Exploring the Andreoy Approach to Better Product Design
Over the past decade I have watched countless product teams struggle with the same question: how do you build something that people actually want to use? The usual answer involves more user research, more prototyping, more testing. But the real problem is often not a lack of effort. It is a lack of structure. When teams do not have a clear framework for moving from insight to execution, they end up with a grab bag of features that satisfy nobody. That is where the concept of andreoy comes into view.
I first encountered andreoy while working with a small hardware startup in Oslo. They had spent six months refining a sensor module for industrial monitoring, and the prototype worked perfectly in the lab. But out in the field the device kept failing in ways that seemed random. The engineers blamed temperature swings. The product manager blamed the factory. Neither was wrong, but neither had a way to connect the two problems. Then one of the senior designers introduced a method she called the andreoy loop. It was not a formal methodology at the time, just a way to force the team to look at the full system instead of isolated parts.
That experience stuck with me. Since then I have applied the same thinking across digital products, physical goods, and even service design. The core insight is simple: every product exists inside a web of constraints, and the best designs emerge when you treat those constraints as raw material rather than obstacles. Andreoy is not a brand or a piece of software. It is a mindset. But it is a mindset that can be taught, practiced, and measured.
Why Most Design Processes Miss the Real Problem
Standard design processes tend to follow a linear path: research, ideate, prototype, test, launch. That sequence works fine when the problem is well understood and the environment is stable. But most real-world problems are not like that. User needs shift. Materials behave unpredictably. Budgets shrink. Team members leave. The linear model assumes you can collect all the important information upfront, and that assumption is almost always wrong.
I remember a project where we were building a booking platform for small clinics. We spent weeks interviewing doctors and receptionists, mapping workflows, building personas. The prototype tested beautifully. Then we deployed it in three pilot clinics, and within a month two of them had stopped using it altogether. The reason had nothing to do with the interface. It turned out that the clinics shared a single internet connection with the pharmacy next door, and the pharmacy's system hogged all the bandwidth during peak hours. No amount of UI polish could fix that. We had optimized for the wrong constraint.
The andreoy framework forces you to ask a different question early on. Instead of asking what users need, you ask what conditions must be true for the product to work. That shift changes everything. You stop guessing and start testing the actual environment where the product will live.
The Three Layers of Andreoy
Over time I have broken the framework into three layers that teams can use as a checklist. They are not steps in a linear process. They are perspectives you revisit throughout the project.
Layer one: Physical and digital constraints
This is the most concrete layer. It covers the hardware, network, power, sensors, screens, and anything else that the product touches directly. When you work on a mobile app, the physical constraint might be the battery life of the phone or the size of the screen. When you work on a medical device, it might be the sterilization temperature or the reading distance for a display. Write down every constraint you can find, even the ones that seem trivial. Then rank them by how hard they are to change. The hardest ones will shape your design the most.
Layer two: Human and organizational constraints
This layer covers the people who build, buy, and use the product. It includes things like team skill sets, budget cycles, regulatory approvals, and user habits. A common mistake is to treat these constraints as negotiable. They are not. If your team does not know Python, do not plan a Python-based backend. If your users check their email only on desktop, do not build a mobile-first app. The human layer is often the messiest because it involves emotions and politics. But ignoring it is a recipe for failure.
Layer three: Temporal and market constraints
This layer covers timing, competition, and lifecycle. When do you need to ship? What are competitors doing right now? How long will the technology stay relevant? These constraints are easy to overlook in the early stages because they feel distant. But they determine whether your product survives long enough to matter. I have seen great products die simply because they arrived six months too late or because the market had already moved to a different standard.
The beauty of the three-layer model is that it gives you a shared language. When the engineer says the sensor cannot handle that temperature range, that is a layer-one constraint. When the sales director says the pricing has to be under fifty dollars, that is a layer-two constraint. When the CEO says the product must launch before the trade show in June, that is a layer-three constraint. Andreoy makes these conversations explicit rather than implicit.
A Concrete Example from the Field
A few years ago I consulted for a company that made portable air quality monitors for construction sites. Their first version had a sleek design and accurate sensors. But it broke constantly. Dust got into the casing. Workers dropped it from scaffolding. The battery died mid-shift. The team had focused entirely on accuracy and ignored the physical abuse the device would endure.
We ran the three-layer analysis. Layer one revealed that the casing needed IP68 protection and the battery needed to last two full shifts. Layer two showed that the workers had thick gloves and could not operate small buttons, so the interface had to work with a single large press. Layer three showed that a new regulation was coming in nine months that would require real-time data transmission, so the device needed built-in cellular connectivity from the start. The team redesigned the whole product around those constraints. The second version was bulkier and less elegant, but it survived on site and the company won the contract.
That is the trade-off. Andreoy does not promise beautiful design. It promises design that works under real conditions. For many products, that is far more valuable.
How to Start Using Andreoy Tomorrow
If you want to try the framework on your next project, here is a practical way to begin. Gather your team for a two-hour session. Bring a whiteboard or a shared document. List every constraint you can think of for the product, from the obvious to the absurd. Do not filter yet. After you have a long list, group the items into the three layers. Then vote on the top three constraints in each layer that you think will cause the most trouble if ignored. Those nine constraints become your design brief for the next sprint.
You will be tempted to skip this exercise when the deadline is tight. Do not. The time you spend upfront identifying constraints is almost always less than the time you spend later fixing failures that could have been predicted. I have seen teams cut their prototyping cycles in half simply because they stopped building for a fantasy environment and started building for the real one.
Andreoy is not a magic bullet. It does not replace good judgment or domain expertise. But it gives those qualities a structure to work within. And in a world where products are getting more complex and timelines are getting shorter, structure is exactly what most teams need.
