are worse. When someone hears the number, they have to work out if it is a three or a five, and if you are on the phone or in a meeting you will get it wrong. It also makes the domain easy to confuse with a dozen others.

Match the search. This is the one people push back on, so let me put it plainly. If your business name is one thing and your domain says something else, you are making your customer do extra work for you. The domain should be the name, in the extension you can defend.

One word, or two that are obviously one thing. Compound names with two normal words usually work. Names built out of three or four syllables do not. Life is short and so is the attention span.

Check the social handles before you commit. Not because it matters legally. Because if you are going to want the same name on a social account later, you will want to know now rather than in six months when the name is taken. It changes which of two good names you pick.

What I would not do

Some advice, given plainly, because it costs me nothing and might save you a year.

Do not buy a name you are not sure you will use yet. I know the feeling. You have an idea, it might not work, but the name is available today and maybe tomorrow it is not. If you are not going to use it in the next month, do not buy it. Most people end up holding names for domains they never built anything on, and then they renew it out of guilt for another three years.

Do not buy twelve domains "just in case". One good domain and maybe one defensive variation is plenty. Twelve domains is a hundred dollars a year you will never look at, and a confusing signal to anyone who finds you.

Do not pick a name you have to explain. Including to yourself. If you cannot say the name in one breath without a story, it is too much.

Do not let price be the only reason. If a name is available at a dollar, ask why. Sometimes the answer is that nobody wanted it. A cheap domain that nobody wanted is not a bargain, it is a warning.

A few endings I get asked about directly

Is .io worth it? For a tech product, often yes. For a local business or a content site, almost never. It is a slightly awkward fit, and you will pay for the privilege of sounding like everyone else in your category.

Is .ai worth it? Only if you are actually in AI. If you are not, it reads as a trend and it is a hundred dollars a year for that read. There are plenty of cheaper ways to be interesting.

Should I get .com even if I have .se? If your business is in Sweden and stays in Sweden, .se is enough and it is a good domain. If you have any intention of working internationally, get the .com as well and point it at the same place. Two domains, one site, no problem. And if you ever do want to sell the company, the .com is part of the price.

What about the new generic endings? They are genuinely fine now. .app, .dev, .site and the rest have been around long enough that people recognise them, and .app gets free HTTPS. Do not treat them as second class, and do not treat them as a sign of anything either. They are just endings.

One practical thing before you buy

When you find a name you like, look at it once more with completely fresh eyes. Then look at the price. If the price makes you wince, it is usually because the name is too good for the extension, and there is a simple answer: take the second best name for two hundred dollars less and build the thing.

The name is not the business. It is the handle you put on the business. A good name helps a lot. A clear idea, a working site, and something people actually want will do more than any extension on earth.

Go and find the name. The shop is right there.

Sources