3.5 Building a Minimum Viable Product (MVP)
An MVP is the simplest version of your product or service that lets you start the learning process. It has just enough features (or manual processes) to test your core hypothesis with real customers. Lean Startup boils down to this: build the minimum to learn and then iterate.
Different types of products call for different MVP strategies:
Physical Product MVP: If you’re making hardware or a tangible good, build a quick prototype or even a mock-up. You might 3D-print a basic model or assemble a rough version by hand. The goal is to get something in a user’s hands so you can see if they use and like it. This could be a simple product sample, or even just a handcrafted model. The point is to test the concept before expensive production.
Software MVP: For tech products, you often start with wireframes, mockups, or videos instead of full code. A classic example is Dropbox: rather than build the entire syncing system, Dropbox’s founder made a short explainer video demo and shared it online. That video showed the core functionality and value. It immediately validated demand without any backend. Another approach is to create a simple landing page or limited-function web version. Buffer did this by launching a two-page site explaining their social-media scheduling app. They even added a fake pricing page to test if people would pay. These low-effort versions let you learn quickly.
Service MVP (Concierge MVP): If you’re offering a service (or a complex product), do it manually at first. In a “concierge” MVP, you deliver the service by hand behind the scenes. For example, Zappos tested online shoe retail by first posting photos of shoes from local stores on a barebones website. When customers ordered, the founder actually went to buy the shoes and shipped them himself. This proved people would buy shoes online before Zappos ever held inventory. Similarly, Groupon started by manually sending PDFs of deals to subscribers. These manual steps validated the business idea with minimal tech. Once demand was proven, founders automated the processes.
Each MVP type follows the same principle: reduce waste by avoiding building unneeded features. First test if people want the solution; then improve it. As Ries says, focus on validated learning.