Register the domain first, then set up email on it, then buy hosting, then build the site. That order works because each step depends on the one before it: email needs the domain, the site needs somewhere to live, and a domain that turns out to be taken invalidates everything already built on the assumed name.
Most of the advice on how to get your business online treats these as a shopping list. They are not a list. They are a chain, and doing them out of order is how people end up rebuilding a finished website because the name on it was never available, or discovering three weeks after launch that the contact form has been quietly failing.
The Chain, and Where It Snaps
A dependency chain has one useful property: you can work out the correct order without knowing anything about the tools. Email cannot exist before the domain does, because the address is part of the domain. The site cannot go live before hosting exists, because there is nowhere to put the files. Hosting is best chosen once you know roughly what the site runs on, because a static brochure page and a store that has to keep a database, a checkout and a payment gateway healthy are asking for different things from a server.
The chain snaps in a predictable place. People start with the part that feels like progress, which is always the website, because the website is the part you can look at.
Step 1: Register the Domain, in Your Own Name
Check availability before anything else. Not after the logo, not after the first design draft, and certainly not after somebody has spent a weekend writing page copy that names the business a particular way.
The reason is not superstition. A name that is taken in .com is often taken in a way that matters commercially even if you take the .net, because customers who half-remember your name will type .com and land somewhere else. Decide that trade-off at the start, while changing your mind is free.
Register it yourself, with your own credit card, under your own email address. This is the strongest opinion in this post. Recovering a domain from a former web designer, a former marketing agency, or a cousin who set it up in 2019 and has since changed email providers is the worst recoverable mistake on this list. The domain is the one asset in the whole stack that cannot be rebuilt from scratch: hosting can be replaced in an afternoon, a website can be rebuilt, and a mailbox can be recreated, but a domain sitting in somebody else's account is a negotiation.
It gets slower, too. Under ICANN's transfer policy, changing the registrant on a domain can put a 60-day lock on transferring it away, so the fix for "my contractor owns my domain" is not always same-week even when the contractor cooperates. Buy the name yourself and hand out access. Never the reverse.
The classic failure here is expensive and entirely avoidable: the site is finished, the cards are at the printer, and then somebody finally looks the name up and finds it parked by a squatter since 2011.
So check it this way. Search the exact name you plan to trade under, register it, then log into the domain registration account yourself and confirm the registrant contact shows your address rather than an agency's. If you cannot log in without asking somebody for the password, you do not own the domain yet.
Step 2: Get Email Working Before You Need It
Email is second because it is the piece customers actually use first, and because it is the piece most likely to be broken in a way nobody notices.
An address on your own domain is the cheapest credibility a small business can buy. A quote sent from a free webmail address with the business name in it reads as a side project, fairly or not. And when your domain sends mail, the internet judges it on records you have to publish yourself: MX records to say where mail is delivered, and an SPF record to say which servers may send on your behalf.
Set those up while the domain is fresh and nothing depends on it. Mail routing is the single most disruptive thing to change on a live business, which is why moving it later deserves its own plan. That problem is covered in detail in what to check before you move your website to a new host, and the short version is that you do not want to be doing it under pressure.
Most hosting accounts include mailboxes on your domain, and on the plans here they are administered from Hepsia's email manager alongside the rest of the account. Start there. Mailboxes that come with the hosting are one login and one support contact instead of two, and if you later outgrow them, moving mail to a dedicated provider is a matter of repointing MX records rather than starting over.
What goes wrong is quiet. The contact form on the finished site sends to an address that bounces, or the message lands in spam because the sending server was never authorized by an SPF record. Nobody complains about a form that eats submissions. They just buy from somebody else.
Testing it takes four minutes. From a phone, on mobile data rather than the office network, send a message to the new business address and confirm it arrives. Then reply from that mailbox to a personal Gmail or Outlook account and check that the reply stays out of the junk folder. At a terminal, dig MX yourdomain.com +short and dig TXT yourdomain.com +short show you what the rest of the world sees.
Step 3: Buy Hosting Once You Know What the Site Runs
Hosting comes third because the answer depends on the site. A five-page brochure and a store taking card payments have almost nothing in common except the word "website."
For a first small business site, shared hosting is normally the right call, and the reason is not price. It is that shared hosting is administered for you. Operating system patching and web server configuration are somebody else's job, which is exactly what you want them to be during your first month of trading. Shared, semi-dedicated, VPS and dedicated plans all exist at BestPaidHosting, and entry-level shared plans start at $3.95 per month, but the distinction that matters at launch is managed against unmanaged. You want managed. If you are still weighing a paid plan against a free one at all, what paying for hosting actually buys you is the question underneath this one.
Two things are worth checking before you pay for a year. First, that the plan runs the software your site needs, which for most small business sites means a current PHP version and at least one MySQL database. Second, what happens when you grow: whether an upgrade is a plan change on the same account or a full migration to a different provider.
The trap at this step is a long term bought too early: three years paid upfront on the cheapest plan available, followed a month later by the discovery that the store platform you settled on needs something the plan does not offer.
Prove the account works before you trust it. Create a database, then upload a one-line file that prints your PHP version and load it in a browser. If a shared hosting account cannot serve that, nothing you build on top of it will work either.
Step 4: Build the Smallest Site That Can Take an Order
Now build. And build less than you think.
A first version needs a home page that says what you do and where you do it, a page for services or products, a contact page with a real address and phone number, and something that proves you are a going concern, whether that is your prices or photographs of finished work. That is four pages. It can go live in a weekend.
The instinct to launch a "complete" site is the single biggest cause of sites that never launch at all. Six months of drafting an About page is six months of a competitor being the one that shows up in local search results. A site that exists collects search impressions, gets crawled, accumulates a little history, and gives you something concrete to improve. A perfect site in a folder on somebody's laptop does none of that.
The embarrassing failure at this stage is placeholder copy left in the template, or a contact page still carrying a phone number from a previous venture.
Open the site on a phone. Then tap your own contact link and actually call the number.
What Can Wait
These feel like launch tasks. None of them blocks anything.
Analytics can be added any time, and installing it two weeks after launch costs you two weeks of data about a site nobody has found yet. A logo refresh is a branding project wearing a launch project's clothes; the site works with the logo you have. A mailing list is worth having once there is traffic to convert, and setting up automation before there is an audience is effort spent on an empty room. Page speed work belongs after launch too, because you cannot sensibly optimize a site whose content is still changing, though it is worth understanding which speed metric to fix first when you get there.
How to Get Your Business Online: The Weekend Checklist
| Step | What it blocks | The check that catches the mistake |
|---|---|---|
| Register the domain yourself | Everything. Email, hosting, branding, print | You can log into the registrar account without asking anybody |
| Publish MX and SPF records | Contact forms, quotes, invoices, customer replies | A test message arrives, and a reply from it does not land in junk |
| Choose and set up hosting | The site, staging, databases | A test page loads and a database can be created |
| Build four pages and publish | Nothing. This is the last dependency | The site opens on a phone and the phone number connects |
Work down that table in order and a small business can be genuinely online, with mail that arrives and a site that takes inquiries, inside a weekend. Sequence beats effort here. Every hour of rework on this list comes from doing step four before step one.
If you only take one thing from this: go and check, today, whose name is on your domain. Everything else on the list can be redone at the cost of an afternoon.