A Primer on Product Development

(C) 2004 Hank Wallace

Have a great idea? Want someone to make it real so you can sell it? There are some things you need to know about the product development process that will save you money, time, and stress. As a person who has a life apart from product development, you are not expected to know these things. But you should be open to learning how a hundred thousand companies worldwide work the process. There are no secrets, but things may not happen as you would expect. Be careful because everyone has some subconscious idea of how products are developed and brought to market.

Let’s assume that you have already conducted market research and you know in general what the product will do and how much it should cost the end user.

The typical product development process is outlined below.

1. Write specifications. Create a written document including all the details of how the product works, the conditions under which it must function, the conditions under which it is allowed not to function, and all other conditions it may experience in the field. If you are not a writer, enlist our aid to help you out. Please don’t assume anything. State every function of your product in the greatest detail possible. This saves you money in the long run. Every weakness you miss in the specification process will be found by your customers!

2. Qualify the designers. For a consumer product, you’ll need:

  • Industrial designer — creates the form and feel of the product
  • Electrical engineer — designs the guts of the product and implements the functions in hardware and software.
  • Mechanical engineer — draws up detailed physical plans for manufacture of the product
  • Plastics engineer — designs the molded parts of the product for reasonable cost and manufacturability

For an industrial product with a utilitarian look, you may not need the industrial designer.

These people may exist in one company, or you may choose to contract their services separately. Some job functions overlap, so you may find many of these skills in one person or group. Ask for references and examples of previous designs. Usually if you find one of the above, he or she can refer you to the others. Most of these people advertise only sparsely, so you will have to take advantage of the colleague network.

3. Qualify manufacturers. You should also qualify the prospective manufacturer(s) at this point so the designers can know their capabilities and design to fit their processes. Each manufacturer specializes in certain products. Some make high volume consumer products, whereas some make only low volume (but more expensive) industrial equipment. Asian and Mexican manufacturers produce products at lower cost, but will shy away from low volume work and have preferred component vendors.

4. Review specifications. Before any costly design work is done, have each designer review and quote your specification (under a non-disclosure agreement). This can be done inexpensively, in only a few hours. The designers will have many comments and suggestions on reducing cost, minimizing complexity, and speeding the product to market. These people do this for a living, so listen to them closely.

If there are disagreements between the designers, listen to the arguments and make a decision according to you goals. Not all product decisions are black and white; most are difficult trade offs between attractive alternatives.

Be aware that once you give the designers the start signal, charges accumulate, and major changes in direction result in wasted money. Making design changes while the product exists only on paper is cheap.

5. Build prototypes. It is not possible to go from paper design to production units in one step. Prototypes of the product must be constructed to prove its function and engineering design worthiness. It takes some weeks to design the prototypes, and some weeks to build them, typically. Once you receive the prototypes, it is up to you (and your designers) to identify weaknesses and areas for improvement, as well as to evaluate the product’s form, fit, and function.

If you wink at the prototypes and are hasty to production, you will invariably waste time and money. Each new product has certain problems that must be identified through use, and then designed out.

6. Redesign. Incorporate the modifications you decided upon from your evaluation of the prototypes. Go to step five and build another set of prototypes to ensure your product functions as required. You must hold in your hand several working prototypes that are functionally exactly like the finished product before you go into production. Repeat steps 5 and 6 until this is so.

If you are detailed in your analysis of the prototypes and realistic in your expectations, you’ll only have to build a few prototypes before going to production. This is where most inventors quit; don’t lose your nerve. You can shorten this process greatly by writing comprehensive specifications and sticking to them. See step 1.

Some inventors quit here because they have some unrealistic expectations. If you must have a product with a brilliant backlit display that runs on batteries for ten years, well that’s not going to happen (in my lifetime). Compromise. If you are trying to change the laws of physics, it’s not going to happen. If you want to fit twice as much function into that $2 microprocessor chip, and it’s not big enough to do the job, you have to either increase product cost or decrease function. Make the decision.

7. Pre-Release. The design is released to production after you are satisfied the prototypes function correctly. However, don’t sign an order for 10,000 pieces if you have only seen five prototypes working. The rule of thumb is to ramp up in factors of 5 or 10. Build 25 or 100 pre-production units. Your contract manufacturer will do this at a much elevated piece price. Get these units into the hands of users as soon as possible, then follow up with a survey to get their reactions. Retrieve the units and look for physical or electronic problems. If problems are found, go to step six and fix them.

If your pre-release models function well, you should consider next building 250 to 1000 units, keeping a close eye on quality and defects. If there is a 10% incidence problem, it’s easier to fix 10% of 250 than 10% of 1000, or to just scrap them if the piece price is low.

8. Release. Once you see a batch that works well with an acceptable defect rate, increase production lot sizes as needed.

9. Monitor. Monitor each batch religiously. Buy a boring book on statistical sampling and quality control and learn to test each production batch to identify problems before sending them to customers. Or hire a boring engineer to do this for you, independent of your designers and manufacturer.

You should roll up your sleeves and test some units yourself, just to get a feel for the variation in look and feel, and to identify any problems that your manufacturer missed. It’s your money on the line, and no one cares as much as you.

Please understand the tradeoff between high volume, low cost manufacturing, and low volume, high cost manufacturing. If you can purchase a similar product at WalMart for $9.95, achieving such a low price for your product will entail manufacturing in Asia, lengthy tooling and qualification times, and a significant fraction of payment to the manufacturer up front before your get the first working product. Once you have a couple working products under your belt, the manufacturer may extend credit terms, but generally not on the first product.

Manufacturing at lower volumes drives up the cost per unit, but you can usually get the product from the manufacturer faster. However, it is not usually possible to start with a low volume product and smoothly ramp up to high volume production at the lowest per unit cost without some redesign.

The following table compares high and low volume manufacturing options.

Low unit cost Higher unit cost
Longer tooling delays Few tooling delays
Higher total tooling costs Lower total tooling costs
Lower per unit tooling costs Higher per unit tooling costs
Parts availability delays Parts available from stock
High risk in case of problems Low risk in case of problems
Large upfront investment Small upfront investment
Can satisfy large distributors Large distributors not interested

The threshold between low and high volume depends on the product. For consumer products, the threshold is between 1K and 10K units, maybe more. Low volume products may extend to 5K units per year. Industrial products are high volume over perhaps 1K units because they are generally higher in cost.

Your job as trail boss is to ride herd on your designers. Make lists of tasks, goals, and problems, and ensure each is addressed. Keep tabs on billings and talk with your designers frequently about what they are doing. Distributing timelines you draw with Microsoft Project is no substitute for getting weekly updates from each designer and re-projecting your completion dates. For every product, there is a number of hours, X, it will take to make it real in your customer’s hands. But since you cannot know X in advance, you have to keep making estimates along the way, correcting where needed.

If you have read this far, then you may have the perseverance, tenacity, grit, determination, and mettle to field a new product. It’s not easy, but it is rewarding, seeing a customer give your product, your baby, great reviews!

And if you have read this far, there are some other things you need to know, specifically about the engineers you will be working with. There is a personality and cultural difference between business people and engineers. Let me describe this briefly so you will know what to expect.

Business people are about business, which is about making money. They are schooled in business communications, finance, marketing, sales, and customer focus. These people are generally good communicators, and are interested in the bottom line.

Engineers have highly specialized training that allows them to create products (in the present context). These people think abstractly, are not always the best communicators, and can assume others think like they do when that’s not always the case. But they have the ability to make products that people want to buy, with proper guidance.

As a business person, if you have never worked with engineers, you have some surprises ahead. Let me illustrate with the classic “what he said” and “what they heard” format:

What the business person said What the engineer heard
I need this working in 4 weeks I need you to do the impossible
Our customers won’t tolerate software bugs I need you to do the impossible
The cost of this is twice what our customer will pay I need you to do the impossible
Didn’t you say that we’d have this feature? I don’t listen to you during project meetings
I don’t understand all this engineer-speak I don’t care how it works, just give me something to sell
What do you mean, “tooling costs”? I’ve never developed a product before
Why is the 100 lot price three times the 1000 lot price? I’ve never developed a product before
What the engineer said What the business person heard
Looks like my state machine has a bug Blah blah blah
We had to add a few more parts to solve that problem Our budget is shot
This will take about two weeks Our customers will buy from our competitor
Take a look at this flow chart Blah blah blah
This might not be ready in time for the trade show I’ve been reading engineering magazines in the men’s room
Making that part longer will require a total redesign I’m stonewalling
I can’t make the Internet faster More excuses

I’ve seen this a thousand times. What’s the solution?

The business person looks at engineering as a cost. The engineer looks at it as a benefit. There is truth to both points of view, and we need to understand each. The only solution is for each person to learn something about the other’s work, their goals, and their way of thinking. Engineers think abstractly, while business people think concretely, in terms of immediate goals. If you, as a business person, have never created anything with your hands, give it a try. Write a computer program. Buy an electronic kit from Radio Shack and build something. Then get your engineers to read your balance sheet, and teach them how to interpret each number. Teach them about cost of sales and time to market. No engineer learns that in school. And ask them to help you get your electronic project working!

The solution is communication, but even with good communication product development is still difficult. Without good communication, however, it’s even more difficult.