Where the weeks actually go
The timeline surprises people because the coding is the short part. On a six-to-ten week marketing site, roughly the first half is spent before anything is built.
- Discovery, one to two weeks — who buys, why they hesitate, what the competition owns, what the site has to accomplish, sitemap and content structure agreed.
- Design, two to four weeks — creative direction, then structure, then interface, reviewed against your real content rather than placeholder text.
- Build, two to three weeks — frontend development, content system, integrations, motion, responsive work across breakpoints.
- Launch preparation, one week — quality assurance, accessibility checks, performance against the budget, analytics wired, redirects mapped if a site already exists.
- Post-launch, two to four weeks — watching real traffic and fixing what it exposes, because it always exposes something.
Why projects run late, in order of frequency
In our experience the developer is rarely the bottleneck. Four things are, and three of them sit on the client's side — which is good news, because it means the timeline is mostly under your control.
Content. 'Client to provide copy and images' is the single largest cause of stalled websites. A project can sit for two months waiting on product descriptions and team photographs. If writing is not in scope for the studio, someone on your side needs it as an actual assignment with a date, not an intention.
Feedback cycles. A design review that takes six days instead of one adds a week per round, and there are several rounds. The worst pattern is serial feedback — three people commenting on three different days, each contradicting the last. Consolidate it into one response.
Approval chains. If the person who can say yes was not in the kickoff, expect a late reversal on something already built. Get them in early or accept the risk knowingly.
Access. Domain registrar logins, hosting credentials, analytics accounts, payment gateways, the previous developer's handover. Chase these in week one, because in week nine they become critical and the person who has them is on holiday.
What genuinely takes longer
Some things add weeks legitimately and should be planned for rather than discovered.
E-commerce adds catalogue work: product data, specifications and photography in a usable, consistent state. This is almost always the critical path on a store, not the build. A 500-product catalogue with inconsistent naming is weeks of work before a page can be designed.
Integrations add uncertainty rather than fixed time. A payment gateway is predictable; a twelve-year-old ERP with no documentation is not. Feasibility gets checked in discovery, and any studio that skips that check is quoting on hope.
Multi-language sites roughly double the content timeline, because translation happens after copy is final and is usually somebody's side task.
A migration adds URL mapping and redirect work. Skipping it is how a rebuild loses years of accumulated search equity in a weekend.
What can be compressed, and what cannot
If you have a hard date — a trade show, a funding round, a season — say so at the first conversation rather than in week four. There are legitimate ways to move faster.
Reduce page count instead of quality. Six pages that do their job beat eighteen half-finished ones, and you can add pages later without re-arguing the site.
Launch in phases. Get the pages that matter live and follow with the rest. A live site earning attention beats a complete site three weeks from now.
Have the studio write the copy. This usually shortens the project despite adding scope, because it removes the longest wait.
Pre-book the review slots. Put the design reviews in calendars at kickoff, with the decision-maker in them.
What cannot be compressed is discovery. Every project we have seen go badly wrong skipped or rushed it, and the time saved came back multiplied as rework.
A realistic answer for your situation
If you need a simple, credible site live in three weeks and it will not be a sales channel, a well-configured template is the honest recommendation. Buy it, spend the difference on something that earns, and revisit in a year.
If the site needs to explain a business properly and convert, plan for eight to twelve weeks including your own time, and treat content as the thing that decides whether you hit that. If you are being quoted three weeks for that scope, ask what is being left out — usually strategy, copy and performance work.