The short version
A two week build works because the scope is decided before the clock starts, not because anybody is working faster than is sensible. What fits is a defined set of pages on a system we have already built and tested many times, with your real business facts in them. What does not fit is anything that depends on a third party we cannot schedule, anything requiring content that does not exist yet, and anything where the requirements are still being discovered. Those are not refusals, they are second phase items with their own dates. The reason to publish this list rather than bury it is simple: nearly every overrun in this industry comes from an exclusion nobody wrote down, and the argument that follows is always about what someone assumed rather than what anyone agreed.
Why a timeline needs an exclusion list to mean anything
Any agency can promise two weeks. The promise costs nothing to make, and when it slips the explanation is always the same shape: something came up that was not really in scope. The client heard a deadline. The agency heard a deadline for a project that turned out to be different from the one that arrived.
So we write the exclusions down before anyone signs, because a deadline without a scope is a guess with a confident tone. What follows is that list, and the reasoning behind each item, because the reasoning is what tells you whether your project fits.
What actually fits in fourteen days
A twelve page site on a system we have built many times, which is the important part. We are not inventing a layout engine, a booking flow or an analytics setup for each client. Those exist, they have been through real traffic, and the fortnight goes into fitting your business into them rather than into building the machinery.
Inside that: home, services overview, individual service pages, the trades or areas you serve, about, contact, and the legal pages. Your photos, your prices or price ranges, your service area, your people. A working quote form that arrives somewhere a human checks. Tracking that records the whole chain rather than just sessions. Structured data so search engines and assistants can read what the business is and where it operates. A speed budget the site has to pass before it goes live.
That is a real website that can take a customer from a search result to a booked job. It is not a placeholder, and we do not ship a phase one that needs a phase two to function.
What does not fit: anything waiting on a third party
The most common cause of a slipped launch is not development. It is a domain registrar login nobody has, a hosting account tied to a former employee's email, or a merchant approval sitting in a queue at a payment provider.
We cannot schedule other people's companies. So access and approvals sit outside the fourteen days by design, and the clock starts when we have what we need, not when the contract is signed. That one rule removes most of the disputes this industry has about deadlines.
What does not fit: content that does not exist yet
If your prices are not decided, your service list is still being argued about internally, or there are no photos of your work, the build is not the bottleneck and a shorter timeline will not help.
We can write, and we do. What we cannot do is invent facts. A page about your emergency response time needs to know your emergency response time. A page listing the towns you cover needs a real list, not a hopeful one. When those answers are missing, the honest move is to pause rather than to publish something plausible, because plausible invented details are the ones that cause arguments with customers later.
The intake call exists to find these gaps in the first two days rather than day eleven. It is a ninety minute conversation and it is the part of the process most likely to change your launch date, in either direction.
What does not fit: custom application work
Customer portals, real time inventory, multi step configurators that price a job from thirty inputs, integrations with field service software that has no public interface. These are software projects. They have their own discovery, their own testing and their own failure modes, and pretending they fit next to a marketing site is how both end up late.
We will scope them separately with their own dates, or we will tell you we are not the right shop for it. The second answer is more common than you might expect and it is cheaper for everyone than finding out in month three.
What does not fit: rankings
This one needs saying plainly because the whole industry is vague about it. Fourteen days gets you a site that is technically capable of ranking. It does not get you rankings.
A new page has to be crawled, indexed and then evaluated against everything already occupying those results, and that runs on a timescale nobody controls. For a local service business in a normal market, meaningful movement starts around the ninety day mark, and the competitive terms take longer than that. Anyone giving you a shorter number for rankings is either quoting a term nobody searches or hoping you stop checking.
What can change immediately is conversion. A site that loads quickly, says what you do, shows a price and has a phone number people can actually tap converts better from the traffic you already have, and that shows up in weeks rather than months.
What does not fit: unlimited revisions
Two rounds, which covers essentially every project that came in with a clear picture of the business. Unlimited revisions sounds generous and behaves badly. It removes every reason to make a decision, and a project where nothing is decided runs until somebody gets tired.
The revision rounds are also positioned deliberately. The first comes after structure and before visual detail, because moving a page at that point is cheap. The second comes near the end, on real content in a real layout. Reordering navigation on day twelve is expensive in a way that is not obvious from the outside.
What happens to the things that do not fit
They go on a written list with dates, and it is a real list that we work through, not a way of ending a conversation.
Location pages come after launch because each one needs facts gathered by an interview, and interviews take calendar time. Content programmes run monthly because publishing twelve articles in a fortnight produces twelve mediocre articles. Integrations get scoped once we can see how the business actually runs, which is usually clearer a month after launch than during it.
None of that is a trick to sell a second project. It is an admission that some work is gated by time and information rather than effort, and squeezing it into a fortnight produces something you would have to redo.
How to test any fixed timeline you are offered
Ask for the exclusion list in writing, before you sign. Ask what specifically starts the clock. Ask what happens on the day the deadline is missed, and whether there is a stated consequence or just an apology.
If the answers are vague, the timeline is decoration. A shop that has actually delivered the same build repeatedly knows precisely where it breaks, and can tell you without checking with anyone. A shop that cannot describe its own failure modes has not hit them often enough to have learned them, which is a different kind of answer to the same question.
Why we would rather lose the job at the quote
Saying all of this out loud costs us work. Some people want to hear yes to everything and will hire whoever says it, and that is a fair choice to make with your own money.
The trade is worth it, because the projects that go badly are almost never the ones where somebody was told no early. They are the ones where nobody wanted to have an awkward conversation in week one and everybody had a much worse one in week seven. A quote that names its own limits is the cheapest form of project insurance either side can buy.
Check it on your own site
Write down what you already have
Domain access, hosting login, photos, price list, service area, the names and roles of your people. Everything missing from that list is a day the clock has not started, no matter what anyone promised.
Separate marketing pages from software
Go through your wish list and mark each item as a page or as an application. Portals, calculators and integrations belong in the second group and need their own timeline, whoever builds them.
Decide your prices before the build
Even ranges. A site that cannot show a number loses to one that can, and this is the item most likely to stall a launch because it needs a decision only the owner can make.
Ask for the exclusions in writing
From us or from anyone else quoting a fixed timeline. The list itself matters less than whether it exists, because a shop that has one has been through this enough times to know where it goes wrong.
Agree what counts as launched
Live on your domain, forms delivering to a monitored inbox, tracking recording, speed budget passed. Write that definition down so the finish line is a checklist rather than a feeling.