SSIR x Bridgespan: Achieving Transformative Scale
SSIR x Bridgespan: Achieving Transformative Scale
Achieving Transformative Scale is an article series exploring pathways that social sector leaders around the world are pursuing to take solutions that work to a scale that truly transforms society.

Running a nonprofit is complex, we’d all agree. There are countless variables you could be monitoring. Yet, as a means of scaling for impact, tracking all of them is neither possible nor desirable. What to do? In the for-profit world, there’s a concept called “the one metric to manage.” Its message is that instead of reducing dozens of possible variables affecting success down to a more-manageable-but-still-large number, boil it down further—all the way to the few that really matter. Here are three stories nonprofits can use as inspiration.

Margin per store

The first is about a little restaurant chain in Boston known as Au Bon Pain. When I joined the company in 1985, it consisted of 19 French bakery cafes that had never made a dime. My mission was to scale it to profitability. We had no money to duplicate traditional methods, so we operated on the assumption that the managers who ran these stores might know how to do it better than the people who were telling them how. We were making 18 percent margins in the stores and asked ourselves, “Why don't we just pay managers 50 percent of every dollar they generate over 18 percent?” Their salaries averaged $31,000. But within six or seven months we had managers on pace to make $160,000 a year.

Teach this case at the Harvard Business School, and most MBAs would say this is a stupid solution and a shocking waste of revenue, because this is a business that doesn't really require any talent. Yet we began making money within 18 months. And within three years, were generating unit-level profitability that was unheard of in the industry. We grew from 19 to 125 stores in three years. Just one solitary metric—margin per store—combined with giving managers the power to affect it drove all this improvement.


The second story involves Taco Bell, which in the 1990s had the distinction of being the one quick-service restaurant chain in America that could generate volume only by selling its food at 59 cents, 79 cents, or 99 cents. Its core customers were thrift-oriented, high frequency, fast food users.

We started interviewing these core customers, and we discovered that just four variables—not the other 55 that management worried about—could explain 90 percent of the variance in their purchasing habits. We called them FACT: fast food fast, accurate order fulfillment, clean environment, and temperatures for food (hot food hot, like meat, and cold food cold, like lettuce).

We presented this to the senior management team, and someone raised his hand and said, “You haven't said anything about the food.” My answer was, “Neither did the customer.” What followed were restaurants without kitchens—re-conceptualizing the notion of a restaurant as a point of distribution. Our nod to transformational scale was the concept of “taco ubiquity.” In other words, a customer should always be just a short distance away from their next Taco Bell.

Conversion and units sold

The last story is about Limited Brands, which I joined in 1998, and where I ultimately became vice chairman. It operated a whole portfolio of apparel brands, plus Victoria’s Secret and Bath and Body Works. At its peak, it ran about 5,000 stores, with some 160,000 employees during the holiday season. And the variance in sales and profits from store to store was totally unacceptable.

We started by asking the question, what do we hold people accountable for in a store? And the answer was sales. But doesn't the merchandise really matter? Doesn't the location really matter? Don't the hours of operation really matter? Doesn't the stock’s position really matter? If you keep asking questions like these, pretty soon you have a list of too many variables to measure again. We had to focus.

Long story short: We didn’t succumb to the temptation to keep adding those variables. Instead, we reengineered and retrained an entire force of thousands to understand that they had control over just two things that really drove sales—the conversion of traffic into customers and the units sold per transaction—and we began to measure them on those two metrics. Store staff knew what to do: Greet customers, help them meet their needs in our store, and find more solutions than the customer had imagined. The result was a reduction in store level performance variance that drove marked increases in sales throughout the system.

Let me end by sharing a conversation I had with a Harvard Business School student who'd been a fighter pilot. I asked him about all of these buzzers and bells and gauges and everything that's in the cockpit. It's just amazing. I said, “So, how do you use all of that?” And he said, “By and large I don't. It's there kind of ‘just in case.’ Most of the time I don't even know where it is.”

There's and old expression called flying by the seat of your pants. I'm not asking nonprofit leaders to operate this way. But I am suggesting that they get very, very rigorous about what buzzers, bells, and gauges they place in their cockpit.