How long it takes to build a custom website in 2026 depends on three things: how big the brief is, how fast the client makes decisions, and whether the content is ready when design starts. Most projects we ship at doubleSix Studio land between 3 and 14 weeks, calendar time, with the spread mapping cleanly to the three tiers we offer. Anything outside that range is usually a sign of either an undersized scope (microsite work that does not need 8 weeks) or an oversized one (a platform replacement masquerading as a "website"). This article walks through the realistic timeline at each tier, where the time actually goes, and the reasons projects slip past their dates.
One thing to set straight up front: agencies that promise "first-page Google ranking in 4 weeks" are different from agencies that promise "site live in 4 weeks." The second is reasonable. The first is not. Building takes weeks. Ranking takes months. They are different conversations.
What does a typical web build timeline actually look like?
Three timelines, mapped to the three tiers most studios price at.
Small site (3 to 4 weeks): A focused 5 to 10 page marketing site with custom design, basic CMS, and on-page SEO. We block 18 to 30 studio days across the four weeks, with the bulk in design and engineering. Right for a Series-A startup, a portfolio, a product microsite.
Mid size site (5 to 7 weeks): A full marketing site with 10 to 25 pages, headless CMS, blog, custom motion, light third-party integrations. We block 35 to 50 studio days. Right for a Series-B startup, a D2C brand, an agency, a hospitality group.
Large custom build (8 to 14 weeks): A platform-class build with custom application work (auth, dashboards, gated content), e-commerce, editorial systems, heavy motion, multi-region hosting. We block 80 to 180 studio days. Right for a SaaS marketing site that ships alongside the product, a fashion brand with a real catalogue, an editorial publication.
Above 14 weeks for a marketing site, scope is usually the issue. Long projects are almost always two projects (a rebrand plus a rebuild) sharing one timeline.
How long does each phase take?
Five phases on every project. The proportions stay similar across tiers, the absolute durations scale.
Phase 1: Discovery (5 to 10% of timeline)
One sharp call, one written brief, one signed contract. We surface goals, constraints, and the actual metric of success. For a small site, this is 3 to 5 working days. For a large custom build, it stretches to 7 to 10 days because there are usually more stakeholders to align.
Discovery is fast when the client has thought clearly about the brief before kickoff. It stretches when "we will figure out the goals as we go" is the answer to the success-metric question.
Phase 2: Design (35 to 45% of timeline)
The largest phase by time. Two structured rounds: an initial concept that proposes the visual direction, and a refined concept that resolves feedback. After concept sign-off, we extend the design system across the full page count.
For a small site: 6 to 9 working days for concepts plus 5 to 8 for the page extensions. For a mid size site: 12 to 18 working days for the same phases, scaled up. For a large custom build: 25 to 45 working days, including custom illustration or photography direction where applicable.
Phase 3: Build (35 to 45% of timeline)
The second-largest phase. Front-end engineering, CMS modeling, integrations, motion implementation. We typically build in vertical slices (one page or one feature complete from end to end, then the next) rather than horizontal layers. This catches problems earlier.
For a small site: 8 to 11 working days. For a mid size site: 15 to 22 working days. For a large custom build with custom application work: 35 to 80 working days, depending on how much app surface there is.
Phase 4: QA and content load (8 to 12% of timeline)
Cross-browser testing, mobile-device verification, accessibility audit, performance budget enforcement, content load into the CMS, copy QA, link checking. This phase always takes longer than people expect because real-content edge cases (long names, awkward image ratios, missing fields) only show up after content is loaded.
Two to four working days at Small site. Three to five for a mid size site. Six to ten for a large custom build.
Phase 5: Launch and handover (3 to 5% of timeline)
DNS, SSL, sitemap submission, analytics verification, training the editor team, transferring admin access, documenting handover. One to two working days at Small site, two to three for a mid size site, three to five for a large custom build.
Launch is the smallest phase but the most expensive to get wrong. A bad DNS cutover or a missing redirect can cost a client a week of organic traffic.
Phase breakdown at a glance
| Phase | Small site (3-4 weeks) | Mid size (5-7 weeks) | Large custom (8-14 weeks) |
|---|---|---|---|
| Discovery | 3 to 5 days | 5 to 7 days | 7 to 10 days |
| Design | 11 to 17 days | 22 to 32 days | 40 to 80 days |
| Build | 8 to 11 days | 15 to 22 days | 35 to 80 days |
| QA + content | 2 to 4 days | 3 to 5 days | 6 to 10 days |
| Launch + handover | 1 to 2 days | 2 to 3 days | 3 to 5 days |
| Total studio days | 18 to 30 | 35 to 50 | 80 to 180 |
| Calendar weeks | 3 to 4 | 5 to 7 | 8 to 14 |
What causes most projects to slip past their deadline?
Five reasons, in the order they show up.
1. Content is not ready when design starts. The biggest cause of slip on every web project, full stop. The client says content will be ready in week 1, but the marketing lead has not started writing it. By week 4, design is waiting for copy. By week 6, the site is "almost done but missing 12 paragraphs and 8 images." Plan for content to lag, not for it to be ready on time. According to industry surveys reported by sources like Smashing Magazine and Web Almanac data, content delays cause more web project slip than engineering issues by a factor of 3 to 5.
2. Decision-makers are not aligned. The CEO loves the concept, the CMO does not, the founder wants something else entirely. Each new opinion adds days of revision and de-alignment. Projects ship on time when one person owns the design decision and the rest of the team trusts that decision.
3. Scope creep that nobody priced. "While you are at it, can we also..." The pattern is innocent each time and devastating in aggregate. Five "small additions" eat 8 days of timeline. The fix is a written change request process: every addition gets quoted in writing inside 24 hours, the client decides yes or no before work starts.
4. Slow client review cycles. The studio sends concept on Monday. Client gets feedback together on Friday. That is 4 lost days per round, every round. On a Growth project with 4 review cycles, that is 16 lost days. Right pattern: 24 to 48 hour review turnaround, with a single point of contact consolidating internal feedback before sending it back.
5. Third-party dependencies the proposal underweighted. The CRM the site needs to integrate with is owned by a vendor who responds in 5 days. The legal team has not approved the privacy policy. The photography shoot is booked for week 8. Each of these adds calendar drag that engineering cannot accelerate.
Why content is usually the delay, not engineering
This deserves its own section because most clients refuse to believe it until they see it.
On a typical mid size project, engineering takes 15 to 22 working days. Content production for the same project (writing the homepage manifesto, the about-the-founder paragraph, the 8 service descriptions, the 4 case studies, the 6 FAQ entries) takes the client team roughly 30 to 50 hours of focused work spread across the 7-week timeline. Most clients do not block the time, so the work happens in 90-minute fragments, late, the night before review meetings.
The result: by week 6, design is done, engineering is 80% done, and the team is waiting on three case studies that the marketing lead promised to write four weeks ago. The site cannot launch until those are written. Engineering has nothing left to build, the studio is on the next project's design phase, and the client wonders why everything stopped.
The fix is dull but reliable: write the content during the discovery phase, before kickoff, with the studio reviewing draft drafts as design proceeds. Content that is locked at week 2 saves 7 to 10 days of slip at week 6.
The single biggest accelerator we have ever shipped on a project: a client who replied to every concept and review email inside one business day. That project ran 11% under timeline. Most projects run 8 to 15% over.
How do you keep a custom build on track?
Six habits that move projects ship-on-time. Adopt three of them and most slip risk evaporates.
- Lock content during discovery. Write everything before design starts. Studio reviews drafts as design lays out pages, and the two converge by mid-build.
- Designate one decision-maker. One person on the client side owns the call on every revision. Internal stakeholders feed into them, not into the studio.
- Block 90 minutes a week for review meetings. Same time every week, 9am Tuesday or whatever fits. Both sides protect it. Reviews never slide.
- Approve concepts inside 48 hours. Faster than that is great. Slower than that and the project starts losing days that compound.
- Use written change requests. Anything outside scope gets quoted in writing inside 24 hours. Client decides yes or no before work starts.
- Schedule the launch date in week 1. Put it on the calendar before design starts. The deadline is real, the launch happens, the team works backwards from it.
Can you ship a custom website in under 3 weeks?
Honest answer: usually not, with caveats.
A genuinely custom website (bespoke design, hand-written code, real CMS) takes a minimum of 3 weeks of calendar time at our studio, and we have not seen a credible custom build ship faster anywhere in the market. The lower bound is set by the design rounds (concepts, refinements, sign-off), not by engineering speed.
What ships under 3 weeks: Webflow microsites at 1 to 2 weeks, Squarespace sites at 5 to 10 days, single-page pre-built templated sites in days. These are all valid for the right brief. None of them is a custom website. The misuse of the word "custom" is how clients end up disappointed.
If your timeline truly is under 3 weeks and the brief is small, we recommend a Webflow build (we will scope it honestly) or a Squarespace setup with a designer to customise the theme. Both ship inside 2 weeks for a 5-page brief. Both leave room to upgrade to a custom build later when the budget and timeline allow.
What is the trade-off between speed and quality?
Real but smaller than people expect.
The 18-to-30 day Small site ships work that is 90% as polished as the 35-to-50 day Mid size site on equivalent scope. The remaining 10% is in motion design, accessibility refinement, and edge-case handling. For a marketing site whose only job is to land impressed visitors, the 10% is worth paying for. For a microsite running for 6 months, it is not.
Speed only damages quality past the point where engineering shortcuts have to be taken. We have not shipped that point in 2 years. If a studio is offering a custom 12-page site in 9 days, ask which corners get cut. The honest answer is performance, accessibility, and aftercare. Sometimes that is acceptable. Just go in knowing.
The shortest version
How long it takes to build a custom website is not a function of engineering speed. It is a function of decision speed, content readiness, and scope discipline. Plan for 3 to 4 weeks at Launch, 5 to 7 for a mid size site, and 8 to 14 for a large custom build. Block content production during discovery. Designate one decision-maker. Hold weekly reviews. Most projects that slip slip on those three. To talk through your timeline against your scope, our contact page is the fastest route, and our services page lists what we ship at each tier.