“How long will this take?” is the second question every client asks, right after what it costs. The useful answer is not a single number, because the same brief can honestly be a six-week project or a six-month one depending on decisions nobody has made yet. Here are the real timelines we quote, and the four things that move them.
Typical build times by project type
- Internal tool or dashboard: 3 to 6 weeks. One user group, one system of record behind it, a clear list of screens. The fastest thing we build.
- Customer portal: 4 to 10 weeks. Status, documents, messaging, and roles. The interface is quick. The API and permissions model underneath it are where the weeks go.
- AI document intake pipeline: 3 to 8 weeks. A bounded set of documents, a bounded set of fields, and a human review queue for anything uncertain.
- Two-sided marketplace: 10 to 16 weeks. Two user types, two onboarding flows, payments, trust and verification. Every feature exists twice.
- Mobile or desktop app: 6 to 12 weeks. Add app store review time on top, which is calendar time you cannot compress by working harder.
- Legacy modernization: 6 to 12 weeks per phase, after a short paid audit. It runs in phases by design, because carving a system seam by seam is what keeps the business running while it happens.
Those are build weeks, starting the day we write the first line of code. Add one to three weeks before that, because we take a limited number of projects at a time and start dates are usually a couple of weeks out.
The four things that actually add weeks
Integrations. Every third-party system is a scope item, not a footnote. A clean, documented, well-behaved API adds days. An undocumented SOAP endpoint owned by a vendor who answers email twice a week adds weeks, and none of that time is spent writing your software.
Approval chains. This is the one nobody prices. If every design decision needs sign-off from four people and one of them travels, the build does not go faster than their calendar. The single biggest schedule risk on most projects is not engineering difficulty, it is how quickly someone on your side can say yes.
Data migration. Moving ten years of records out of a system that let people type anything into any field is usually harder than building the thing that will consume them. Budget real weeks for reconciliation, not an afternoon for a script.
Undefined scope. “And it should probably also do X” is how a six-week project becomes a fourteen-week project without anyone deciding to. The fix is not a stricter contract. It is a demo every Friday, so a change gets raised and priced the week someone thinks of it instead of discovered at the end.
Why we ship weekly instead of quoting a launch date
We give you a launch date, but the number that actually protects you is the weekly one. Working software goes on a staging link every Friday from week one. That does three things a Gantt chart cannot.
It makes progress checkable instead of reported. You click the thing. Nobody has to trust a status percentage.
It surfaces the wrong assumptions in week two rather than week nine, when they are cheap to fix. Most timeline overruns are not slow engineering, they are three weeks of building the wrong understanding of a workflow.
And it means you can stop. At any weekly demo you can end the project and pay only for the weeks delivered. Everything shipped up to that point is already deployed, documented, and yours. A studio that cannot let you do that is asking you to bet the whole timeline up front.
How to make your project faster, for free
The things that most reliably compress a timeline cost nothing:
- Name one decision-maker. One person who can approve scope without convening a committee is worth more to your schedule than an extra engineer.
- Get your data and credentials ready before week one. API keys, sample exports, access to the staging environment of whatever we integrate with. Waiting on access is pure dead time.
- Cut the second user type from v1. Two roles is not twice the work, but it is close. Ship for the group that matters most, then add.
- Decide what you are not building. A clear no-list is worth more than a long yes-list.
The honest version
Any studio that gives you a timeline before a real conversation is guessing, and a guess that is too optimistic is not a favor. Ours comes out of a free 30-minute scope call, in writing, within 48 hours, alongside the fixed price and the scope it covers.
If you want the real number for your build, see what a web application project typically involves, or get an estimate.