Embracing Failure

Framing failure to unleash innovation

A few weeks ago, Jon Williams and I shared our insights on the benefits of failure in a blog post, focusing on blameless postmortems and failure workshops. The latter is an extension of a pre-mortem, maintaining team focus on potential pitfalls of a project or initiative rather than immediate solutions. The former, a method adopted by Etsy and numerous other companies, treats failure as a learning opportunity rather than a platform for blame. Despite establishing an open discussion on failure, it remains a challenging concept for many to embrace. After all, falling short of goals isn't on anyone's agenda.

Attaining goals is crucial, but so is embracing the possibility of failure. Without this, one cannot truly maximize potential gains. This conundrum frequently arises in contexts like OKRs or nested goals. As a leader in product engineering, it's common to have a portfolio of projects with varying probabilities of success, some of which promise higher rewards despite their lower likelihood. Using the 3X framework as an example, some projects aim at extracting value, others at expanding, and yet others at exploring. Typically, those lower on the S-curve bear greater risk. An educated guess would suggest that around 80% of the projects in your portfolio will meet their targets. You might incorporate this assumption when setting your goals, aiming for the 80% achievement mark. However, if this conservatism scales at each organizational level (VP, Director, Senior Manager, Manager, etc.), the end goal becomes unduly modest. Conversely, neglecting this factor could lead to falling short of top-level goals.

When managing a large project portfolio, accurately estimating the success rate enables goal attainment. Yet, as a team leader for a single project, achieving or missing your goals is a binary outcome - you either succeed or fail. This illustrates the difficulty with failure. While organizations may openly discuss failure through tools like blameless postmortems and failure workshops, individuals often avoid personal failure, leading them to choose safer paths, lower goals, or less bold strategies. This reticence can result in overall diminished company performance. How can we address this issue?

Mark Rober, CrunchLabs founder and ex-NASA engineer, known for his YouTube channel, designed an experiment examining human reactions to the risk of failure. He devised a simple programming puzzle and challenged 50,000 of his followers to solve it. Unbeknownst to the participants, two slightly different versions of the puzzle existed. One group started with 200 points, remaining unaffected by failed attempts. The second group also began with 200 points, but lost five points each time they failed. Despite these points being purely symbolic and void of real-world value, the group facing no point-loss played the game 2.5 times more and succeeded 16% more than their counterparts. Rober likens the no-loss group's attitude to that of video game players who persist until they win. His insight, that he shared as part of his commencement address at MIT, was to "Frame failures like a video game, and you'll not only iterate more and learn faster but also sweeten the taste of subsequent successes." How can we transfer this video game approach to product engineering?

A direct answer is to disassociate achieving product goals from employee evaluations, promotions, or raises. Believe me, even if this is declared publicly, and yet still discussed in calibration sessions behind closed doors, employees will deduce the truth from promotions or rating patterns. Perhaps you're wondering, "How can he advise against conflating successful launches with achieving business outcomes while discouraging product goal achievement consideration in evaluations?" The distinction lies in collective success or failure in achieving product goals. Teams should receive credit not for product launches, but for achieving business goals. Consider a team working on search improvements to boost conversion rates on search pages by 0.5%. Numerous changes may be launched, which the team should indeed take pride in. However, they should celebrate their success only upon reaching the targeted 0.5% conversion rate improvement. When it comes to individual evaluations, promotions, or raises, being part of a successful or unsuccessful team should not be a determining factor. These decisions should be based on demonstrated behaviors and competencies.

By reframing failure as a learning opportunity and decoupling it from professional evaluations, businesses can foster an environment of boldness and innovation. This approach, inspired by the video game-like mindset, encourages teams to take calculated risks and push boundaries. Evaluating individual performance should be grounded on the demonstration of competencies and behaviors, rather than solely on the outcome of projects. By making this shift, companies can unlock the full potential of their teams, nurturing a space where growth, exploration, and risk-taking become the bedrock of success. This is an essential evolution, and while we spoke about it in terms of product teams it’s not just for product engineering, but for any modern, forward-thinking organization.