I recently hosted a discussion on a topic I’ve spent a lot of time thinking about. Failure. Specifically, the role of failure in innovation.
I know this can be a touchy topic – but in full disclosure, I’ve experienced enough of it in my career to develop a pretty strong perspective on it.
The history of failure extends back as long as mankind has been trying to invent new things. And it’s not like we’re really getting any better at it. Yes, we learn from the wisdom of each passing generation. Our tools have become more and more sophisticated. And yet, we still seem to find ways to fail spectacularly and regularly.
But dramatic, headline-grabbing, sometimes tragic failures like Theranos and the Boeing 737 Max are just the tip of the iceberg. The vast majority of failures are far less public, but still leave their mark. They’re products that just don’t work well – cause frustration or disappointment and get cast aside. Or products that work as intended but never sell. Or products that can’t quite sustain a business. And it gets worse - the products that make it into the marketplace are just a small fraction of the failed projects that consume remarkable amounts of time, money and resources but never see the light of day or make it onto a shelf.
There are no accurate statistics on product failures. A Harvard study a while ago estimated that between 80-95 percent of all products introduced into the market fail. Perhaps the most jaw-dropping statistic is that it’s been estimated that about half of all the resources, time and money allocated to development ends up being spent on failed products!
The reason for these failure rates lies in our insatiable desire for the innovative and new. Innovation improves our lives, drives our businesses, and separates us from our competition. But striving for innovation inherently implies taking risk. And the very definition of risk is something that might end in failure. Innovation, Risk and Failure are intractably linked and always will be.
There’s no possible way you can drive your business forward in an intensely competitive environment without some level of controlled failure. Pick any company you admire and I’m positive that you’re going to find a fair degree of failure littered in their past.
Part of this can be explained by the sheer number of things that have to go perfectly to succeed in new product development and innovation. It’s perhaps the only activity that requires the integrated collaboration of just about every aspect of a company – strategic planning, finance, accounting, marketing, legal, engineering, manufacturing, distribution and sales.
If you consider that you were likely to be successful 75 percent of the time on each of these isolated activities, when you combine them the cumulative probability of something being the weak link in the chain helps explain why so many projects are doomed to fail.
So, let’s spend a bit of time putting some context around failure and then talk about ways you can increase your chances for success and minimize the impacts of the failures that do occur.
All failure is not created equal – and much of this has to do with the point in the development process when failure occurs. A reasonable development process starts with strategic planning, determining user need, and developing a business case. If things look favorable, you might progress from considering concepts into engineering, programming, or development and then finally a launch into the market. Each of these activities represents an area of risk and a possibility for failure - although I think it’s actually a misnomer to call them failures when projects are truncated early in the process. Early on, the costs committed to projects in terms of people, time, and money can be pretty insignificant. So intuitively, you want to tolerate failure as early in the process as possible.
All innovation isn’t equally ambitious. The further your company is “reaching” from its current products and services, the higher the risk of failure. New technologies, emerging or previously unmet user needs, or brand-new business models carry a much higher threshold of risk and higher development failures. Conversely, if you’re introducing “Version 4.5a” of an existing product that a modest update, there should be a much higher expectation of success.
Products fail for a really broad range of reasons. One pretty useful taxonomy for deconstructing failure is shown in the graphic below.
At its core – a product or service has to be desirable. It needs to meet a need in the marketplace and be valued and desired by your customers. It also needs to be feasible –it needs to work as intended and expected. It must be safe, reliable, conform to applicable standards, and deliver the features and functions expected in a dependable and delightful manner. Finally, it needs to be viable, creating and supporting a sustainable business. You need to be able to manufacture, distribute and price the product or service in a way that covers your costs and leaves some left over for that most capitalistic of endeavors – the profit margin. In the attached video of my recent presentation, I give some real-life examples.
Sometimes the truth behind a failure isn’t quite as simple as our public perception. Consider the case of New Coke – perhaps the most compelling failure story in modern commerce.
It’s been argued that New Coke might have been the most thoroughly market-tested product in modern history and the product was uniformly preferred to existing Coke and its rival Pepsi in blind taste tests. But when New Coke was introduced, it was a colossal, embarrassing and remarkably expensive failure. Yet, if Coke is to be believed, there are many aspects of the New Coke experiment that were a resounding success for the company.
“Yes, it infuriated the public, cost a ton of money and lasted only 77 days before we reintroduced Coca-Cola Classic. Still, New Coke was a success because it revitalized the brand and reattached the public to Coke.”
— Sergio Zyman, Coca-Cola head of marketing
The “Stepping stone”
Sometimes failure is an effective stepping stone to future success. The “failed” Apple Lisa went on to inspire the remarkably successful, iconic, and groundbreaking Apple Macintosh.
Or consider the Toyota Prius. In 1997, the first-generation Prius took some bold technological steps – and suffered from some pretty dramatic teething pains including a string of recalls and failures. And it only sold a fairly paltry 30,000 units. But Toyota has a great track record of playing the “long game” and stuck with it to create what might be one of the most beloved, dependable and successful small cars in modern history.
Compare that to the GM EV-1. Introduced in 1996, it had a very favorable reception and deep owner loyalty. But GM concluded electric cars were an unprofitable niche and bailed on the market in 1999. Imagine what might have happened if GM had taken Toyota’s approach and had a 10-year head start on Tesla.
Tools and techniques for managing risk
Perhaps one of the largest influences on a companies’ long-term track record of innovation is how their culture deals with failure. It isn’t fun, but all is not lost. There are things you can do that increase the odds of success, reduce cost, and ensure that failure is minimized and sequestered to the greatest extent possible.
Have an honest and realistic discussion about risk and the possibility of failure. Whether you call it wargaming, scenario planning, worst case, common sense or just plain realism, encouraging dialogues about risk and failure is a really critical part of the process.
Engineers have a wonderful process they use for formally assessing risk – called an FMEA – or Failure Mode Effects Analysis You can do an FMEA for just about any aspect of the development process from strategic planning through distribution.
The cost of failure rises dramatically during the development process. So, it stands to reason that failing earlier will conserve precious resources. Tolerating failure earlier in the process and making sure your process anticipates and accepts some degree of “infant mortality” just makes sense. I like to say that failure should be encouraged in strategic planning, expected in conceptual design, tolerated in engineering, avoided in manufacturing deployment, and shunned like the plague after product launch.
And there is a huge difference between “project failure” and “product failure.” Neither are fun, but product failures can mean the end of a company or worse.
3. Align expectation
Recognize that the probability of failure increases the more innovative you’re trying to be. Take the time as an organization to assess your appetite for risk. Make sure you don’t bite off more than you’re willing to chew.
4. Patience, perseverance and pivots
While being willing to kill deeply flawed projects is a characteristic of winning companies, likewise, recognize that your first effort is seldom your best. Just about any successful company has endured dramatic shifts from their initial plans. But too few companies are willing to plan for that. Think what might have happened if GM had considered the EV-1 to be the first step in a multi-phase journey.
I’m deeply skeptical of any innovation effort that doesn’t start as modestly as practical and plan for iterative evolution in the marketplace. The principles of lean startup, sprints, and minimally viable products are all good steps toward this concept. Accordingly, a development plan that recognizes the potential for warts, flaws, and missteps in the initial launch can be a very powerful tool.
5. Design Spirals
Whenever we talk about failure there is going to be a lot of talk about rapid prototyping and iteration. I think a better metaphor is the spiral concept – where there are a series of design/test/prototype iterations that all progressively move you closer to a final product.
Whenever I see a development plan that moves right from a first prototype into production – or skips the prototype entirely – I’m worried. Make sure your development plans include progressive iterative spirals designed to move toward an increasingly optimized target.
6. Fight the hottest fire
It’s important to devote the most resources and efforts to the riskiest element of your project. This one might seem obvious, but I’m constantly amazed at how many companies throw significant resources into areas that represent minimal risk. Part of this is the tribal nature of our siloed development resources. Truly skilled and empowered project teams bump up a level and make sure they are committing the most effort to the area of greatest uncertainty. Building an expensive prototype of a deeply flawed business concept is a true waste of time and effort.
As outlined earlier, success in innovation lies at the intersection of desirable, feasible and viable. Each has tools that can be used to assess, qualify, quantify and modify if needed.
7. Parallel path
Another prudent technique is to formally decide what aspects of your innovation require a parallel path. By parallel path, I mean an alternative, less risky approach or fallback that’s co-developed in conjunction with highly risky aspects of the design. These consume costs and resources, but often the cost and time of a parallel path easily offsets the risk of failing an aspect of your product that has no contingency.
Let’s talk about how we can help move your business forward.