Home / Blog

How long does it take to build custom software?

The honest range runs from a few days for a small internal tool to many months for a large platform. For a system that runs a mid-sized company, it should be weeks. Here's what decides where your project lands.

5 min read · 28 September 2026

The short answer

For a company of 20 to 300 people, a well-scoped system that replaces spreadsheets and scattered tools should go live in weeks, not quarters. With us, a single-company system with up to five roles is typically live in three to four weeks. A system with the AI layer, integrations, automations and a client portal takes six to eight weeks. Groups of companies across several countries are rolled out in phases.

And before any of that, you should see a first working version on your own data. With us, that's within 24 hours of kickoff. It's small, but it runs, and it tells both sides whether the understanding is right.

Those numbers assume a responsive client and a clear first phase. They're not the time it takes to write code; they include mapping the process, migrating and checking data, testing with real users and a staged go-live.

Why some projects take years

You've probably heard of software projects that ran for two years and still didn't launch. They tend to share the same causes.

  • Everything in one release: nothing goes live until every module is finished.
  • Long discovery phases: months of workshops and documents before anything is built.
  • Hourly billing: no commercial pressure to finish, and every unclear point becomes more hours.
  • Committee decisions: every question needs a meeting, and meetings need weeks to schedule.
  • Late data migration: the data doesn't fit the new structure, discovered in the final month.
  • Changing goals: the business changes during the long project, so the target moves.

What actually decides the timeline

The same variables that drive cost drive time.

  • Roles: how many different kinds of users need their own screens and permissions.
  • Hand-offs: how many times work changes hands, and how many rules govern those hand-offs.
  • Integrations: each connection to accounting, banking, email, calendars or industry platforms adds its own small project.
  • Data: how much there is, how messy it is, and how quickly someone on your side can confirm it's correct.
  • Decisions: how fast your side answers questions. This is often the biggest factor, and the one you control.

Typical timelines by type of project

These are rough market ranges for well-run projects, not guarantees. They assume a clear scope and a responsive client.

  • A single internal tool for one team and one workflow: days to a few weeks.
  • A CRM or operations system for one company with a handful of roles: a few weeks.
  • A full company system with integrations, automations, AI and a client portal: one to two months.
  • A group-wide system across several companies, countries and currencies: several months, delivered in phases so parts go live early.
  • A public software product for thousands of external users: months to years, and a different kind of project altogether.

Why a first version in 24 hours is possible

People are sometimes sceptical when they hear 'working version in 24 hours'. The reason it's possible is not magic or shortcuts. Most business systems share a large common base: users and permissions, records with statuses, lists and filters, forms, files, notifications, audit logs, mobile views. We've built that base many times, so we don't rebuild it for each client.

What we build in the first day is the part that's specific to you, at small scale: your core unit of work, your main roles, your statuses, on a sample of your real data. That's enough to check whether we've understood how your company works. Everything after that is the careful work of covering every hand-off, migrating all the data and making it solid.

How to speed things up on your side

Most delays in business software projects come from waiting, not building. You can remove a lot of that.

  • Name one decision-maker who can answer questions within a day.
  • Share the real material early: spreadsheets, exports, chat screenshots, forms.
  • Give the builder access to a sample of real data in the first week.
  • Get the actual users to try each new part as soon as it's ready.
  • Keep a 'later' list, and be strict about what goes in the first phase.
  • Plan go-live away from your busiest season.

How to tell a credible schedule from a hopeful one

A credible schedule is tied to scope you can check: these roles, these hand-offs, these integrations, this data. It has milestones where you'll see working software, not just documents. And it comes from a builder who has already looked at how your company actually works.

Be wary of two extremes. A quote of many months with nothing to see until the end is a risk to your budget and your patience. A promise of a complete company system in a few days is usually a template with your logo, or a demo that will need a lot more work before anyone can run a business on it. A first working version in 24 hours is realistic because it's deliberately small. The full system still takes weeks of careful work with your team.

Does faster mean worse?

Not necessarily. Speed in well-run projects comes from starting with proven building blocks, scoping tightly, and cutting the waiting time, not from skipping testing or security. We start from parts we've built many times and spend the time on what's specific to your company.

What should never be rushed: data migration checks, permissions, backups and the go-live plan. If a fast schedule leaves no room for those, it's the wrong kind of fast.

Timelines for groups and phased rollouts

When a system spans several companies, countries or currencies, the timeline changes shape. Instead of one go-live date, there's a sequence. The first company or division goes live with the core system, usually within the same weeks a single-company project would take. The next ones follow, each adding what's specific to them: local rules, languages, currencies, integrations with their own accounting.

Consolidated reporting across entities comes once enough of the group is on the system to make it meaningful. This approach means the first part of the business sees value early, and each later rollout is shaped by what the earlier ones learned.

What happens after go-live

Launch isn't the end. The first months of real use always bring adjustments, so the timeline should include a support period. Ours runs three, six or twelve months depending on the plan (Core €18,000, Maxxed €48,000, Empire from €120k), with Maxxing Care at €2,900 per month afterwards.

Get a real timeline for your company

The fastest way to a credible timeline is to show a builder how your company runs today. Apply and walk us through your spreadsheets, chats and tools. We'll map the hand-offs, give you a fixed price and a go-live date, and show you a first working version on your own data within 24 hours of kickoff.

Questions

Can custom software really be built in a few weeks?

A focused system for one company can, when the builder starts from proven parts, scopes tightly and the client answers questions quickly. Larger systems should be phased so parts go live early.

What is a realistic timeline for a custom CRM?

With us, a single-company CRM is typically live in 3 to 4 weeks. With AI, integrations and automations, 6 to 8 weeks.

What causes the most delays?

Waiting for decisions, data and feedback from the client side, and data migration left until the end.

What is a first working version?

A small but real system running on a sample of your own data, which your team can click through. With us, it arrives within 24 hours of kickoff.

Does a phased rollout take longer overall?

Sometimes slightly, but you start getting value from the first phase within weeks, and each phase is shaped by real use of the previous one.

Keep reading