Scaling iteration

Fostering an experiment-driven culture and driving behavior change with forcing functions and gatekeepers

The problem with using outcomes to define success is that outcomes are lagging indicators (hindsight) — they can only be measured after investing time and resources in launching a project or initiative. In a fast-changing world, it’s important to have leading indicators (foresight) to predict much earlier whether a team or individual will have success. At Alpha, we’ve learned that teams that iterate tend to outperform teams that don’t, so we measure iteration as a leading indicator for success. The subsequent challenge is in fostering a culture of iteration at scale, especially when it requires behavior change.

We’re not the first company to recognize that iteration is the key ingredient to success. Many of the world’s leading companies share a similar view and work tirelessly to optimize employee behaviors accordingly.

Status quo

Ray Dalio, the founder of the world’s largest hedge fund, Bridgewater Associates, famously measures process over outcomes. “Knowing how people operate and being able to judge whether that way of operating will lead to good results is more important than knowing what they did,” he wrote in his book, Principles. “We call this ‘paying more attention to the swing than that shot.’”

Unlike Dalio, most organizations reward the shot more than the swing, measuring the quality of a decision by the quality of its corresponding outcome. Psychologist and professional poker player, Annie Duke, refers to this as the fallacy of resulting. While the worst poker player can get lucky and win big on any given hand, you’d be wise to determine who the better poker player is overall and put your money on them.

Resulting rears its ugly head in the war for talent. That most companies almost exclusively measure and reward outcomes presents a challenge for companies like ours trying to scale a culture of iteration. Virtually every prospective candidate and future employee has already developed biases toward polish, showmanship, and shifting risk to after an initiative’s launch, rather than toward exposing ideas, confronting risk, and iterating early and often.

Forcing functions

The most painful and tedious method for driving behavior change toward iteration is quite obvious: simply expose teams to its absence and let projects fail. I have found that, in any organization, iterative and customer-driven teams tend to be the ones with the most painful and humiliating failures in the rearview. They want to avoid making the same mistakes and reliving those experiences, like overly prudent drivers who earlier narrowly survived catastrophic crashes.

The challenge of course is that this method of learning has real costs and damage, and isn’t ideal for companies that want to drive behavior change faster and less expensively. Instead, consider repeatedly exposing teams to simulated failure by creating a forcing function with guard rails.

In any given industry, you are likely to find a methodical and clever way in which professionals simulate the real-world to expose themselves to criticism and ‘road test’ certain strategies. Chefs use test kitchens, lawyers use mock trials, and entrepreneurs experiment with ads and landing pages. Chris Rock famously frequents a small comedy club in a college town to evaluate jokes before incorporating them into his mainstream routines. The simulations are gatekeepers to the real world, offering foresight into potential risks and opportunities much earlier.

There are countless ways to incorporate simulations and optimize iteration in a startup or corporate environment.

If a product manager has to achieve predefined levels of user engagement in a beta release before being able to push an update to an entire customer base, they will rapidly learn how to get feedback on prototypes earlier. If a salesperson has to collect detailed information from a prospective customer before being able to issue a contract, they will rapidly learn how to ask tough questions earlier. Valve, the renowned video game company, takes gatekeeping orders of magnitude further, enabling employees to freely assign themselves to projects they perceive have demonstrated the most value and momentum. If an employee there wants resources and a team to launch a new game, they’ll have to make a compelling case with evidence.

In all these scenarios, teams and individuals operate with autonomy and accountability, while creatively confronting risk, iterating, and generating evidence early in order to progress to successive stages of a project.

Success criteria

As our company scales, three metrics will make or break our culture and ability to innovate:

1. The time it takes for a new hire to shift biases toward iteration. Many of our best and most talented employees come from large organizations where polished work and delaying risk is rewarded. For them, iteration and exposing unpolished ideas may at first be unnatural and uncomfortable. How can we accelerate this shift?
2. The percentage of employees that adopt a bias toward iteration. Even at best-in-class organizations, it’s unlikely that everyone (or even most employees) have a bias toward iteration. How can we achieve a critical mass?
3. The virality of iteration. The most important internal metric for scale isn’t how many or what percentage of employees have a bias toward iteration, but the number of employees who can teach and hold others accountable for iteration. How can we empower employees to create effective forcing functions within their own teams and purviews?

The pace of change is going to continue accelerating. Organizations that drive behavior change across their workforces to optimize for foresight will continue to outperform organizations that exclusively and comfortably reward hindsight.