Key takeaways
- There's a difference between builder hours (the actual work) and calendar weeks (how long you wait) — and almost every timeline quote only counts the first one.
- A typical small-business site takes 3–6 weeks with an agency, an online store 4–10 weeks, and a custom web app 2–6+ months. The exact number depends more on your route (DIY, freelancer, agency) than on the technology.
- In our experience across 150+ builds, the two biggest timeline killers are content collection and slow feedback loops — both are on the client's side, not the developer's.
- AI builders can generate a prototype in an afternoon, but a launch-ready site still takes weeks. "Generated" is not "launched."
- You can genuinely compress a timeline by 30–40% with a 7-day prep checklist before you hire anyone.
Ask ten people "how long does it take to build a website?" and you'll get ten answers: a weekend, two weeks, three months. They're all technically correct, and most of them are useless to you. Here's why: almost every answer quotes builder hours — the time someone spends typing — while you're living in calendar weeks — the time between "let's do this" and "it's live."
We've shipped 150+ websites since 2022, and the gap between those two clocks is where every surprise lives. So let's give you the honest answer: what things actually take, where the time really goes, what the AI-builder hype changes (and doesn't), and how to buy back weeks before you hire anyone.
The honest timelines: what to actually expect
First, the numbers. These are real ranges from real projects — ours and the industry's — not marketing. The columns are your route, because who builds it changes the timeline more than what they build it with.
| Site type | DIY (you build it) | Freelancer | Agency |
|---|---|---|---|
| Landing page / one-pager | 1–3 days (plus learning curve) | 1–2 weeks | 1–2 weeks |
| Small business site (5–8 pages) | 1–4 weeks | 2–4 weeks | 3–6 weeks |
| Online store (Shopify/WooCommerce) | 2–6 weeks | 4–8 weeks | 4–10 weeks |
| Custom web app / platform | Not realistic | 2–4 months | 2–6+ months |
Notice something? The DIY column looks fast until you add the invisible row: learning the platform, troubleshooting at midnight, redoing the design three times because you hate it on mobile. WordPress.com's own guidance says a first-timer building a five-page site can spend anywhere from a few hours to several weeks — the range is that wide because it measures your learning, not the tool.
For the agency column, there's industry data backing the range: GoodFirms' 2026 survey of 300+ web development firms across 31 countries found half of them pick 4–8 weeks for a midsized website, and another third pick 8–16 weeks. Those are self-reported estimates from the firms themselves, not observed delivery times — but they line up with what we see on our side of the table.
The code is the fastest part of building a website. Decisions are the slowest. Every timeline overrun we've ever seen was a decision problem wearing a technical costume.
Builder hours vs. calendar weeks: the math nobody shows you
Here's a number that surprises every client: a small-business site that takes six calendar weeks to deliver usually contains only 60–120 hours of actual work. That's 7–15 working days. The other four weeks aren't laziness. They're the five phases below — and look at which column each one stalls in.
| Phase | Actual work | Calendar time | Where it stalls |
|---|---|---|---|
| Discovery & scope lock | 4–8 hours | 3–5 days | Waiting on your decisions |
| Design | 15–30 hours | 1–2 weeks | Revision rounds + feedback latency |
| Development | 30–80 hours | 2–4 weeks | The actual building (rarely the bottleneck) |
| Content entry | 5–10 hours | 1–3 weeks | Client-side bottleneck — see below |
| QA & revisions | 8–16 hours | 1–2 weeks | Device testing + fix cycles |
| Launch | 4–8 hours | 3–5 days | DNS, SSL, analytics, SEO basics |
Read that table again. Development — the part you're paying for — is the one phase that almost never stalls. The stall happens everywhere else. This is the single most important thing to understand before you sign anything: you control more of the timeline than your developer does.
The 5 things that actually eat your timeline
In 150+ builds, we've seen every delay pattern. Five of them account for nearly all overruns. Four are on the buyer's side.
1. Content collection (the #1 killer)
This is the one nobody budgets for. The developer needs your final text, your photos, your logo files, your product descriptions, your team bios — and you don't have them ready. So the build sits at 90% for three weeks while everyone waits on a paragraph about your company history.
The fix: before the project starts, put everything into three buckets: words (every page's final text, in a doc, approved by whoever approves things), images (photos, logo, brand assets, high-res), and access (domain registrar, hosting, social accounts, payment processor logins). A project with these three buckets ready on day one starts 2–3 weeks ahead of one without them. No exceptions.
2. Feedback latency
The designer sends mockups Monday. You look at them Friday. Your partner looks at them the following Wednesday. Comments arrive as "make it pop." Each round-trip costs a week, and most projects need three rounds. Do the math: feedback latency alone can triple the design phase.
The fix: agree on a 48-hour feedback rule before signing, and name one decision-maker. Two approvers with equal votes is a recipe for a three-month project. We'll say more about this in the checklist.
3. Scope creep (the polite kind)
Nobody storms in demanding a redesign mid-project. Scope creep arrives politely: "while you're in there, can we add a small booking thing?" "Just one more page." "What if the homepage had a video background?" Each addition is small. Ten of them add a month.
The fix: write the scope down and freeze it. Keep a visible Phase 2 list — everything that isn't launch-critical goes there, in writing, with the promise it ships after launch. Good developers love this; it protects their schedule too. If you're comparing quotes, our pricing page shows how we scope projects in fixed tiers instead of open-ended hourly fog.
4. Third-party integrations
Connecting your site to your CRM, booking system, payment processor, or inventory tool is never "just an API." Each integration means accounts to create, permissions to grant, webhooks to test, and — the classic — discovering that the tool you bought doesn't actually do what its sales page said. Budget 2–10 days per integration, not hours.
5. Launch polish (the invisible 10%)
A site isn't done when it looks done. SSL certificates, DNS propagation, 301 redirects from the old site, Google Search Console, analytics, speed checks, mobile testing on real devices, form deliverability testing — this "invisible 10%" is what separates a site that launches from a site that launches and works. Anyone quoting you a timeline without a launch phase is quoting you a half-finished project.
Rule of thumb from 150+ launches: take the builder's estimate and add 50% — not because developers are slow, but because your calendar has weekends, holidays, indecision, and a day job. The ones who beat their timeline are the ones who prepared like it was their project, not the developer's.
The 2026 wildcard: AI builders changed the question
Here's what's new since every other timeline article was written: tools like Lovable, v0, and Bolt can generate a working prototype in an afternoon. Clients now walk in saying "but the AI built it in ten minutes — why do you need six weeks?"
Fair question. Honest answer: "generated" and "launched" are different planets.
What the AI gives you in an afternoon: a good-looking prototype with demo data, no real authentication, no payment processing, no security review, no SEO, no analytics, no accessibility checks, no mobile QA, and copy full of placeholder confidence. Turning that prototype into something you'd stake your business on — real user accounts, real money moving, real customer data, Google actually finding it — is still 2–6 weeks of professional work, depending on what it does.
We've written a full breakdown of exactly what AI builders skip on the road to production — the security gaps, the data problems, the "works in the demo, breaks with ten users" issues. And if you're already holding an AI-built prototype that's stuck, that's literally what our rescue service exists for: we take AI-generated apps from "impressive demo" to "production system you can sell from."
The honest version of the AI timeline claim: AI compresses the first 20% of a build (scaffolding, boilerplate, first draft of the UI). It does almost nothing for the other 80% — decisions, content, integrations, testing, launch. Anyone selling you "a full website in a day" is selling you the first 20%.
Timeline red flags: promises that should scare you
Not every fast timeline is a lie — but these ones usually are:
- "Fully custom website in 3 days." Possible only if "custom" means "a template with your logo." Real custom design needs at least one revision cycle, and revision cycles need your feedback, which needs calendar days.
- No staging site mentioned. If they never show you a private preview link to review before launch, you're not getting a review phase — you're getting whatever they built.
- "We'll figure out content later." Translation: the project will stall at 90% for a month while you write copy under pressure. Content-first or content-parallel, never content-later.
- 100% payment upfront, no milestones. Milestones exist to keep both sides honest — and to keep the timeline honest. A developer with nothing due at the design milestone has no incentive to hit it.
- They can't tell you who's doing the work. "Our team" with no names means your project gets whoever is free. Ask for the actual designer and developer, and ask what else they're working on this month.
Flip side: a developer who gives you a longer timeline than competitors isn't necessarily slower. Sometimes they're the only one who budgeted for your feedback rounds. We've seen the "2-week" quote turn into a 3-month project more times than we can count — the honest 6-week quote would have been faster.
Your 7-day compression checklist (steal this)
Do these seven things before you hire anyone, and you'll genuinely cut 30–40% off a typical timeline. This is the highest-leverage week of your entire project.
- Write the sitemap on one page. List every page and the one job that page does ("Homepage → get the call booked"). If a page doesn't have a job, delete it. Fewer pages = fewer weeks.
- Fill the three content buckets. Words (final, approved text for every page), images (high-res, organized), access (every login the developer will need). Day one readiness beats day-one kickoff calls.
- Name one decision-maker. One person with final say on design. Everyone else advises. Write it down and tell the developer who it is.
- Freeze the scope in writing. What's in, what's explicitly Phase 2. Sign it like a contract, because it is one.
- Commit to the 48-hour rule. You'll review and respond to every deliverable within two business days. Put it in the agreement — it protects you too, because it forces the developer to keep deliverables coming.
- Limit revisions to two rounds per phase. Two rounds is enough for any competent designer. Unlimited revisions is how 4-week projects become 4-month projects.
- Define "done" before starting. Write the launch criteria: pages live, forms tested, analytics installed, speed target met, you trained on updating content. A project without a definition of done never ends — it just fades out.
Do this and something interesting happens: developers will want your project. An organized client with ready content and fast decisions is every good developer's favorite kind of work — and favorite clients get priority scheduling. We've bumped prepared clients ahead in our own queue more than once. Check our services page to see what a scoped, milestone-driven build looks like from our side.
The bottom line
So, how long does it take to build a website? The code takes days to weeks. The project takes weeks to months — and the difference is decisions, content, feedback, and launch polish. The timeline is mostly yours to control, which is good news: it means the fastest way to launch sooner isn't finding a faster developer. It's being a faster client.
Want to know where your current site stands before you rebuild or launch? Run it through our free site audit — you'll get a brutally honest teardown of what's actually slowing it down, in about 60 seconds. Or if you're ready to talk timelines for real, book a call and we'll map your project week by week, no guesswork.
