Learning From Mistakes

As software engineers, we’re faced with new problems and challenges every day. No matter how well we know a programming language, how many projects we’ve worked on throughout our careers or how much time we’ve spent creating repeatable solutions to common problems, there will always be something new that requires critical thought.

Inevitably, then, we will make mistakes. Deadlines will not always be met, solutions may not always be correct, and critical tasks may be overlooked. We may even accidentally break the software. These are not Bad Things.

When we’re walking an unfamiliar path, we’re bound to make the occasional wrong turn. What’s important is that we know that those mistakes have a lot to teach us, and that we’re in an environment where it’s safe to make those mistakes.

Creating a safe environment to fail

Avoiding Blame

From an early age, we’re taught that when we do something wrong, there are negative consequences. We learn to associate those consequences with the action that caused them, and we actively avoid it in future. We’re also encouraged to assign blame when we see others doing something wrong, if only to again avoid negative consequences directed at us.

This continues into adulthood and our careers, with the assumption being that your employees work as hard as they do to, in part, avoid making mistakes and being made an example of. The belief is that without negative consequences to failure, your employees will be less engaged and less motivated.

This attitude is counterproductive to a healthy working environment. Giving your team a space in which mistakes and failures can be accepted and learned from doesn’t mean encouraging lower standards, but ensuring your team and your organisation as a whole can continue to evolve and grow.

It’s also important to recognise that mistakes and failures are not necessarily the result of wilfully deviant behaviour. Tolerating mistakes, and recognising that they are opportunities for everyone to learn rather than for one person to be blamed, is a skill.

Retrospectives

Your team needs to know that mistakes can be tolerated, and the best way to convey this is to have open discussions about problems that have arisen, without playing the blame game. Take the time to talk about why a mistake happened. Once you’ve discovered the mistake, you need to find out what it can teach you and how it can help you in the future.

Regardless of what we discover, we understand and truly believe that everyone did the best job they could, given what they knew at the time, their skills and abilities, the resources available, and the situation at hand.

– Norm Kerth

This quote, known as the Prime Directive of retrospectives, illustrates perfectly the attitude that should be taken when trying to create an environment that looks at mistakes and failures in a positive light.

Retrospectives happen at the end of a sprint, usually every week or two, and give everyone a platform on which to highlight things that went well, and things that didn’t go so well. Providing you’ve built an environment in which it’s safe to be honest about shortcomings and mistakes, retrospectives are a great way to uncover process failures and to voice concerns about the work that may lead to avoidable mistakes.

An example we’ve encountered is realising that our mistake was not getting enough detail on requirements during the planning stage. This led to us moving down a path of work that we ultimately discovered was incorrect when we presented it to the customer.

Within the subsequent retrospective, we were able to safely discuss this as a team, and to admit that there were things we could have done but didn’t. We spent time figuring out why this had happened, and then deciding on actionable steps we could take to prevent it happening again in future. We came away from the retrospective having realised we needed to spend more time early on discussing requirements with the customer, and then making sure that information was disseminated across the entire team.

Embracing mistakes

We’re in the discovery business, and the faster we fail, the faster we’ll succeed.

– Amy Edmondson

The best way to deal with mistakes and failures is to treat them as opportunities to learn, both individually and as a team, after all, while it’s our job to design solutions that meet requirements, we’re not infallible. When we’re embarking on a new challenge, making mistakes is a crucial part of the discovery and experimentation process.

Our solutions may fail in unforeseen ways, or we may need to revisit the task and find that we’ve made more work for ourselves by creating something inefficient. Either way, you have the opportunity to reflect on what went wrong, why, and how you can simplify the process to either reduce or eliminate mistakes.

Big mistakes are easy to spot and discuss. In software, you know something’s gone wrong if, for example, a build fails, critical data is lost or a website goes offline. Steps are immediately taken to fix those mistakes and resume normal service. The trick is being able to identify and learn from smaller mistakes, as they’re much more easily hidden, both passively and actively. The earlier these are discovered, the better.

This mindset of actively discussing and learning from mistakes, rather than blaming anybody for them, doesn’t mean you’re encouraging your team to slack off and take shortcuts to the detriment of the project. Even in situations where a mistake can be attributed to an individual’s lack of care or inattentiveness, there’s the chance to dig deeper and discover what led to that behaviour, and what you can do to improve the situation for the individual and your team.

Using mistakes to uncover requirements

In most software teams, strides are taken at the beginning of a cycle of work to gather as much data and information as possible from the customer to understand requirements as completely as possible. Nevertheless, it’s not unheard of for a seemingly unimportant detail to be overlooked during this phase of the project, only to either become a blocking problem midway through development, or to go completely unnoticed and later be revealed as a key requirement whilst you’re showcasing your work.

Use these situations as opportunities to figure out how you can improve next time: what information didn’t you have that you wish you’d had? How could you have elicited that information from the customer? Could you have broken tasks down further to discover hidden requirements? Questions such as these will help your team improve with each new project, and ultimately you’ll deliver better work and make your customers even happier.

One way to try to discover hidden requirements is to carry out research spikes. On a software team, this would typically involve one engineer dedicating a small but significant amount of time, such as half a day or a whole day, to investigating whether a potential solution is worth spending more time on.

It’s a long enough period of time that some thorough research can be done, but short enough that, if it doesn’t pan out, the loss of time isn’t too much to bear. The researcher is also safe in the knowledge that, should the research lead nowhere, the team won’t consider it a failure.

Using mistakes and mentoring to help teach new skills

Less experienced members of the team may struggle with tasks other engineers find simple. As we’ve said, engineers at every level are constantly facing new challenges, and mistakes are bound to happen. However, when you have junior engineers working alongside senior engineers, the environment you’re creating should allow juniors to feel safe enough to approach their more experienced peers for guidance. To take it even further, encouraging your senior engineers to take an active interest in mentoring is a great way to quickly upskill newer members of the team.

Using mistakes to analyse common problems and automate them away

Software engineers love to automate all the things, but there’ll always be the occasional process that’s still being performed manually and, no matter how often the process is performed, the more convoluted it is, the more likely it is that a crucial step is overlooked, leading to a failure.

Back in the day, before the advent of source control, something as fundamental as deploying changes to a production environment was a manual process, and involved massive amounts of risk. If the deploy broke anything on production, you had to cross your fingers and hope that someone on the team had a historical copy of the offending file. That problem was solved with a combination of solutions such as Git and Jenkins, which give us the ability to easily deploy and move software through various testing environments all the way to a production environment at the touch of a button. If anything is broken, we can then easily roll the latest changes back.

Within your organisation, there are likely several risky and complex processes that are causing your team frustration. By allowing your team to identify and discuss these problems, you’re giving them the ability to work together to find a solution that reduces the risk and transforms the process from one that causes frustration to one that is almost mundane.

Conclusion

Mistakes and failures are not something to be feared, in fact, celebrating them is perhaps more appropriate. That statement sounds a little ridiculous but, when you consider how much a failure can teach you about the work you’re doing, the way you’re doing it and how you can help others do it, there’s too much valuable knowledge to be garnered from a mistake to set about reprimanding someone for making it.

About the Author

Scott Mason

Former Software Engineer at Made Tech. Probably listening to soundtracks right now.

Avatar for Scott Mason

We are hiring! Find out more about a career at Made Tech.

Download a copy of our new book

Legacy technology is one of the biggest threats to public sector organisations.
Whether you’ve started your journey already or don’t know where to begin, this 160-page book has been written to guide you to define and implement the right approach for your organisation.