The Honest Version, Up Front

Yes, you can build your own website in 2026, and for the first time ever a DIY site can genuinely compete with professional builds. AI changed that. This guide is the straight version of how to do it: the right order of steps, the three routes worth taking, and the traps that quietly kill most DIY sites.

Two things before we start:

We build websites for a living, so you should know our bias. We'll flag the moments where paying makes more sense than DIYing, and we genuinely encourage you to try yourself first. You'll either end up with a good website or a much clearer idea of what you're paying for. Both are wins.
The most important sentence in this guide: a website's job is to get found and convert. Function over form. Most DIY builders spend 90% of their time on how it looks and 10% on whether anyone will ever see it. Flip that ratio and you're ahead of half the professionally built sites in the country.

First: Pick Your Route (Don't Pick a Platform Yet)

Most guides start with "choose Wix or WordPress." Wrong first question. The right one: what does this website need to do?

| Your situation | Your route | |---|---| | Selling products online | Shopify. Don't overthink it. Best checkout, best apps, built for stores. | | Business site that needs to bring in customers | AI tools (the Claude route). Steeper start, dramatically better result, and you own everything. | | Simple brochure, all work comes from referrals | A builder (Squarespace or Wix). Quickest to something presentable. |

For stores, our Shopify guide covers the ranking side. For the builder route, we've compared them all honestly. The rest of this guide focuses mostly on the AI route, because it's the one that's new, the one nobody else is explaining properly, and frankly the one we'd take in your shoes. It's how we built our own website, and that site now brings us clients every week.

Step 1: Keyword Research Before Anything Else

This is the step almost every DIY builder skips, and it's why their sites get zero visitors. Before you think about design, pages, or platforms, find out what your customers actually type into Google.

Do this today, it's free:

Open Google and start typing what you sell ("plumber hami...") and note the autocomplete suggestions. That's real search demand.
Search your main service and scroll to "People also ask" and "Related searches." Write them all down.
If you want volumes, Google Keyword Planner is free with a Google Ads account.

You're building two lists: money searches (someone ready to buy: "emergency plumber hamilton", "wedding photographer waikato prices") and question searches (someone researching: "how much does a bathroom reno cost nz"). Your site needs pages for both.

Step 2: Plan Pages Around Searches, Not Around Yourself

The classic DIY site is five pages: Home, About, Services, Gallery, Contact. It's a brochure, and Google treats it like one, because a single "Services" page can't rank for ten different services.

Instead, sketch a page list where every service gets its own page and every area you serve gets its own page. A Cambridge electrician who does rewiring, switchboards, EV chargers, and heat pump installs across three towns shouldn't have 1 services page; they should plan for 12+ pages, each answering one specific search. Add a page for the big cost and how-to questions from Step 1 and you have a real site plan.

This single decision, more pages, each more specific, is most of the difference between a site that ranks and a site that doesn't. It's also where the AI route destroys the builder route: dragging together 20 pages in Wix is a slog you'll abandon. Asking Claude to build them from your notes takes an afternoon.

Step 3: Buy Your Domain, and Own It Yourself

The non-negotiable: register the domain in your own name, with your own account, on your own credit card. Not your web guy's account. Not bundled inside a builder subscription. Businesses lose their domains every year because a freelancer disappeared or an agency held it hostage, and the domain is the one part of your website you can never rebuild.

The practical bits: a .co.nz from any reputable NZ registrar costs about $20-30 a year. Pick a name that's your business name, keep it short, skip the hyphens. Done. Don't spend a week on this.

Step 4: Build It (the AI Route)

Here's the route that didn't exist three years ago. Tools like Claude can now build you a genuinely custom website: you describe your business, feed it your page plan from Step 2, and iterate in plain English. No code knowledge required to start, and the output is a real, fast, modern website rather than a template.

We wrote a full walkthrough of building a site this way, so we won't repeat it here. The honest summary:

What's genuinely easy: producing pages, writing content drafts, getting something that looks professional, making changes ("make the header green", "add a testimonials section").
What's harder than it looks: the plumbing. Connecting your domain, setting up hosting, forms that actually email you, analytics. None of it is difficult for someone who's done it before, and all of it is a wall of jargon the first time.
The good news on hosting: modern hosting (Vercel, Netlify and friends) is nearly free for a small business site, fast, and secure. If someone quotes you $30-50/month for "website hosting", that's a legacy business model, not a necessity.

A middle path we offer, because this exact wall is where most DIYers stall: build the site yourself, and get us to handle the domain, hosting, forms, and launch plumbing. You keep the DIY savings and the ownership; we make sure the thing actually goes live and the enquiries actually arrive. Get in touch if that's the flavour of help you want, it's a small job for us and a big unblock for you.

Step 5: Build for Function Over Form

Tattoo this on the project: traffic first, then conversion of that traffic. A beautiful site nobody visits is a $0 asset. An average-looking site that ranks and converts is a salesperson that never sleeps.

Concretely, function over form means:

Every page has one job and one obvious next step. Call, book, quote, buy. If a visitor has to hunt for how to say yes, they won't.
Make it easy to say yes. Reduce the ask: "Get a free quote" beats "Contact us". A two-field form beats a ten-field form. Click-to-call on mobile beats a phone number in a footer.
Give something to the not-ready-yet visitor. Most visitors aren't ready to buy today. A lead magnet, a free checklist, a pricing guide, a "what it costs" download, turns those visitors into an email list instead of a bounce statistic.
Answer the money questions on the page. Price ranges, timeframes, areas served. Hiding prices doesn't create enquiries; it creates back-button clicks.

Design still matters, but it matters in service of trust: clean, fast, real photos over stock where possible, reviews visible. That's it. Every hour spent past that point on pixel-nudging is an hour not spent on the pages that bring traffic.

Step 6: Test Like a Customer, Not Like a Builder

You built the site on your laptop. Your customers will visit it on a cracked iPhone in a ute. You're not building for one screen.

Before launch:

Open every page on your actual phone, plus borrow one other phone and a tablet. Check nothing overlaps, buttons are thumb-reachable, and text is readable without zooming.
Submit your own forms. Did the email arrive? Check spam. A form that silently fails is the most expensive bug a small business site can have, and it's alarmingly common.
Click every link. Call the click-to-call number.
Run the site through Google's PageSpeed Insights (free). If it scores badly on mobile, fix that before adding anything else.

This step is also where DIYers discover there's a bit more to making a site function how you want than the builder ads suggest. That's not a reason to quit. It's a reason to test properly.

Step 7: Launch, Then Get Found

Going live is not the finish line; Google doesn't know you exist yet.

Set up Google Search Console (free) and submit your sitemap. This is how you ask Google to index your pages and how you'll see what searches you appear for.
Set up Google Analytics (free) so you know what's working.
Claim your Google Business Profile and fill it completely: photos, services, hours, and start collecting reviews. For local businesses this is as important as the website itself.
Add your site everywhere: email signature, social profiles, invoices, van signage.

Then keep publishing. One genuinely useful page or post a month, answering a real question from Step 1's list, compounds. Sites that launch and freeze fade; sites that keep answering questions climb.

Where the Honest Line Is

Here's our straight answer on DIY versus paying, and it's the same logic as our advice on hiring marketing help:

DIY is the right call when the website is a supporting act: your work comes from referrals, you have more time than money, or you're testing an idea. You'll spend a weekend or three, you'll learn heaps, and you'll own the result.

Pay for the build the moment leads depend on it. If the website's job is to generate customers, the game changes: keyword-mapped page architecture at scale, conversion structure, technical SEO, schema, redirects, speed engineering. That's the difference between the site existing and the site producing. Our builds start at $2,000, which is one or two customers for most businesses, and when we rebuilt Airmax's site, the phone went from 2 calls a day to 10 within two months. That's the ROI mathematics that makes "should I pay for a website" a short conversation.

And genuinely: try it yourself first if you're curious. Worst case, you arrive at the quote conversation knowing exactly what you're buying, which makes you a better client for anyone you hire. We'd rather work with someone who's been in the trenches than someone buying a mystery box.

Frequently Asked Questions

How much does it cost to build a website yourself in NZ? Domain ~$25/year, modern hosting free to ~$10/month, and a builder subscription ($300-700/year) only if you take the builder route. The AI route can be nearly free in cash and 40-80 hours in time. Your time is the real cost, which is why the DIY-versus-pay maths depends on what your hours are worth.

Can I build a website for free? Sort of. Free builder tiers exist, but you get the platform's ads on your site and a domain you don't own, which reads as "not a real business yet" to customers. If it's a hobby, free is fine. If it's a business, spend the $25 on a domain at minimum.

Do I need to know how to code? No. Builders never needed it, and the AI route now doesn't either: you describe what you want in plain English and iterate. Where technical knowledge still helps is the plumbing (domains, DNS, hosting, forms), which is exactly the part you can pay someone a small amount to handle once.

How long does it take to build a website? A builder brochure site: a weekend. A proper multi-page site via the AI route: a few weekends, mostly spent on content. The trap to avoid is the opposite problem: months of tweaking without launching. Ship at 80% and improve live.

Should I use WordPress? We'd say no in 2026. It's the ageing middle option: harder than a builder, weaker than the AI route, and it becomes a part-time maintenance job (plugins, updates, security). Most of the "basic site that never ranked" projects we replace are WordPress. Full reasoning in our builder comparison.

What about ecommerce? Shopify, full stop. Build the store yourself with their tools, then invest your effort in the content that makes stores rank: collection pages, buying guides, product depth. We covered exactly that in our Shopify guide.

How to Build a Website in NZ (2026 Step-by-Step Guide)