2014: The year of hardware?


A few days ago I attended the Hardware Workshop in San Francisco. It was a great way to speed up the learning curve by looking at what others have done and the obstacles they encounter. The simple truth I learned is that hardware is still much much difficult than software and that is hard to change unless technologies like 3D printing becomes pervasive. Also, I saw a world that considers software as a necessity rather than primordial need to differentiate and compete. From my point of view, the need for hardware will only increase but the fact is that software is indeed, eating the world.

This is a summary of what I learned. I am certain I missed something but as soon as Marc Barros, the guy behind this great program, sends us the presentations, I will post them in here –if he approves, of course.

Customer interviews matter.

The customer interviews can show you the problem from a different point of view. Maybe the solution you are thinking about is not a good match to the problem. Make sure to understand the problem well. For example, Lumo thought they were solving the problem of back pain but when they talked to customers they understood they actually knew they had a problem and (somewhat) how to solve it. They just needed more help. They wanted more resolution. The problem turned into solving “posture” for many reasons beyond back pain.

When you are asking the questions, make sure to start as general as possible. Don’t make assumptions about customers’ habits. Let them tell you. The questions will help you to avoid being a technology company chasing a problem. The solution should probably be as technology agnostic as possible. Getting users to give you feedback using Mechanical Turk, Task Rabbit, Craigslist and alumni network and others is key.

Describe, absorb and analyze the customer journey:

The customer journey has a beginning, during and after. Make sure to understand the issues the customers are trying to solve are aligned with the solution you are trying to build. Avoid premature optimization of your solution as you might start developing an overly complicated solution. That could be distracting. Lumo vs. Contour illustrate two different problems with two different approaches. Lumo was a brand new solution to an existing problem that had not been solved efficiently. Contour was a new camera that replaced solutions that users hacked on their own. Both products started as a efficient homemade “hacks.” Countour lost to more efficient and focused rivals. Lumo is succeeding but it is too early to declare victory.  As an interesting fact, Lumo’s the three founders did not have hardware experience so they found people that can help them. Interns from Stanford, advisors from design firms like IDEO.

Early stage venture fund for hardware startups –– Eric Klein from Lemnos labs:

I liked Eric’s presentation. He was down to earth and very straightforward. He had been at Sun Microsystems and that part of his experience shines through. There’s a very specific way in which former Sun Microsystems engineers tackle problems. I saw that over the four years I spent at the company.

How does the product create a revenue generating relationship with the customer. Your first product is the start of an ongoing mutually valuable relationship. For a successful development of a hardware product you can look at cost, features, quality and time. Thing is, you can lock 1 or 2 variables out of the four. There’s no such thing as a perfect product. Also, you have to able to define your customer at a high resolution. Avoid thinking that the crowdsourced customer is your target customer. Think about HHI (household income), age, ZIP codes and DMAs.

The Rise and Fall of Flip by Andre Neumann-Loreck

I never understood the appeal of the Flip. Maybe I was never the target market but I saw a low end camera with a USB cord attached to it. For me, that was interesting but not worth buying. It was so hot that a year later after launch, I decided to buy one, but by then the offering was so complicated with all the different types of Flips that I opted out. It was a great success nonetheless and below you can find a few notes from the talk that Andre gave at the Hardware Workshop.

Flip chose to go to the UK instead of going to China. The situation was that the Chinese market supported a high end video camera with all the features that they needed to capture their special moments. Flip fell somewhere in between the low and high end cameras. The UK was a better fit for Flip. The value proposition resonated more there and made it one of the best markets for Flip. For Flip the UK was 10% of the US revenue. France, Germany, Spain, Italy added up to another 10%, but why? In the UK, the US product has no import stigma. Also, it is probably easier to distribute in the UK. If you’re planning to go international, then think about the repercussions of your bottom line and operations.

How much should I charge for my product?

This is an example for consumer products. First, it has nothing to do with your COGS. Simply assume that you are going to get the “volume” pricing. Also, it doesn’t matter if you get to volume either. The perception of the consumer matters. For MSRP not ASP (average selling price)your pricing should be usually 3x or 4x of COGS. COGS need to include all operational and overheard costs. Distributors usually will cost between 5% to 10%.

How much do retailers vs. online retailers charge me?

50% is normal. If customer spends  $100 MSRP then your ASP is $50. For Amazon, 20% to 25% is normal. Some retailers give you MDF (marketing development funds) but that really is taking away from your margin. Retailers usually pay 30-60 days after product is received while you have to pay in full for production batches. Also, marketing is a real expense so you should be planning for 15% to 25% of revenue in the launch year.

Building ugly early, by Flextronics.

Innovation: Something different that has impact. The key is to think about the future needs of the user not the present problems. For that you have to find the fastest most efficient way to go to market.

Sketch, score, prototype.

The first step is to build an ugly sketch of what you’d like to build. When you do that you find early issues that can be fixed and this helps you make better decisions in terms of what to keep and what to throw away. Now you move onto an ugly prototype using materials that you can find at Staples or Fry’s. Tools that you can user are: strata sys, protomold, CAD, Fineline, makerbot, etc.

A picture is worth a thousand words. A model is worth a thousand pictures. You can do a 2X, 4x or 10X model if necessary.

Where to build your hardware product?

China is a low cost labor base. It goes up 20% year over year but it is still very affordable. The supply chain is tightly integrated as opposed to the US where it is more distributed. They have tremendous depth of manufacturing knowledge. The infrastructure is amazing. The distance, language and culture could be considered as negatives.  If you’re building less than 5,000 then it probably doesn’t make sense to go to China. Manufacturers are interested in volume not in your groundbreaking ideas.

The Lean Hardware Startup, by Cyril Ebersweiler from HAXLR8R

Life of a hardware startup:

Concept, Minimum Functional Prototype (MFP), Complete Functional Prototype (CFP), design to manufacture (DFM), First factory run (FFR) and retail. Remember that no hardware plan survives contact with factory.

Concept phase. Become the master of the problem. Talk to customers. While doing this, bootstrap with your own money because fundraising is almost impossible. Many investors want to see something that works.

MFP –– Finally works. Off-the-shelf, 3D prints, handmade parts. Your BOM is in the tens of thousands of dollars. Some cash is coming in from FFF but people will think it’s the final product. Be direct and tell them that at this point they are highly likely to lose their investment.

CFP — You’ve done a prototype that looks better. You’re ready to talk to suppliers. You’re still at the angel, bootstrapping level. Maybe join an accelerator. VCs are still far away.

DFM — You are ready to roll and to crowd fund. Identify suppliers, negotiate, reduce fees. Master your BOM and your profits. You’re still at the angel, bootstrapping level. You’ve joined an accelerator (or not). VCs are closer but still far from reach.

FFR — Making! You’re still at the angel, bootstrapping level. Maybe join an accelerator. VCs are closer than ever but they will need pre-orders to get involved.

Closing thoughts: A category leader requires a lot of capital. Pick your battle: single hit, portfolio, platform or ecosystem. In hardware, success is not binary.

Creating the right roadmap and what’s after the MVP?

Ariel, the co-founder of Flip Video. Simplicity and breaking away from the common design philosophy of packing in features. When building a roadmap, there are no rules. There are some considerations to keep in mind. Your roadmap is your strategy.