Why Your Website Redesign Will Take 11 Weeks

Most redesign quotes say something like "four to six weeks". Most redesigns take two to three months. The gap is not usually the agency being slow. It is the work that sits on your side of the table and never appears on anyone's schedule: writing the About page, finding the GST-compliant terms of service, deciding who signs off, tracking down the login for a domain registered by a cousin in 2016.

Below is an honest timeline for a fifteen to thirty page business site with a contact form, a few service pages and an existing site to replace. Not an ecommerce build, not a portal. Where things slip, I have said so and said why.

Before week one: the things that decide everything

There is a stretch before any design happens that most owners treat as admin. It is not. Two decisions made here shape the next ten weeks.

Who approves. Name one person. Not a committee. If your spouse, your sales head and your brother-in-law all get a veto at different stages, every round of feedback becomes three rounds, and each round is a week. The single most common cause of a three-month build becoming a six-month build is unclear approval authority.

Where the content comes from. Somebody has to write the words. If you have not decided whether that is you, a freelancer, or the agency, you have not decided, and the project will stall at week four. This is worth being blunt about: content is the slipping point in nine redesigns out of ten.

Also gather, before anything starts:

• Domain registrar login (the actual registrar, not your hosting panel)
• Current hosting login and, if possible, cPanel or SSH access
• Google Analytics and Google Search Console access, or an admission that nobody knows who set them up
• Any Google Business Profile that points at the site
• Your logo in a vector file — .ai, .eps or .svg — not a JPEG someone pulled off a brochure

If the domain login is lost, budget an extra week and expect a support ticket with GoDaddy or BigRock involving identity documents. This happens often enough that it is worth checking on day one rather than day sixty.

Weeks one to three: discovery, structure, first design

Week 1 — discovery and sitemap. A couple of conversations about what the site has to do. Not "look modern". Do people ring you, fill a form, or walk in? Do you need enquiries qualified before they reach you? The output is a sitemap: an ordered list of every page and how they nest. This is the cheapest hour in the whole project to argue about. Moving a page on a sitemap costs nothing. Moving it in week eight costs design, build and internal links.

Week 2 — homepage design. One page, designed properly, in desktop and mobile. Resist the request to see everything at once. If the homepage direction is wrong, seeing eight wrong pages does not help.

Week 3 — feedback and revision. Where this slips: feedback arriving as "I don't love it" rather than "the enquiry form should be above the fold and the blue is too close to our competitor's". Vague feedback produces a second design that is differently wrong. Ask for specifics, in one document, from one person.

Weeks four to six: the content wall

This is where the schedule breaks. Design is approved, the developer is ready to build, and the site needs about 4,000 to 8,000 words of real copy that only you can supply the substance of.

What typically happens: the agency sends a content template, the owner intends to fill it on Sunday, three Sundays pass, and the project sits at 60 per cent for a month. Nobody is being unreasonable. Writing about your own business is genuinely hard and it is nobody's urgent task.

Three ways to break the wall, in order of how well they work:

1. Interview and draft. You talk for forty-five minutes per section, someone else writes it, you correct it. Correcting a draft is ten times faster than facing a blank page.
2. Reuse ruthlessly. Your existing site, brochures, WhatsApp replies to common questions and old proposals contain most of what you need. Editing beats inventing.
3. Ship in stages. Launch with the pages you have. A live site with twelve good pages beats a staging link with twenty-five perfect ones that nobody sees.

Photography sits here too. Stock images of white people in glass offices actively hurt an Indian services site — visitors read them as a signal that there is nothing real behind the page. Half a day with a local photographer covering your premises, your team and your work is usually the highest-return spend in the whole redesign. And do get the raw files: a photo straight off a phone is often four megabytes, and it needs resizing before it goes anywhere near a live page.

Weeks six to nine: build, and the parts you cannot see

Weeks 6–7 — page build. Templates get built, content gets poured in, forms get wired to an inbox. If enquiries should land somewhere more organised than a shared Gmail, this is the point to connect a CRM or enquiry management system, not after launch.

Week 8 — technical and redirects. The genuinely important, entirely invisible week. If your existing site has ranking pages and the URLs change, every old address needs a 301 redirect — a permanent forwarding instruction — to its new equivalent. Skip this and old links from Google, directories and other people's sites hit a 404 error page. Rankings that took two years to build can evaporate in a fortnight.

Also in this week: image compression, page speed work against Core Web Vitals (Google's public thresholds are 2.5 seconds for Largest Contentful Paint, 200 milliseconds for Interaction to Next Paint and 0.1 for Cumulative Layout Shift), title tags, structured data and an SSL certificate that actually covers both the www and non-www versions.

Week 9 — review. You go through it properly, on a phone and a laptop, and list what is wrong. Expect forty items. That is normal, not a disaster.

Weeks ten and eleven: launch and the fortnight after

Launch is a Tuesday morning task, not a Friday evening one — you want two working days to catch problems while everyone is available.

On launch day: DNS change, SSL verified, forms tested from a phone on mobile data, Search Console updated with the new sitemap, old analytics still recording. In the week after: watch Search Console's coverage report for 404s you missed, check that your Google Business Profile links resolve, and confirm enquiries are actually arriving. It is depressingly common for a form to look like it submitted while the mail never leaves the server.

Traffic often dips for two to four weeks after a launch while Google recrawls. If your redirects are right, it recovers. If it has not recovered in six weeks, something is wrong and it is worth a proper technical SEO review rather than waiting hopefully.

What I would not buy in a redesign

A full brand identity bolted on. If you came for a website and leave with a logo redesign, a colour system and a brand book, your eleven weeks became five months and the website is no better at generating enquiries. Do the brand work separately, if you need it at all.

Custom-built anything that a plugin does adequately. A bespoke booking module or blog engine costs more to build and much more to maintain. Custom code earns its keep where your process is genuinely unusual — see our note on when a custom build is worth it — and nowhere else.

An unlimited-revisions promise. It sounds generous. In practice it removes the pressure that gets decisions made, and projects with unlimited revisions finish later and worse. Two named rounds per stage, with a clear cost for more, serves you better.

A launch-day "SEO package" bought as a bolt-on. Redirects, titles, speed and structured data belong inside the build. If they are being sold to you as an extra afterwards, ask what the build actually included.

What to do next

Before you speak to anyone about a redesign, do three things this week. Confirm you can log in to your domain registrar. Decide, in one name, who approves designs. And open a document and write the three sentences you would say to a stranger who asked what your business does — that document is the seed of your homepage, and starting it now removes the single biggest cause of delay.

When you are ready to scope a build properly, tell us what you have and what is broken about it, or read more in the blog first. Either is fine. Losing the domain login is not.

Questions we get asked