The Portfolio Effect – Odds of Success (Part 3)

Here are a few more thoughts to add to our discussion of the portfolio effect and risk.

I grew up admiring the concept of the “Lone Wolf” inventor. The Wright Brothers, Edison, Tesla, Bell. Parts of what we all have heard are myth, and some is reality. Several of these “individual inventors” created what looked a lot like a corporate entity at their time. However, the myth of the Lone Wolf lives on – e.g. consider Steve Jobs and the iPhone, or Mark Zuckerberg and Facebook.

It is dangerous to subscribe to this myth. At Micrex, 98% of our customers are large corporations. Only they have the ability and resources to launch a new product into a competitive global marketplace. They understand that the best way to develop new products is by having a broad portfolio of initiatives at various stages of maturity. This approach takes most of the luck and some of the risk out of product development.

The portfolio approach for product development shares the theoretical underpinnings with the same concept in finance. In the 1970’s, “Modern Portfolio Theory” conclusively demonstrated that maximizing return while systematically controlling risk required that one work with a diversified portfolio of investments.

At Micrex we have a soft spot for drunks, sailors and entrepreneurs. Despite our better senses — we occasionally get to work with an entrepreneur in a startup. In almost all cases, they run out of time or money (which in many ways is the same thing) before their product is ready for market.

How do we handle this same risk at Micrex? Through our customers. At any one time, Micrex is working on 300 – 400 projects. Many will not succeed, but some will. It is a matter of averages.

Fail Fast: It Should Be Part Of Your Strategy

At Micrex, we are big believers in the concept of “Fail Fast”. We have always talked about it with our customers – and what we have learned is that there are different interpretations of what “fail fast” means.


From an engineering perspective, fail fast is about system safety and optimization. A system that is designed to fail fast is one that will go into a failure state, rather than continue to operate in a possibly failed state. This allows back-up or redundant systems to immediately be activated.

The key is to minimize the delay between a likely failure and a truly flagging failure, so that an alternative solution can be implemented.

Extrapolating this to product development leads to two more alternatives:

  • Fail but learn: failure produces learnings, which inform next steps
  • Fail often: design the system to make it easy to try the next thing.

Celebrating Failure Doesn’t Seem to Make Sense

Why celebrate failure – particularly when we know that the corporate world rewards success?
The product development professional must strive to foster an environment where failure can occur without negative consequence. The good news: this is not as hard as it sounds.

The Solution: Test Early Or Inexpensively

It means looking for ways to test either early or inexpensively. Test components before testing the entire system. If something seems to work, then test variations that fail – it will give you confidence in the solution’s robustness.

Counterpoint: I was chatting with a long-time customer about changes at his company because of a recent merger.

Me: “What’s new at your company?”

Customer: “They’re training us to be more entrepreneurial”.

Me: “How’s it going?”

Customer: “Great – until something does not work out”.

Hope Is Not a Strategy – Odds of Success (Part 2)

Recently I corresponded with a reader who was concerned about the implied negativity around the “odds” of success described in these posts.

I have done quite a bit of research on the odds around product development and am comfortable with my numbers. It is not much different than saying that 1 out of 5 small business startups succeed. I have actually been pushed toward these terrible odds by the senior people I have worked with in product development.

My view is to embrace the numbers. Success comes from dealing well with failure. Run through the things that don’t work. Kill dumb ideas. If you have only one or two new products, it is really hard to prioritize and pick between the winners and failures. View it from a portfolio perspective — an investor who buys one stock is gambling.

Or you could choose to look at it another way: how does one actually calculate the success ratio? What is the numerator, and what is the denominator? A researcher working at the bench may see lots of failures. The chairman, who validates the decision to launch a fully refined product, sees completely different ratios.

It is a lot like sales. When I was young and naive, if someone called and asked about my product, I figured that was about as good as a sale. Now when I have a new lead, it does not even register until there has been a great deal of qualification. As we manage a sales effort, we recognize that it takes multiple leads to result in a sale, and we organize accordingly. We need to recognize that similar metrics apply to new product development.

The Chairman Wants Options

chairmanWhile the odds of any particular product being a commercial success are low, commercial success may not be the only measure of accomplishment.

Next year will consumers want little cars or big cars? Are heating costs going up or down? A good product development manager provides a portfolio of options to management. By definition, if there are a wide range of options, only a limited number will be exercised.

If you have product development responsibilities, are your expectations of success aligned with your senior management?