How to Design and Build Complex Platforms with Shape Up

Building a complex platform — a SaaS, an internal tool, a management system — is an exercise in managing uncertainty. You do not know exactly how long it will take, you do not know all the problems you will encounter, and initial specifications always change when reality shows up. The question is: how do you manage that without the project spiraling out of control?

At Villelabs we use Shape Up to build complex platforms. It is not the only way to do it, but it is the one that has worked best for us. Here is how.

The concept of appetite

In Shape Up we do not estimate how long a project will take. Instead, we define how much we are willing to invest: the appetite. It is a subtle but fundamental shift.

When you say "this feature will take 6 weeks," you start negotiating time. When you say "we are giving this feature 2 weeks," you start negotiating scope. And negotiating scope is exactly what you need when building something complex, because it forces you to separate the essential from the nice-to-have.

For a typical SaaS platform, we divide development into 2-week cycles. Each cycle has a clear appetite and a defined deliverable. The first cycle might be "authentication + basic dashboard." The second "the main user flow." The third "integrations with external APIs." Each one is independent and each one delivers value.

Shaping: the invisible work that makes the difference

Before the development team starts working, the project goes through a shaping phase. This is not a 50-page requirements document. It is a process where we define:

• The real problem we are solving (not the solution the client imagines, but the problem behind that solution).
• The boundaries of the cycle: what is in scope and what is explicitly out.
• The technical risks that could make the cycle fail. If we identify a big risk, we tackle it first.
• Rough sketches of the solution: not pixel-perfect wireframes, but flow diagrams and whiteboard sketches that give direction without limiting the team's creativity.

Shaping takes between half a day and two days per cycle. It seems short, but that time prevents weeks of rework.

Hill charts: knowing where you actually are

One of the most useful concepts in Shape Up is hill charts. Imagine a hill: the uphill side is the "figuring out how to do it" phase (high uncertainty), and the downhill side is the "executing what we already know" phase (low uncertainty).

Each piece of the project sits on the hill. If something has been stuck on the uphill side for days without moving, it is a signal that there is an unresolved problem — and it is better to address it now than in the last week of the cycle. This replaces the meaningless "we are 60% done" with real information about where uncertainty lives.

Cool-down: the secret to not burning out the team

After each cycle there is a cool-down period of a few days. The team fixes minor bugs, explores ideas for the next cycle, or simply recharges. In platform projects that last months, this is not a luxury — it is what keeps quality consistent. A burned-out team in cycle 5 will produce fragile code that costs twice as much to fix later.

Real example: lead management system

A client needed a system to manage leads from multiple channels (web, WhatsApp, Instagram). Instead of planning 12 weeks of development, we split it like this:

• Cycle 1 (2 weeks): Lead ingestion from web form + list view with basic filters.
• Cycle 2 (2 weeks): WhatsApp Business API integration + automatic assignment to salespeople.
• Cycle 3 (2 weeks): Metrics dashboard + visual Kanban-style pipeline.
• Cycle 4 (2 weeks): Instagram DM integration + automatic AI-based lead scoring.

By the end of cycle 2, the client was already using the system in production. Cycles 3 and 4 were adjusted based on what we learned from real users. That would not have been possible with a rigid 3-month plan.

At Villelabs, we use Shape Up to build platforms that ship in functional increments, not as a big bang at the end. If you have a complex project and want to know how we would approach it, let us talk.