Ten years of RAAK: what changed in product development since 2016
We started in Leuven in 2016 with two designers and no clients. A short, honest list of what is different about making hardware now, and what is not.
Kasper and I started RAAK in 2016 in Leuven, after working in London and New York. Ten years on there are teams in Brooklyn and Shanghai, more than 110 launched products behind us and a network of over 55 manufacturing partners. This is not a retrospective. It is a list of what has changed in how physical products get made, from where we sit.
What changed
Prototyping got cheap and fast. In 2016 a good-looking prototype was a line item with a lead time. Now a desktop printer gives us a form to hold the next morning, and a service bureau gives us a near-production part in a week. The effect is not just speed. We test more ideas physically, earlier, and we argue less about renders.
Electronics became the default. Most products we touch now have a battery, a sensor or a radio. That changed who is in the room: an electronics partner from the first week, and certification planning from the discovery sprint, not from the end.
China stopped being only a place to make things. Our Shanghai team started as manufacturing support. It is now also where some of our clients' markets are, and where some of the best tooling and finishing suppliers we know are. The direction of the relationship is no longer one way.
Sustainability became a specification. Ten years ago it was a value in a presentation. Now it is a material choice, a repairability requirement, a packaging constraint and, increasingly, a regulation. It is better this way. Specifications can be designed to.
Clients got more technical. More founders come to us with a working prototype and a clear idea of the electronics. The brief is further along, and the question is less "what should this be?" and more "how do we make ten thousand of it?".
Software ate the interface. Buttons, dials and displays have been replaced by apps on many products. That removed work from industrial design in one place and added it in another: the few physical controls left have to be exactly right.
What did not change
Tooling still costs what it costs. A steel injection tool is still the biggest cheque in most projects, and it is still cut once. All the speed before it does not change the care needed at that moment.
The pilot batch still decides the launch. Whatever the renders promised, the first few hundred units from the line are the product.
A brief is still the best investment. The projects that went well in 2016 had a clear owner and a clear problem. So do the ones that go well now.
Factories still reward visits. Email and video calls have improved, and a factory still works harder for people who have stood on its floor.
What we would tell ourselves in 2016
Start the manufacturing conversation on day one. Price the phases separately. Say no to projects with no decision-maker. Go to China sooner. Hire the engineer earlier. Most of it we learned the slow way, which is why these articles exist.