1.1.30. Creating a Prototype: Physical Goods

Entrepreneurs often come to me and say, I’ve got an idea. I just need someone to build a prototype. Usually that masks a deep misunderstanding of what it takes to get a hardware product to market. And it views a Prototype as a single endpoint rather than as a process. I’m going to use a bunch of examples to illustrate key principles of prototyping. And I’m going to close with a discussion of working with suppliers and factories to get Prototypes built. Let me first make a distinction between Focused Prototypes and Comprehensive Prototypes.
0:35
A focused prototype reflects just one or a few dimensions of performance of the product that you’re developing. These usually are referred to as say, test rigs or mock-ups.
0:47
A Comprehensive Prototype is a fully functioning prototype that usually goes through successive refinement and often goes by the term, Proof of concept Prototype, Alpha Prototype, Beta Prototype, Pre-production Prototype. Let me give you some examples. I worked to develop a new cordless nailer as part of a new venture. And one of the concepts we were considering was an electric powered cordless nailer. One of the key questions the development team faced was, how heavy would the tool be, and how would it feel balanced or not?
1:26
We built this prototype to answer that question. This is a Focused Prototype. It’s made of plywood with steel rods inserted throughout to give it the right weight and the right center of gravity. This of course can’t be used to drive a nail, so it only mirrors one attribute of the product, which is what is its size and weight. But we can use it to answer the question of does this tool feel light enough and how does it feel in terms of the moment of inertia and balance? This is an example of a Focused Prototype.
2:03
I developed a product called the cross rack bicycle rack. The big idea behind the cross rack was that it would put the cargo right behind the rider. And it would do so by attaching to the seat post. This was the very first prototype that I built. It used an old stem from a bicycle. To attach to the seat post. And then I made this in my garage just from bent steel tubing. This I called the proof. This is called a proof of Concept Prototype. I was actually able to use this on my bicycle and to verify that the concept had validity. This is a comprehensive proof of concept prototype. This was the second prototype which I called the alpha prototype. And this addressed a couple of concerns I had with the proof of concept prototype. The first was that the rack wasn’t big enough and wouldn’t accommodate a standard bicycle bag and so I made this larger. The second was to accommodate the feedback that it would be nice if I could remove the rack from the bicycle. And so I adapted.
3:12
Another part from an existing bicycle rack with another stem from a bicycle in sort of a cobbled together way, in order to get that complete functionality. This was an Alpha Comprehensive Prototype.
3:27
This was the Beta Prototype for the cross rack. It has all of the features and functionality of the eventual product but it was actually made by the factory that would eventually make the product. And I was able to actually give this to consumers to test. That’s often the characteristics of a beta prototype. So, this is a Beta Comprehensive Beta Prototype but it still had a few issues, in particular the factory got the finish wrong, so this had a glossy finish and this had a matte finish and I also recognize that I really wanted there to be a little bit of an angle between the way this clamped and the angle of the support structure.
4:04
This is the Pre-production Prototype for the cross rack, and this is in every respect the production intent design of the cross rack. You can see it even has the logo on it. It’s got the proper angle. The surface finishes are all exactly the same. In fact, this could have been sold to a consumer, and this is an example of a Pre-production Prototype. All of those prototypes are examples of Comprehensive Prototypes. They’re fully functioning but they typically go through cycles that help the team achieve milestone and demonstrate progress and to workout challenges in the design and development of the product.
4:43
What are the motives for building prototypes? Why do you build prototypes? Well for you, for the development team itself. You’re building prototypes primarily to learn, to solve problems and answer questions about do you actually have so called product market fit? That is, does your solution actually work to adjust the needs of your target customer? What isn’t so obvious is that you also build prototypes for an external audience. The external audience, for instance, might include suppliers, investors and partners. The questions or issues that you’re typically addressing for an external audience are proving that your solution works, technically. Proving that the dogs will eat the dog food. That’s slang for proving that customers will actually adopt your solution. Those prototypes are very useful for communicating what it is you’re doing, and for demonstrating that you as a team can implement. That this isn’t just an idea, but you’re actually serious about implementing the product. Let me give you a few examples. Shown here is a furniture product. It’s actually a media stand and the production version is shown on the right. But the one on the left was made with just found objects bought a home improvement store. That was meant to prove that it would accommodate the different elements of the cable box, the DVD player, the remote, those sort of things, and would fit in the room. It was probably made for
6:19
about $10 US worth of materials, and probably in a couple of hours. A Quick Prototype to answer questions and to learn. Does it fit in the room? Does it accommodate the function that it’s intended for? This is a prototype that illustrates just how simple a prototype can actually be. This was for a cosmetic Concept that I was working on with a cofounder. The company was called Blend8. And we just wanted to be able to show to others just a representation of what our cosmetic product might look like and how the brand would play out. And so we just mocked up a label with a laser printer, stuck it on a generic Jar and used it in demonstrations. It’s very effective as a way to communicate that we are serious about this product and the product concept. This is a team of my students in my product design class and they had the idea for a clothes iron that was integrated with a glove and it’s shown there by the woman there in the red sweater, she’s wearing it on her right hand. And the idea was that after you had your clothes on if you had a little wrinkle you could sort of reach with your iron glove and take care of it. Now, that may not sound like a great idea to you but in fact one of the great things about they’re building this prototype was they could answer that question, does that work, are you burned, can you do it at the right temperatures, how does it feel, how does the customer feel about this idea and I show you this example because none of this four students really had the technical background. To build a new iron but they were able to put together a markup or a demo using found objects. It was actually quite effective in demonstrating this product concept.
8:07
I want to show you a good trick for prototype in challenging products where it’s very difficult to build a Comprehensive Prototype that both works like and looks like the production-intent product. This is a prototype that is functional for a nailer concept that we were considering In fact, the concept here is to use a small explosive charge to drive a piston that then drives the nail. And it integrates the existing nail magazine from an existing product with some new hardware that we designed to prove out that explosive piston concept. This doesn’t really look like a product. It’s a works like prototype, but it’s not a looks like prototype. If I were to show this to investors, I don’t think they’d be too impressed that this looks like a product, and yet it would be quite effective in demonstrating that the core principle behind the product works. So what we did was to develop a separate looks like prototype. Let me show that to you. This is the looks-like prototype for that same concepts. We were able to illustrate where the magazine,
9:20
the strip of explosive charges would be stored. We are able to illustrate that it could have two handle positions depending on how it would be used, and we were able to illustrate the basic aesthetic size and shape and ergonomics, all with a very simple foam prototype that was built by a model maker in a few hours. This, of course, doesn’t have to work, but it is quite effective in illustrating to outsiders what the eventual product would look like. So a very nice trick that you can play is to use two different prototypes. It looks like and it works like In tandem to illustrate what the eventual product will be. I’d now like to walk you through the evolution of a product from the original prototype all the way through to production and talk about some of the issues that arise as you go through the process. Shown here are the eight prototypes that I developed to test original solution concepts, and they embody many different approaches to the design of an ice cream scoop, and they’re mostly made of found objects. So if you look at the one that’s third from the left, it’s actually made from a tomato paste can and the end of a baseball bat. And it was made in probably 20 or 30 minutes in the shop using just conventional shop tools. Some of the others you can see there are made by taking existing products and cutting them up and reconfiguring them in order to change the geometry. So this gives you a sense of the kinds of techniques you can use just to explore solution concepts. Eventually I decided to pursue the concept that embodied an angled scoop in which the scoop head is angled relative to the handle.
11:10
One of the key questions was what the ergonomics and form of that might be. And so, I built a model out of a material called balsa foam, which is just a foam material that can be easily sanded and shaped into the desired eventual handle and scoop shape. And that’s shown here. The problem, of course, is that this is just a solid block of foam and can’t be used to scoop ice cream. So I had to figure out how to make a working prototype from this basic approach. The way I did that was to create some drawings, that were dimensioned based on the model. Shown here is one of two drawings that I created that were based just on measurements from that foam model, that I was quite happy with. I then hired a graduate student here at the University of Pennsylvania to create for me a computer model from those dimensions. That was a skill I actually don’t have personally. And so I just hired someone for a modest amount of money to be able to create a geometric model based on my sketch. Shown here on the left is the geometric model, or an image of the geometric model that he created in the computer. And once you have that there are many suppliers who can create Physical models using a technology called 3D printing. So, on the right, you can see the 3D printed artifact that came about from the computer model that was created from my hand sketch. The 3D printed model on the right is one that I could actually use to scoop ice cream, and could verify that. In fact, the handle shape and the angled scoop were quite effective. What came next was a series of iterative refinements from the original foam model shown on the far left all the way through to the eventual production design shown on the far right. I show you this to invoke a little bit of humility about what it really takes to build a hardware product. Even for a very simple product like a single-part ice cream scoop. We go here through dozens of iterations, each one of which prototyped in order to verify the approach. What we eventually end up with on the far right became the production version of the scoop, which is shown here, sold by the company Belle Vie. This product, if you look at it now, looks quite beautiful, looks highly refined, looks like it was always meant to be. But what I want to do is to show you that behind that experience, behind that result is a lot of iteration of dozens of product variations to eventually come up with That final product. So, when people say they ned someone to help them build a prototype, often, what that masks is a misunderstanding of what it actually takes to build even the simplest hardware products, which is, it takes a lot of iteration, a lot of trial. And a lot of refinement to get from original proof of concept through to the final production intent. Let me also say that to really do hardware products at a very high level of professional expertise that allows you to be successful as an entrepreneur. You absolutely have to have the expertise on the team that’s appropriate for the task at hand. So while I was adequately skilled to come up with the original solution concept of that angled scoop, I don’t have, and I didn’t have, the sophisticated industrial design skills to create that beautiful aesthetic final product. For that, I brought in a partner, in this case a design firm, Lunar Design, to help me to build that additional level of refinement.
14:57
I want to turn now to the question of how to work with suppliers. And I want to give you just two examples that represent to two basic approaches that you take to work in with suppliers to produce prototypes and ultimately to producing your final product. I’m shown here at a factory in Taiwan that make die cast parts, that is metal parts that are cast by molding machines. And I’m shown here with the plant manager in the middle and with my agent on the right. And in this case, the model is that I provided to the factory a fully described detailed drawing and set of computer models for exactly the part that I wanted them to produce. They then produce that part exactly to my specifications, they don’t even have to know what the part does or what it’s for. They’re simply working from a detailed technical specification usually in a form of a drawing that fully specifies the geometry of the hardware that I want them to create. Now, the good news is that when you do that, that means they have no proprietary information other than the drawing. They don’t even know what you’re doing, necessarily. The bad news is that you have to do the hard work of actually creating all the details of the design before they can produce it.
16:21
There’s an alternative model called the ODM model, which stands for original design manufacturer.
16:28
I’m shown here meeting with an ODM in mainland China. And in this case this is a motor manufacturer. And in this case, I went to the motor manufacturer and I said, I need a motor that does this, and is about this big, and has this kind of torque, and uses this much power, and that hopefully I can buy for about $15. And then the factory goes off and actually designs that for me, and produces a product to my functional specification. Now that, of course, is really nice because I don’t have to do the work of designing that product. But let’s just review the basic strategies and approaches you can take to working with suppliers for prototypes. In one approach, you work with an original design manufacturer. In that case, the factory has the know-how. Their staff created a design based on your requirements, and the advantage is that they often know how to do that in a way that’s very efficient to produce. And the design work they do is often free. And I guess by free, I mean it’s bundled into the cost of procuring the parts from them. The disadvantage is that you have nothing proprietary at this point. Anyone could go to that factory and give them the requirements and they would produce it to those requirements. The alternative is that a factory that produces to your proprietary design where you do all the design work and where you provide the details of what it is you want them to produce. And the advantages and disadvantages are exactly complementary to the ODM. That is, their staff doesn’t do the work. That means the product might not be efficient for them to produce. And you have to pay to get the design work done, but the advantage is that you own the design, that you keep it completely proprietary. They don’t have any of the proprietary information that goes into the product. So I guess is you think about which approach to take, you ought to think about the source of your enterprise’s proprietary advantage. Where does your unique advantage come from? Do you have something that’s unique and proprietary in terms of technology? If you do, then you really have to own your own design. You have to do the design work required. Bringing to the task partners or experts that you either get to join your team or that you contract with very closely. And then you provide the detailed specifications to the factory and they produce to your design. That strategy allows you to keep what you’re doing unique and proprietary. But it comes with it It comes at a cost. Meaning, you have to do the design work, you have to have the expertise. So those are the fundamental trade-offs you face as you think about working with a supplier in order to create prototypes and eventual production units for your enterprise.