Doing a hardware startup is, well, hard

I am not an engineer, neither have I any experience in manufacturing. When I first started what eventually became Berrolia, I entered a whole new world of making goods. This world turned out to be far more surprising than I ever expected. Being an economist, I belong to this group of people for whom actually „making” something means:

• creating an Excel spreadsheet
• coming up with a new bright idea and creating a stunning PowerPoint presentation
• writing thousand lines of code
• ..and so on and so on.

Actually, we never do anything practical.

The list of professions that do not know how to make things is long and includes teachers, programmers, accountants, marketing guys. And, of course, economists. With such a background, believe me, creating a hardware startup can be a challenging experience. I’d like to share with you some of the lessons that I learned about producing things. Again, I have no manufacturing background.

If you belong to one of these groups, read on. If not, and you actually know how to do something useful, read on too, and enjoy your growing ego.

When you look at a, say, can opener, it looks like a simple thing. You just do not see the army of people that look after materials, processing, quality control, packaging and shipping. If you are trying to create a new gadget, these are all areas that you will have to take care of. I am not saying that it is impossible. You just need to be aware that there is a lot to be learned before you are ready to sell your first 50 pieces.

With Berrolia, from the very beginning I’ve had a pretty good idea what the end product should look like. I created a number of prototypes in a modelling program and had them printed on a 3D printer. At that point, I thought that I am just millimetres away from launching the product to the market. But it turned out that I was merely at the beginning of the road.

1) Learn to talk to the manufacturing people. They have their own language

After the 3D-printed prototypes were ready, I contacted a local machine workshop and asked if they could cut (or „machine”) something similar from aluminum. To my surprise, they did not take me seriously. I was sure that, since I am potentially a paying customer, they would treat me like in any regular business relation. Only after several unsuccessful approaches to other workshops, I realised that I just did not speak their language. I approached from the end customer’s point of view, keeping in mind what I want the product to look like.

But they had in mind their machines and technologies. They approached from a different angle: what they can do with their machinery, at the same time keeping cost and time at reasonable level. At last, this is what engineers are educated to do. It took long time before I slowly started talking their language so that we could understand each other. Of course I had to re-design everything so that the product could actually be manufactured using the technologies that were available locally. Of course, I could make use of one of companies that boast that they can take you from „concept to end product” and source everything in China. I did not and I think that it was a good choice. I promise to write more about it in another post.

When I learned to talk to the manufacturing guys, it turned out to be one of my most interesting experiences. Most of these people are very friendly and informative. I do not mean that they are all nice. But if you want to do something or solve a technical problem, chances are that they will give you their time at no cost, because these guys are passionate about solving problems. If you give them your bad solution, they will spare you half an hour, sit down with you and give three good solutions, just out of pure pleasure to discuss technicalities.

2) Learn the key aspects of manufacturing. Guys in the factory will be more likely to talk to you

I had to learn more about CNC machining to get an idea of what can be easily done and what can not. There are shapes that can be easily (this also means – inexpensively) cut, while others require specialised custom tooling or waste a lot of material. My advice – talk to people that will be manufacturing your gadget. Learn as much as you can about the machines they are using, their jargon, their technical drawings, finishes, paints and hundreds of other things. Watch videos on YouTube about manufacturing. Robert Grzesek has a cool webpage (http://www.grzsoftware.com) with a lot of information about digitally-controlled machining. I enrolled to his free online course about programming the CNC machines. I did not become a qualified CNC programmer, but now I can imagine what can be machined easily and adjust the designs respectively.

Believe me, it will pay off.

3) Keep in mind that prototype is something entirely different from a scalable production

The first contacts with manufacturers dated back to early 2014. By the fall of that year I had some working prototypes manufactured using techniques that could potentially be repeatable. Again, I thought I was just millimetres away from a scalable business. I thought that since a workshop can do two pieces, it can do two thousands as well.

Well, wrong again.

It is another pitfall that can be potentially deadly to any hardware startup. Producing two prototypes is usually done with extra care, by the most skilled employees. A scale production can be a totally different story. You cannot just walk in and say „Hey, I want 1998 more of these”. Believe me, each piece will be different and you can throw 50% of them out of the window.

4) Learn to set quality requirements

Production of more than 20 pieces is supposed to be managed by quality requirements. If you do not set any, chances are that you will get a lot of useless pieces and waste a lot of money. On the other hand, it is also likely that you’ll have no idea how to set the quality standards, since you do not know yet, what can possibly go wrong.

Any solution?

I chose a path to order 20 pieces first, then 200, then 500, and each time check, what went wrong. This was a costly process, but it did work. Each time errors allowed us to discuss what went wrong and set the quality standards to our suppliers. Anand Srinivasan shares some valuable thoughts on the importance of taking baby steps at startup phase.

5) You need more time than you think to manufacture your big invention

Eventually, spring 2015 was the first time when I had first 200 pieces ready for sale. I cannot say if it was good timing, or maybe I was super-slow. One year from idea to semi-scalable production can be seen as a very long time for some, but we really needed that time to learn about the production process. Maybe at that point was the right time to launch a Kickstarter campaign and use these 200 pieces as „early bird” rewards? Possibly. However, I felt that there were still many loose ends – web page, product photos, payment systems, taxes, marketing and so on. So we decided to launch slow and get early feedback from customers.

Did I say that a hardware startup was easy?

If you are like me and you have no background in manufacturing, and just invented the „next big thing”, my suggestion is that you take your time. Do not set hard time constraints on yourself, because chances are that you will do a premature launch and fail. Try to manufacture small numbers first – you will learn a lot. Launching a Kickstarter campaign before actually manufacturing a small lot of your gadgets can be a waste – it will cost you nerves and chances are that you will let down your supporters. Jonathan Albert from Bresslergroup shared useful thoughts about launching the Kickstarter campaign prematurely.

From my experience, there was a lot more to learn about manufacturing than I thought. And job experience with Excel worksheets and big nice yearly reports did not prepare me for that.