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.

Fail?

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”.

Secrecy

I am going to argue that as a group, product development professionals have a bias toward excessive secrecy.

By nature, PD requires some degree of secrecy. It is obviously a poor idea to disclose all the details of your company’s strategy.
Top Secret
Inventors by instinct keep their best ideas under wraps. After all, you are only as good as your last idea. I have had customers with cubicles within yards of one another working on projects with us, but they never share the information with their co-workers. There are also the dark cautionary tales about people who disclosed confidential information through carelessness or negligence.

Simply put — in most organizations you can get in trouble for sharing too much.

Here’s the dilemma: by definition, product development relies on an exchange of ideas. New products do not occur in a vacuum.

Of course, as practitioners we will never be criticized for excessive secrecy, but our organizations will pay the price over time through a failure to innovate.

I have seen very few ideas lost or stolen, but thousands that have failed because of poor execution.

This brings me to another of my favorite aphorisms: “Every time I think I see conspiracy, in the end all I find is ignorance or sloth.”

Just One More Feature – Please!

Very often, when we are in the final stages of wrapping up a new product, we become so close to the task that we miss the big picture. The temptation to add one more feature or improvement is almost irresistible.

This can also have huge negative consequences.

256px-10th_mountain_division_ssi-svg

During World War II my father served in the Quartermasters Corps of the U. S. Army. His stories of the war tended to center around cocktail parties in Washington, but a few cautionary tales survived the rigor of wartime in the capital.

One was about the pitons used by the mountain troops. Apparently the piton manufacturer and the Army were quite proud of the work done behind enemy lines by troops using these pitons. They agreed to stamp “US” on each piton.

Tragically, the pitons began failing in the field. Eventually they discovered that the stamping compromised the piton.

Repeat after me – “Perfection is the enemy of good.”

Luck

When Napoleon was asked what kind of general he preferred, he said, “Just give me a lucky general”.

406px-Jacques_Louis_David_-_Bonaparte_franchissant_le_Grand_Saint-Bernard,_20_mai_1800_-_Google_Art_Project

While I am not a big believer in luck per se, this blog makes many references to luck. It is understood that invention is all about the unexpected occurrence. Having the insight to recognize these events and to run and seize opportunity could be considered making your own luck.

In “The Black Swan” Taleb argues that the inventor in Manhattan has a huge advantage over the inventor living on a mountain top. The Manhattan inventor gets invited to cocktail parties. At cocktail parties he meets investors, and investors are what you need to be a successful inventor.

The One Best Way (Part 6): Cowboy Coding

Software development is hard. Whether it is combining incredible levels of creativity and complexity, or talking about the high rates of failure – the debate about software design and development techniques has a lot of relevance to other disciplines.

We have already touched on the use of a grand design. This is sometimes referred to in software as “Waterfall Design”, as all parts of the product are supposed to flow together and work as one in the end.

At the other end of the spectrum is “Cowboy Coding” where the lone programmer is given freedom to do what seems right. There are “nicer” versions. For example, “Agile Programming” has the patina of an intellectual framework.

B. F. Skinner said, “A first principle not formally recognized by scientific methodologists: when you run into something interesting, drop everything else and study it.”

Just another Harvard trained cowboy.

WHY THIS BLOG:

Companies today need to become much better at innovation. While most organizations claim to foster new products, their actual track record is poor. Managers are using old tools and methods to shape the future.

Except for the rare companies, such as Google or P&G that are richly endowed with a culture of innovation, most managers are left to fumble or improvise a process for developing new products. Regrettably — these improvisations are likely to fail.

Yet it is the nimble smaller firm, when combined with an appropriate innovation strategy, that has the greatest chance for success.

My goal is to help those who are struggling with PD to find a successful path through the art and science of innovation.