If you are asking should I build my own website or hire someone, you have already run the comparison in your head: a small monthly subscription against a few thousand dollars up front. That comparison is wrong. Not because the subscription price is fake, but because it is one line item out of four, and the other three decide it.
I build websites for a living, so weigh my bias accordingly. I am also handing you the math I would use to argue against myself.
Why “monthly fee versus build fee” is the wrong comparison
Because the monthly fee buys software and the build fee buys labor, and the labor happens either way. The only question is whether you pay in dollars or in hours.
The comparisons I read count one number. There are four.
| Number | What you write down | Where it comes from |
|---|---|---|
| 1. Your hours | Hourly rate times realistic build hours | What you charge, or would have earned |
| 2. The stack | Three-year subscription total, add-ons included | The vendor’s pricing page, today |
| 3. The delay | Annual customer value times customers per month times months to finish | What a new customer is worth to you over a year |
| 4. The exit | Cost to move off when you outgrow the platform | Migration quotes, plus what stays behind |
Number 1: your hourly rate times the hours it really takes
Plan on 40 to 60 hours for a first build, over two or three months of evenings. That is the number when it goes well.
Here is where it goes. Learning the editor well enough to stop fighting it. Writing copy for eight to ten pages, which nobody budgets and everybody underestimates. Sourcing photos. Connecting a domain. Testing a form. A pass on mobile. Then the revision loop, where you look at the homepage and hate it.
And here is what never makes the comparison: those hours do not stop costing you when the calendar looks empty. They come out of sleep, weekends, and the follow-up calls you skipped to fight a hero image.
If you are reading this between meetings and do not have forty evenings for stock photo licensing – that is your answer. You will not lose this on skill. You will lose it on calendar.
But AI can just build a website now, right?
Partly. And the honest version is more useful than either of the loud versions, especially if this is the first you are hearing of any of it. You are not behind.
What AI does genuinely well is a first draft. Pages, headings, placeholder copy, a layout, in an afternoon, and it will look like a website. What it hands you is a starting point, though, not a finished business asset, and the distance between those two is most of the work.
If you already tried it, you know the feeling. Fine in the preview, wrong on your phone. Spacing almost right, section to section, in a way you could feel but not name. Copy that described a business instead of your business, with no proof, no prices, no offer. A form that said thanks and never told you whether the email arrived. That is not you missing a setting. That is the gap between a draft and something that holds up in front of a paying customer, and it is not one prompt away. It is the job.
We use AI every day, for the record. It sits in our process the way a nail gun sits in a framer’s truck: removes the slow part, does not decide where the walls go.
Number 2: the three-year subscription total, with the add-ons counted
Look them up yourself, today. Prices change and I will not hand you a stale one.
What you write down is not the headline plan. It is the plan plus the commerce tier if you sell anything, the booking app, the form upgrade the moment you need conditional logic, the custom domain once any free year ends, email hosting. Multiply by 36.
Then be fair to the other side. Hiring is not a one-time number either: a WordPress site still needs hosting, and some need paid plugins. If you want someone watching it, our Care plan is $300 a month, no contract. Three-year total next to three-year total.
One line never moves to the cut column: maintenance. Every site that got wiped or quietly turned into somebody’s spam relay was running software that was fine last year. Skipping backups and updates does not save $300 a month, it borrows it against a rebuild. Cut a page. Cut the video. Not that.
Number 3: what months without a working lead path cost you
You do not need a study for this. You need one number you already know: what is a new customer worth to you over a year?
Multiply by the customers a working site would bring in per month, then by the months to finish. If a customer is worth $600 over a year and the site brings in two a month, four months of evenings cost you $4,800. Twice the price of our $2,400 Launch build. If your customers are worth five figures, it is not close.
And if your instinct says zero because you run on referrals, check that. A referral still looks you up before calling. The site is not generating the lead, it is confirming it.
Number 4: the exit cost, and what does not come with you
The content usually comes with you. The design almost never does.
Hosted builders export text and images another system can read. What does not survive is the layout, the templates, and often the URL paths, which means redirects, which means somebody maps old addresses to new ones so your links do not fall in a hole. Forms are a separate mess.
I know that one from the inside. We moved our own blog, 177 posts, off a hosted CMS onto WordPress. Everything made it, and it still took real work: URL mapping, redirects, re-hosting images. When we quote that job, site migrations run $700 to $1,800. So the cheap start does not remove the build cost. It defers it, staples a migration fee on top, and hands you the bill when you are busiest.
Is there ever a case for doing it yourself?
One. If nobody has paid you yet and you are still testing whether anyone will, throw up a rough page and go get ten conversations. That stops the day someone hands you money. From then on the site is what buyers judge you by before they speak to you, and every evening on it is taken from the business that just proved it works. Cross that line and tell us what you are building.
And if what you actually want is control, not savings, you can have both. You keep the domain, the code, the design, the content, and your own logins. That is how we hand sites over.
Four times going it alone reliably costs more
- Money moves through the site. Checkout, tax, inventory, deposits. The edge cases break quietly, in the direction of you not getting paid.
- Search is your main channel. If you compete for real local terms, structure matters more than looks, and builders make structure easy to get wrong.
- You have a hard date. Learning a new tool under a deadline is how people pay twice. We schedule around the date.
- You already stalled once. The second attempt costs what the first did, plus your morale.
The three jobs that go wrong most
Forms that do not deliver
The big one. Invisible, which is what makes it expensive. Submissions land in spam, or the notification goes to an address the platform cannot send from. You never find out. You just think the site is not working. Test it like a customer: fill out your own form from your phone on cell data, confirm it is not in junk, and store entries in a database, not only in email. We set that up on every build. “The form was broken for six weeks” is the most expensive sentence in this job.
Structure that buries your services
One long homepage with everything on it, no page per service, identical title tags sitewide, headings chosen because they were the right size instead of the right level. Looks fine. Also gives search engines one page to rank for every service you sell, so it competes for one term instead of five. If a service is worth money, it deserves its own page and title. That structure is the backbone of how we build a small business site.
Mobile
You will build it on a laptop. Most of your visitors will not be, and the editor’s mobile preview lies a little. Open it on a real phone: can you tap the number, is the text over the hero legible, how slow does it feel off wifi. We test on real devices.
You already started and stalled. Rescue or restart?
The deciding factor is almost never the code. It is whether your content is any good.
Copy and photos are the expensive half of a website. If you have written real page content and gathered decent images, you are further along than you feel. Content pages run $275 to $375 each when we write them, and our $2,400 Launch package includes ten pages populated.
Rescue gets expensive the other way around: a half-built site on a platform nobody would have chosen, no content, and somebody has to reverse-engineer it first. Paying a professional to learn your abandoned decisions is not a bargain. Call before the stall, not after.
So should I build my own website or hire someone?
Add numbers 1, 2, 3, and 4, then put the total next to a real quote.
I like that rule because it drags the two costs everybody hides from themselves, your hours and your delay, onto the same page as the ones on an invoice. Run it honestly and for anyone with paying customers, the four numbers land north of the build fee. Our pricing calculator gives you the hire side in a minute.
Common questions
Is it worth paying someone to build a website?
Once you have paying customers, yes, and the case strengthens every month. It is worth it when your time bills for more than the build costs, when money moves through the site, or when search is how customers find you.
How long does it take to build your own website?
Budget 40 to 60 hours for a first build, usually over two to three months of evenings. Writing the copy takes longer than assembling the pages. Almost always.
Can AI build my website for me?
AI will build you a draft, fast, and it will look like a website. What it will not do on its own is get the structure right for search, prove anything specific about your business, hold spacing consistent enough to read as professional, or verify that your form delivers. That stretch decides whether the site earns anything, and we do it by hand.
Is a hosted builder cheaper than hiring a developer?
Cheaper in software, yes. Cheaper in total, usually not, once you price your own 40 to 60 hours, the months unfinished, the add-on tiers, and the migration at the end.
Can I build it myself now and hire someone later?
People do, and it works, but know what the detour buys: a year or two of a site you apologize for, plus a migration fee to land where you were headed anyway. If you go that way, register the domain in your own name and keep your copy where you control it.
Look, run those four numbers on a napkin. Ten minutes, and I would rather you did that than take my word for it. Then run the calculator so the other side is a real figure instead of a rumor. When the two land close enough to argue about, send me what you are working on and I will tell you what it takes to finish it properly.