I take it as a given that when a client approaches us with a new embedded product idea, they will require a very demanding set of features and a minimal price tag. The “minimal price tag” part always applies to the development effort required. For products with a hardware component, it also applies to the product’s cost of goods sold (COGS).
Companies developing high volume consumer products can afford to spend quite a bit of engineering effort to reduce their COGS, since the development costs will be amortized over so many units. However, many of our clients sell low-to-medium volume products—in the range of 1,000 to 10,000 units per year. This volume is not high enough to leverage massive economies of scale, yet a clever design team can still create a very successful and cost effective product if some care is taken during the design phase.
There’s no magic bullet to reduce COGS in the low-to-mid volume range; it’s mostly common sense, coupled with the experience that comes from having built products in this volume range before. Here are some ideas for minimizing COGS:
Stocking distributors can help reduce both product costs and development costs. Distributors have much higher leverage with IC suppliers than a small company, which can help reduce lead times and solve supply problems. One big advantage is that distributors often give forward pricing or high volume pricing (even at low volumes) that can significantly lower your product’s cost. The big ones (Avnet and Arrow) also have dedicated resources that can suggest technology solutions you might not be aware of. They can also facilitate deals with small 3rd party subsystem vendors, such as display manufacturers. As with any vendor, you must negotiate with these companies in order to see the best prices. They are in business to make a profit, but they can also act as a facilitator in negotiating with the end suppliers.
IC vendors always have new products in the pipeline. These are usually lower cost and higher performance than existing product. Of course there is risk in using a new product, especially if there is a lot of software support required. You need a close vendor relationship in order to successfully integrate a new product.
Many high volume consumer electronics ICs are custom designed for a certain customer or retail market. Large IC vendors such as Texas Instruments and Intel frequently offer embedded products that are derived from their consumer products. These are usually low cost, but very high performance products that are guaranteed to be available for many years. One example is the Intel Atom family. The consumer versions of the Atom chipset can disappear at a moments notice, but the embedded versions will be around for at least 7 years after release.
Many component suppliers have wonderful parts that would be a perfect fit for your product, but not all of these parts are actually available in small purchase quantities. Your best chance of a sustainable supply chain is to use parts that are currently stocked. Maxim is a company that is notorious for having lots of great parts in theory, but a much smaller subset of parts that are actually regularly manufactured and available for small purchase quantities. My colleague Todd refers to these types of parts as being made out of “Unobtainium”.
This is a no-brainer, but one that engineering teams are not always comfortable with. Competition between suppliers is one of the best ways to reduce product costs. A lower cost competitive bid is your best negotiating position.
Second sourcing components or subsystem modules creates a permanent negotiating position with your suppliers. A great example is to use a single board computer in a open form factor, such as COM Express. Many vendors supply these modules, with lower cost versions coming out in the future.
Concentrate your efforts on the highest cost components. Unlike high volume products, it may not be worth the overhead cost to squeeze out the pennies.
If you don’t manufacture your own products, you will need to select a CM; this is an area where “bigger” definitely doesn’t mean “better”. It may be difficult to get and keep the attention of a large CM. It is more likely that a large CM will slip your deadlines in favor of their larger customers. You will be able to push on a small CM when you need to.
These are some of the guidelines we use to minimize costs for low to mid volume products. Part 2 of this article will cover minimizing development costs while still producing a cost-effective product.