Why "how long will it take" is the wrong opening question
A number on its own tells you nothing. The same "site in a month" can mean a template with swapped photos or a landing page built around your own argument. The useful question is what the number is made of — that is where you see whether a deadline is real or a sales answer that will slip in week two.
We quote by stage, and each stage has a floor below which work does not speed up, it just gets worse.
Where the weeks go
Structure and wireframes — one to two weeks. This is where you decide what a visitor sees and in what order they get the argument. It is the cheapest stage to change and the most expensive to get wrong: moving a block on a wireframe takes minutes, moving the same block after build takes a day.
Design — two to four weeks. Driven by the number of distinct screens, not the number of pages. A two-hundred-item catalogue is two layouts, not two hundred. A landing page with seven unlike sections is seven separate decisions.
Build — from a week and a half. Markup, assembly, forms, analytics. The number grows with integrations: every external system means agreeing a format, getting credentials and handling the case where it goes quiet. Why we build on code rather than a builder is covered in Next.js versus website builders.
Content — parallel, but not free. Copy, photography, specifications. On paper it runs alongside design. In practice it is the single most common reason a project stalls: the layouts are ready and there is nothing to put in them.
What actually moves the date
In our projects deadlines are rarely lost in design or build. They are lost in three places, all of them on the organisation side.
Diffuse sign-off. When five people review a layout and each sends notes separately, every review round doubles. The working arrangement is one owner who collects opinions internally and brings back a single decision.
"We'll send the content later." The most common cause of dead time. We put the required materials and their due dates into the brief up front. If the substance is not there, moving the start is more honest than designing around placeholders and redoing it later.
Late scope changes. Adding a section during design costs days. Adding the same section after build costs a week, because everything around it has to be reassembled. That is not a contractor being difficult; it is how the work stacks.
The quiet time sinks
Three things eat weeks without anyone noticing until the end.
Credentials. Domain, hosting, mail, CRM, inventory system — each has an owner, and the owner is rarely in the thread on day one. Requesting access belongs in week one of the plan, not the week it is needed.
Legal copy. Privacy policy, terms, company details. Half a page of text and a lawyer with their own queue. The site cannot go live without it, and it is regularly the last open item.
Photography. A shoot has its own schedule: the location, the weather, the people, the retouching. If the project depends on original images, book the date at the start, not when the layouts are waiting.
Rough figures
- Single-project or product landing page — four to five weeks.
- Corporate or manufacturer site — six to eight weeks.
- Online store — from eight weeks, then it depends on catalogue and integrations.
- Property sales site with unit selection — from ten weeks, and most of it is data.
These assume the substance is ready and one person on your side can decide. We fix the estimate and the date before we start: if our own estimate slips, that is our problem, not a reason to invoice more.
Three questions for any contractor
Ask what stages the number is made of, what has to be ready on your side and by when, and what happens to the date if your feedback arrives a week late. A clear answer to the third question is what separates a plan from a promise.
