How Long Does It Take to Build a Website in 2026?

You asked a simple question and got a shrug: 'Four to twelve weeks. It depends.' That range is not a lie, but it is a number about the agency's calendar, not yours. The date your site actually goes live is decided somewhere else entirely — on your side of the table.

This post gives you the week-by-week schedule, names the three things that blow it up, and tells you what should be live in week one so your business is not invisible for a quarter.

How long does a service business website really take?

A 10-to-20-page service business website takes 6 to 10 weeks from kickoff to launch when the client side moves fast, and 4 to 6 months when it doesn't. The spread is not caused by design difficulty. It is caused by waiting.

Strip the calendar down to actual labor and a site like this is roughly 25 to 35 working days of work for a small team: information architecture, wireframes, two design rounds, front-end build, CMS setup, QA, launch. Nothing on that list takes three months.

What takes three months is the eleven days your homepage copy sat unread in an inbox, twice. So the honest answer to 'how long' is: about as long as it takes you to make forty decisions. Any agency that answers without mentioning that is setting you up to be surprised in week seven.

What is the realistic week-by-week timeline?

Here is the 9-week schedule we'd hold a 12-page service business site to, assuming you answer questions within two business days. The last column is the honest part — what actually kills each week.

Nine weeks. Notice that you are on the critical path in three of the six phases. Every week added to that schedule is, almost always, a week nobody answered.

Why is the delay almost always on your side, not the agency's?

Because the two things that push a build late — scope creep and approval latency — are both controlled by the buyer. PMI's 2018 Pulse of the Profession found that 52 percent of projects completed in the previous 12 months experienced scope creep, or uncontrolled changes to the project's scope, up from 43 percent five years earlier. That figure covers projects of every kind, not just websites — but the mechanism is identical.

Scope creep on a website almost never starts with the agency. It starts with 'while we are in there, can we add a careers page?' Each request is small and reasonable. Ten of them is a month.

The second half is sequencing. Design, build, and QA run in order, so a quiet week in week 3 does not get absorbed — it slides week 8. An agency can eat one slow week. It cannot eat four.

Most agencies will never write this down, because 'you are the bottleneck' reads like blame in a sales cycle. We'd rather you be a well-armed buyer than a comfortable one. That is also why we publish what a website engagement costs instead of making you book a call to find out.

Which three client-side bottlenecks cause most overruns?

Three items account for most of the slip on a service business build: copy approval, photography, and owner-only decisions. All three are yours, and all three can be finished before kickoff — which is the single highest-leverage thing you can do to a timeline.

• Copy approval. The words are the project. A design cannot be finished around copy that doesn't exist, so an agency will fill it with placeholder text, you'll approve the layout, then the real copy arrives 400 words longer and the design breaks. Settle copy first, design second.
• Photography. A half-day shoot of your actual crew, trucks, office and finished work is a one-week lead time to book and a one-week edit. Start it in week 1, not week 6. If you skip it, you get stock images of a stranger in a clean hard hat — and buyers can tell.
• Owner-only decisions. Which service do you want more of? What do you refuse to do? What is the phone number that a lead should actually ring? Nobody at the agency can answer these. Block two hours in week 1 and answer all of them at once.

None of this requires a project manager. It requires one person on your side with authority to say yes and a calendar slot to say it in.

What should be live in week one instead of three months from now?

One page: your highest-margin service, a phone number, a form, and a booking link — live within 5 business days of kickoff. A business that goes dark for three months while a redesign is 'in progress' is paying for that redesign twice, once in fees and once in the leads that searched during the blackout.

Phase the launch. Ship the money page for the service you most want to sell, point your ads at it, and let the rest of the site land behind it over the following weeks. You get a lead-generating asset in week one and real conversion data by the time the full site launches.

This is also how you avoid the worst outcome in web design: a beautiful nine-week build whose hero section turns out to be wrong, discovered in week ten. Test the hero and offer that convert service business buyers on a live page while the rest is still being designed.

Redesigns are normal, not exceptional. In Clutch's 2025 survey of 406 US small business owners, 81% said they had already redesigned their website at least once, and 90% planned to invest in their site over the next 12 months. Treat the site as a thing you ship in increments, not a monument you unveil.

What is a 5-day template build actually skipping?

A five-day build hits five days by skipping discovery, original copy, photography, QA and measurement — which is roughly the entire list of things that decide whether the site produces phone calls. The template itself is genuinely fast. Everything that makes the template yours is what got cut.

That QA line is not a nicety. Google's Core Web Vitals thresholds are an LCP of 2.5 seconds or less, an INP of 200 milliseconds or less, and a CLS of 0.1 or less, each measured at the 75th percentile of real page loads. A template loaded with four sliders and six plugins misses those quietly, and you find out from your page speed and conversion numbers, months later.

Honest verdict: if you are pre-revenue and need a URL to put on a truck this week, buy the template. It is the right call. If the website is the thing that has to make the phone ring at $2,500-a-job margins, five days is not a schedule — it is a skipped scope. Upwork's 2026 cost guide puts website builds anywhere from $300 to $5,000 for exactly this reason: the range is not one product at different prices, it is different products.

How do you keep a build from stalling in revision rounds?

Cap design at two rounds, write every piece of feedback as a line-level edit with a named owner and a due date, and lock copy before design starts. Revision rounds do not stall because two rounds is too few — they stall because feedback arrives as taste instead of instructions.

• Lock copy before design. Design against real words or you will redesign against real words.
• One decision-maker. Feedback aggregated from five people in a group chat is not feedback, it is a tie. Nominate one person who can say yes.
• Line edits, not adjectives. 'Move the phone number above the fold and cut the second paragraph' is actionable. 'It feels corporate' costs a week.
• A 48-hour clock on approvals. Put it in the agreement. It protects your launch date more than any clause about the agency's.
• A written change log. Anything requested after sign-off gets logged with a time and cost, and gets shipped after launch. Not never — after.

One more thing worth setting expectations on: launching the site is not the same as ranking. A new site's design goes live in weeks, but SEO results take months to compound. Budget the calendar for both, or you will judge a nine-week build on a metric that was never going to move in nine weeks.

If you want a build that runs on a published schedule instead of a shrug, look at how we scope conversion-focused website design — month-to-month, you own the codebase, no lock-in. Not sure whether you need a rebuild or a repair? Get my free audit and we'll tell you which one, in writing.