How Long Does It Take to Build a Website? A Realistic Timeline

Every founder asks "how long will this take?" and most agencies answer with a range so wide it's useless. Here are the timelines we actually plan to, with the caveats that matter, and the things that quietly blow timelines up.

Marketing website: 3–6 weeks

A custom-designed 5–10 page marketing site with a CMS, basic SEO, and a handful of forms. The standard small-agency project.

• Week 1: kickoff, discovery, sitemap, content audit. We figure out what the site is, who it's for, and what it must do.
• Week 2–3: design. Concept directions, then revisions, then component design, then page-level layouts.
• Week 3–4: development. Front-end build, CMS wiring, animations, performance tuning.
• Week 5: content load, QA, accessibility checks, cross-browser/cross-device testing.
• Week 6: launch and post-launch tuning.

Six weeks is the comfortable pace. Three weeks is achievable when the brief is tight, content is ready on day one, and the client is responsive within 24 hours. Eight to ten weeks means stakeholder review cycles ate the schedule.

E-commerce: 6–12 weeks

Shopify or similar, custom theme, ~50–200 SKUs, payments, shipping rules, basic loyalty/discount logic.

• Week 1–2: discovery, IA, product taxonomy, checkout flow design.
• Week 3–5: design and prototyping. PDPs, PLPs, cart, checkout — these are the make-or-break pages.
• Week 5–8: theme development, app installs, payment + shipping configuration.
• Week 8–10: product import, content load, QA across real-world flows.
• Week 10–12: soft launch, performance tuning, conversion optimisation, full launch.

The honest truth: e-commerce timelines depend on how messy the product data is. A clean CSV with 100 products takes a day to import. A spreadsheet inherited from a 2018 store with inconsistent variant logic takes two weeks.

Custom web app or SaaS MVP: 8–20 weeks

Auth, dashboard, billing, email notifications, ~6 core features. This is where ranges legitimately widen because feature scope dominates.

• Week 1–2: product discovery — flows, edge cases, data model. The work that decides whether the rest goes smoothly.
• Week 3–5: design — UI for every screen, every state (loading, empty, error). Reviews matter.
• Week 5–8: back-end: API, auth, database schema, integrations.
• Week 6–12: front-end build, in parallel with back-end as APIs land.
• Week 12–16: integration work, billing, email, third-party services.
• Week 14–18: beta with real users, bug bash, performance.
• Week 16–20: public launch and stabilisation.

The MVPs that ship in 8 weeks have small, focused scope and a single decision-maker on the client side. The MVPs that take 20 weeks are usually 12-week MVPs that grew during the build.

What slows projects down (the honest list)

Slow decisions, not slow code

Across every build we run, the slowest part of any week is waiting for a client decision. "Which direction do we want?", "Can you approve this copy?", "Do we ship Tuesday or push to Friday?" — every 24 hours a question sits unanswered is 24 hours added to the schedule.

Content that isn't ready

"We'll send the copy next week" is the four most expensive words in agency work. We've watched 4-week projects become 10-week projects on this alone. Have content ready on day one, even if it's draft quality. Polishing copy on a real layout is fast; designing around imaginary copy is not.

Stakeholder math

Every additional approver multiplies cycle time. One decision-maker = 24-hour turnarounds. Three approvers = 1-week turnarounds. Six approvers = nobody approves anything because nobody wants to go first. Choose one DRI on the client side and let them be wrong sometimes.

Late-stage scope creep

Adding "just one more feature" in week 6 of an 8-week project doesn't just add the feature's time — it adds context-switching cost, re-testing of touched areas, and re-design ripple. A good agency will quote a change order. A bad one will absorb it and miss the deadline.

Brand and aesthetic disagreements

Ranking colour palettes for a week is a luxury you can afford pre-build. Doing it mid-build wrecks schedules. Lock the look-and-feel before page design begins.

How to ship faster (without compromising quality)

1. Pick one decision-maker on your side. Internal politics aren't the agency's problem to solve.
2. Reply within 24 hours during the build. Async over Slack/Linear/email. Don't wait for the weekly call.
3. Have draft content on day one. Real words, even rough.
4. Lock scope at week 0 and resist additions. Park them in a "phase 2" doc.
5. Don't ask for a fourth round of design revisions. Two is normal, three is the buffer, four means you don't know what you want.
6. Use real branding, not a moodboard. If you're rebranding, do it before the website project, not during.

How to slip a project on purpose (when you should)

Sometimes a slip is the right call:

• Content isn't ready. Better to launch in 8 weeks with great copy than 6 with placeholder text you'll regret for a year.
• Brand isn't settled. Building a website on a wobbling identity is a guaranteed re-do.
• Scope was wrong from day one. If discovery uncovers complexity nobody saw, re-scope and reset the timeline. Pretending it's fine just delays the bad news.

The fastest project we ever shipped launched in 4 weeks. The slowest we'd planned for 6 took 18. The difference wasn't the team — it was the client's ability to hold scope.

Realistic timelines, summarised

Frequently asked questions