March 29, 2021
In part 1, we looked through some examples of how failure is a part of life and some of my own epic failures and in part 2, we covered how we need to feel ok about failing spectacularly. But how do we fail better? How do we fail in a way where we’re actually working toward success? Toward innovation? And learning from our failures? Let’s look at that now!
There are three ways we learn from failure:
First, by finding errors when they are small and easier to correct and then consistently sharing them. As humans we don’t want to see errors, we have blindspots. We don’t want to bring them up or share them when they are found. If we don’t have the culture and the systems in place to allow this, it’s going to be challenging to learn from failures if we can’t even identify them or talk about them.
The picture above is from a Toyota factory line. The red arrows in the picture indicate cords all along the assembly line that any of the workers can pull. When you pull the cord, it stops production. These cords - Adnon cords - are not safety equipment, but they are ways to prevent defects from moving down the line. They have empowered everyone to prevent the defect, regardless of how much it costs to stop the line. This approach was shocking to western manufacturing companies. That’s a lot of power given to employees, to stop a multi-million dollar manufacturing line.
Amazon has taken this idea and applied it in a couple of different ways. One of those Adnon cords is that any customer service representative can “stop the line” by removing any products on the retail website until a defect is fixed.
Second, we need to systematically analyze failures, to dive deep to find the root cause. Repeating the same mistakes over and over again without learning the lesson is unproductive, expensive, and shouldn’t be celebrated. There is also a lot of emotion involved in the analysis step. Our biases are definitely at play. Our brains want to explain away the hard parts, downplay responsibility, and place blame.
Correction of Errors (COE)
Another process we have, or what we call a mechanism, is a Correction of Errors (COE). A COE at Amazon is essentially a blameless postmortem. It’s a mechanism we use to learn from our mistakes or a problem and to drive continuous improvement, whether it’s a flaw in the tools, processes, or organization. It documents what the problem was and how to avoid it in the future.
A COE answers these questions:
- What happened?
- What data do you have to support your analysis?
- What was the impact on the customers?
- What are the contributing factors?
- What did you learn, and how will you prevent it from happening again in the future?
Then this information is shared widely. The video below goes into more detail on the COE mechanism at Amazon.
Third is to have proactive and strategic experimentation. It’s intentional, you’re seeking new knowledge. This experimentation is required because you don’t know the answers in advance. Think COVID vaccines or brand new products. Failure in these cases is not optional. It’s a core part of innovation. Researchers know their experiments can result in a large percentage of failures and so they need to become comfortable with failure. I read that 90% of experimental drugs fail. That’s obscene. That sounds expensive! But the payoff when you do hit the jackpot is huge. It’s worth that failure so that you can learn what doesn’t work and keep applying those lessons toward success. So that you can save lives.
Working Backwards process
At Amazon, we start with the customer and works backwards with everything we do. We have a process for this, called the Working Backwards process.
This process asks the 5 customer questions:
- Who is the customer?
- What is the customer problem or opportunity?
- What is the most important customer benefit?
- How do you know what customers need or want?
- What does the customer experience look like?
The artifacts from the working backwards process include a press release, an FAQ, and visuals. In addition to putting the customer first, this puts a simple process in place for generating ideas, clarifying them, sharing your ideas, validating the customer value, and getting buy-in. Everyone at the company can do this. And it’s a way for your ideas to be heard. By examining the customer’s pain points, we drive continuous improvement and innovation.
What can you do?
So all of this probably seems daunting and it very well might be depending on the culture of your company right now. Remember that culture change is hard and massive and hairy. It takes a long time to change. And it takes leadership, from the top, to make it stick. But don’t let that stop you! We are all leaders and there are actions you can take today, even if they are on a smaller scale.
- Reevaluate your team values and how they align with your company’s. Define tenets for your team.
- Talk to your teammates about what psychological safety means to them.
- Implement a retrospective after each project or sprint.
- Implement blameless postmortems when things go wrong.
- Create an award for most spectacular fail and recognize teammates regularly.
- Lean into experimentation and iterate with prototypes and minimum viable products or MVPs.
- Learn in public.
- Share your mistakes.
- Ask questions.
- Lead by example.
Finally, we need to accept failure as part of the process. Get comfortable with it and make it a habit! How we handle it and learn from it is what’s important. It’s what moves us forward and what enables innovation. Building a culture that supports psychological safety and encourages failure will be rewarded with innovation.