One question that has long plagued organisations is how to improve performance among frontline workers, the people who actually drive customer experience. Our work with hundreds of companies offers a clear and simple answer.

To show how it works, we’ll walk you through an example. In 2016 the leadership team of a national retail organisation asked us to help boost their frontline performance. They wanted to improve revenue, cost, risk, and customer satisfaction all at the same time. (They reached out to us because we wrote a book describing how these performance outcomes would be improved with an operating model that increases motivation.)

We’ve written before that why people work determines how well they work — that someone’s motive for doing a task determines their performance. Our work has shown that if a person’s motive is play (for example, excitement from novelty, curiosity, experimentation), purpose (the work matters), and potential (they are improved by the work), then their total motivation and performance increase. But if their motive is emotional pressure (shame, guilt, insecurity), economic pressure (mercenary behaviour), or inertia (no motive), then total motivation and performance worsen.

The retail organisation wanted to see how this applied to its stores. So we ran an experiment: We fully transformed the operating model of four stores (which employed around 60 people) for one year, and then compared their performance with that of the other 750+ U.S. stores.

As we predicted, we saw the performance of our experimental stores increase significantly. Productivity (revenue divided by expense) increased by 20% year over year (far more than the 9% increase in revenue that the control group stores averaged); customer satisfaction increased by 11% (the control group saw it decrease by 4%); and sales increased by 8% (the control group saw only a 2% increase).

We should note that this organisation is one of the top performers in its industry, so the baseline performance was already high. But we believe that by taking similar steps, the average organisation can improve performance even further.

Focus on Learning, Not Pressure

Prior to this pilot, the operating model of the stores was focused on creating emotional and economic pressure to drive performance. District managers would often hear, “You need to get your team to try harder,” or “This is really not what we would expect from your store,” or “Other stores are doing better.” On occasion, managers would use special rewards or threats to motivate better performance.

This playbook is the norm in most organisations. Based on what we’ve seen, frontline employees are among the lowest in total motivation.

To engage the front line at the retail organisation, we implemented a new operating model focused on optimising play, purpose, and potential while reducing the pressure. This required four major changes:

Reduce the economic and emotional pressure. To ensure this front line could focus on learning, we eliminated high-pressure motivation tactics, including sales commissions, high-pressure conversations, sales-based promotion criteria, and public shaming. We explained to leaders that great leadership isn’t about pressuring people to do their work. Rather, it is about inspiring your people to want to do their work well, so they can perform adaptively.

Incorporate a spirit of play by encouraging experimentation. To drive performance with play instead, we wanted to focus on increasing experimentation. Experimentation fosters curiosity, allows for novelty, and sets the pace of learning — all of which are important components of play.

To encourage experimentation, each store maintained an idea board that tracked the primary challenges the store had to solve, as well as ideas for solutions. For example, in one store, a challenge was how to get more walk-in customers. Colleagues could add any ideas they had, whenever they wanted. The challenges themselves were used to focus ideation on what mattered the most.

Employees were asked to choose an idea on the board and experiment with it. Every person was expected to have at least one experiment that they owned at a given time. They learned about hypotheses and about how to reduce an experiment to its minimum effective dose to get a useful result. Each store also had a weekly 45-minute meeting for teams to review their past performance and experiments, without shame or blame, in the spirit of generating new ideas.

There were rules. Experiments had to be doable on the job using only the budget, tools, and time already allocated to each store. Colleagues learned how to focus on creating low-risk experiments (experiments that were “above the waterline” so as to not sink the ship). Once an experiment ended, lessons were systemically captured and shared. There was no pressure for an experiment to work as long as something was learned. If an experiment didn’t work, the team wouldn’t throw it out, but instead would seek to understand why and launch a different experiment.

Create a sense of purpose around the customer. To build a genuine sense of purpose and meaning, the employees in the experiment stores were taught how to connect every product, process, and policy to the benefit and impact they had on customers. If they couldn’t connect an action to a customer outcome, they were taught that it was safe to ask questions until they understood.

Systematically manage apprenticeship. While experimentation is focused on learning strategic or process improvements, it is equally important to manage the pace of learning through apprenticeship. In a culture of apprenticeship, people receive high levels of on-the-job coaching by others who are higher in skill. And just like the system of experimentation, the system of on-the-job apprenticeship has to be tightly managed.

Here’s what we did: Each store was given a set of 30 frontline skills for employees to learn. This included things such as “generating ideas through sales and service data” and “building advisory relationships with prospective customers.” Everyone was asked to choose the skills that they believed would most improve their performance if they learned them. While leaders would help with this decision, the choice was ultimately up to each individual, with leaders focusing on finding on-the-job moments to learn.

We told leaders to identify and share the strongest skill (the professional “superpower”) of each individual on their team. The point of this was to make it easier for colleagues to seek help with learning skills from their colleagues. For example, one person’s superpower might be “solving complex service problems,” while another’s might be “introducing new products to customers.”

Every week, employees would have a brief development discussion with their leader on how they were progressing on learning their skills. Goals and metrics were transparent to everyone so that nothing was hidden.

As leaders and colleagues focused on skill development, performance problems were no longer met with blame and defensiveness. Instead, if a colleague was struggling to perform, the immediate focus became learning and teaching.

Seeing Performance Results

Only a few weeks after this new model was put into place, the teams’ behaviours started to fundamentally shift.

For example, in one case, one of the most junior members of the store led a successful experiment on how to move customers through the checkout line faster. Another store member conducted an experiment on how to better explain a product’s features to a customer. Another had an experiment on how to improve signage visibility.

At an individual level, colleagues started to open up to their teams on what they wanted to learn. Coaching accelerated. People became far more engaged in their work. At the end of eight months, year-over-year performance at our pilot stores had increased significantly, compared with the control stores. The approach is now being scaled across the organisation.

Broadly speaking, in front lines today, productivity growth has stalled, while employees are feeling less engaged and more stressed. Moreover, employee retention on the front line continues to be a problem. Too many organisations are responding to these trends with more pressure and micromanagement, which only worsens the problem and increases risk.

Small increases in productivity and retention can have a significant impact on the bottom line. Our rough calculation suggests that improving retention by one percentage point in a 5,000-person front line results, on average, in $2.5 million in annual benefit. (That number assumes a fully loaded cost of $50,000 per person, and a $50,000 cost to replace a new hire and make up for lost productivity while the new hire is found and trained.)

Now is the time for organisations to invest in workers. By implementing a frontline management system focused on driving performance through total motivation, you can build the ultimate win-win.

Originally published by Lindsay McGregor & Neel Doshi on Harvard Business Review.