Five questions, about thirty seconds. You get a real timeline and a real price range, not a brochure.
These are the same prices on our pricing page. The range comes from how long the build takes, not from what we think you can afford.
The five questions are not a lead-capture ritual. Three of the answers go into the arithmetic and two do not, and you should be able to sanity check the range before you send us anything.
A brochure site, a SaaS product with logins and billing, and an iOS and Android app are three different amounts of work. This is the single biggest move in the range, which is why it is the first question.
Extending something live is about a quarter less work than standing the same scope up from nothing. The pipeline, the auth, the data model and the design language are already decisions somebody made.
Sized against concrete anchors written per build type, not a generic small, medium and large. Guessing that you are small when you are not is how an estimate goes wrong before anyone speaks.
Your industry and who builds for you today. Both are asked because they shape the first call, and neither touches the arithmetic. Nobody is priced higher for looking like they can pay more.
Week one is the pilot at $1,000: one deliverable, scoped in writing, live by day five, and credited against your first month if you carry on. Every week after it is billed at the $2,500 a month rate, which is where the rest of the range comes from. The bottom of a range rounds down and the top rounds up, because a ceiling quoted below what the work costs is a worse conversation than showing a little more.
The estimate is a starting point for a scoping call, not a quote. Nothing here is binding until the work is written down and we both agree on it. If the range looks wrong for what you have in mind, that is worth a conversation on its own.
Whatever the calculator shows you, this is what is inside it. Most studios quote these separately, or skip them and leave you to find out after launch. They are in the price of every engagement, including the pilot week.
Flows and interface design that survive contact with real users and real edge cases.
Every merge builds, tests, and deploys itself. Rollback is one command, not an incident.
Tests that fail for real reasons, running on every pull request before anything reaches your users.
Your cloud account, your domains, your control. Nothing important lives on our side of the fence.
Error tracking and uptime alerts wired in before launch, so you hear it from us and not from a customer.
Dependency scanning, secret management, and sensible defaults. Not a compliance theatre exercise.
Agents and retrieval built with evaluation and a human checkpoint, when the problem actually calls for them.
A documented repo and a walkthrough with whoever inherits it. You are never locked to us to keep it running.