Dashboard Thinking vs. Windshield Thinking - Why Looking Away from Metrics and KPIs is an Essential Leadership Habit

On July 16th, 1969, Neil Armstrong, Buzz Aldrin, and Michael Collins launched into space on top of a multi-stage Saturn V rocket. The Apollo 11 program was the most technologically initiative mankind had ever developed, supported by the Apollo Guidance System (AGS), a cutting-edge navigation computer, and the collective manpower of NASA Mission Control in Houston, Texas, hundreds of highly-trained men and women monitoring the most advanced flight instruments ever built.

Three days later Neil Armstrong and Buzz Aldrin began their descent to the lunar surface on board the Eagle Landing Module (LM). The AGS took over as designed, navigating the LM to a flat, clear landing spot NASA engineers had planned and programmed months earlier. Or so everyone thought.

Armstrong looked out the windshield. Something was wrong. He saw the LM was headed for disaster: a boulder-strewn field.

Armstrong took manual control. He disabled the LM’s guidance system and steered towards a nearby clearing, but the Eagle was going too fast. Under pressure, with computer alarms blaring and fuel dwindling, he eyeballed another spot, almost landing before noticing another crater blocking his descent. Armstrong cleared the crater, visually locating another landing site, but lost visibility when dust blown up from the lunar surface covered the windshield. Neil trained his eyesight on the speed of rocks flying past, estimating the LM’s speed and altitude in his head. With 25 seconds of fuel left and hundreds of millions of people watching, Neil Armstrong landed on the Moon’s surface.

Armstrong’s decision to switch to manual control, temporarily ignoring his dashboards, made the difference between triumph and the worst disaster in aviation history.

Apollo 11 was a human and systems engineering triumph. It relied on cutting edge computer systems, and hundreds of experts tracking each metric and measurement every step of the way. But in order to succeed, Armstrong needed to stop and look out his windshield. Likewise, your business may be mature and process-driven. You may have dashboards, metrics, and controls that make Houston Mission Control blush. But to succeed, you need to make time for Windshield Thinking.

Just like Neil Armstrong needed his dashboards and his windshield, modern leaders shouldn’t rely on Dashboard Thinking alone. They must make time to look beyond metrics and KPIs, in search of the insights only a fresh, big-picture perspective can offer.

Dashboard Thinking and Windshield Thinking

You can’t fly a plane, or command a spacecraft, without a dashboard. Pilots depend on flight instruments to know how fast they’re going, how far they can go, and what’s wrong with their craft. But to fly safely, pilots still need windshields. They need to see the world around them, assess their surroundings, and make decisions based on instinct and experience. Even the most advanced instruments won’t tell you everything. In Armstrong’s case, they may even steer you wrong. That’s why balancing these two perspectives, like Neil Armstrong did, is critical.

Effective leaders apply this same balance to their own decision making: Dashboard Thinking for quantitative, metric-driven decisions, and Windshield Thinking for the unstructured, instinct-driven approach for exploring new ideas and analyzing unfamiliar threats. 

In business, metrics are the gauges and instruments on your dashboard. Your metrics help you track things like profitability, sales costs, employee attrition, and so on. Metrics help us make better strategic decisions, just as our dashboard lets us know to slow down when we’re going too fast, or pull over when our oil pressure drops.

Metrics are vital, but they don’t tell you everything. Certain strategic decisions can’t be made based on a spreadsheet or a number on a calculator. As many real-world examples show, companies that overemphasize metrics in driving strategy or rewarding employees often fail. Wells Fargo used metrics to implement a major shift in the bank’s customer identity, resulting in large scale fraud. IBM applied a mainframe mentality to the personal computer revolution, and lost ground to upstarts like Microsoft and Apple. In 2008, some of the most powerful institutions in the world failed to see fundamental weaknesses in the mortgage-backed securities market, an oversight that nearly toppled the global economy.

These cautionary tales are why leaders should be wary of putting their trust in a few metrics, are leaving key performance indicators (KPIs) in place for too long. They’re why leaders need to reserve time for brainstorming, imagining new possibilities, addressing unforeseen obstacles, and experimenting with ideas that fly in the face of conventional thinking. Leaders need Windshield Thinking.

Why Dashboards Are Vital

The world runs on metrics. Investors live by the Dow Jones Industrial Average or the S&P 500. Bankers watch the 10-Year Bond Rate. Retail stores track conversion rate, sales per square foot, gross-margin-return-on-investment, and units-per-transaction. The world of sports tracks batting averages, passer ratings, etc. Pilots use altimeters, attitude indicators, and airspeed indicators.

On July 16th, 1999, thirty years to the day after the launch of Apollo 11, John F. Kennedy Jr. strapped in behind the controls of his Piper Saratoga single-engine airplane in Martha’s Vineyard, Massachusetts. With him were two passengers: Carolyn Bessette, his wife, and Lauren Bessette, his wife’s sister. It was 8:38 PM.

At 9:41 PM, Kennedy’s plane crashed into the Atlantic Ocean, killing himself and both passengers instantly. The National Transportation Safety Board (NTSB) reconstructed the evening’s sequence of events. Kennedy had chosen to fly at night, in harsh weather conditions, despite not holding an instrument rating, the qualification reserved for pilots able to fly safely without the use of a windshield. His reasons are unknown, but many believe Kennedy took this risk out of fear of missing his cousin Rory’s wedding the next day.

Most experts attribute the crash to spatial disorientation. It’s the aviation term for the temporary sensory inability to determine angle, altitude, or speed due to physical or environmental distractions. It’s most common at night or in bad weather, when pilots can’t make out the horizon through the windshield. Pilots have been known to fly straight up or down, thinking they were level with the ground. Kennedy’s Piper Saratoga crashed nose-first into the Atlantic Ocean.

Spatial disorientation is an example of how sensory perceptions fail. It’s the reason planes have dashboards. Altimeters and airspeed indicators can’t fly the plane by themselves, but they can give the pilot information they can’t always infer by looking out the windshield. On a clear day, experienced pilots can get by without their flight instruments. In a bad storm, they become vital.

We rely on our sensory judgement every moment of every day. We use our sense of balance to stay upright, and our vision to avoid other cars on the road. We rely on intuition to avoid dangerous situations, and our instincts to make snap decisions when we don’t have time to stop and think analytically.

The human brain can do a lot of amazing things, but our senses have limits. We can be tricked into seeing things that aren’t there, taking irrational risks, even thinking we’re upright when we’re flying into the ocean. Professional gamblers and less-ethical salespeople use this to their advantage. In the wrong frame of mind, otherwise business-savvy people can make reckless bets or bad investment decisions.

Emotion can also disable rational decision making. People are often reluctant to accept information they don’t want to hear. Drivers regularly risk their lives to catch a yellow light, just like pilots have been known to fly into dangerous weather just so they won’t be late to their cousin’s wedding.

Many areas of practice take pains to correct for human irrationality. The scientific method uses the peer review process, for example, where any new finding is subject to testing by scientists without stake in a positive test result. When pharmaceutical scientists want to test the efficacy of a new cancer treatment, they submit their findings to scientists with no financial interest in the treatment. This independent review is designed to evaluate a hypothesis while avoiding the motivated reasoning or bias that might unconsciously ignore a side effect or exaggerate a result.

Our senses, instincts and intuition are powerful, but they aren’t perfect. They’re prone to bias, suggestion, and error. That’s why cars and planes have dashboards, not just windshields. Dashboards, and the instruments attached to them, provide the quantitative data we need to properly assess the world. And when our senses weaken or fail, dashboards help us make decisions. They complete the picture.

Dashboards and Key Performance Indicators - The Good, The Bad, The Destructive Incentives

Dashboards are everywhere in business. But instead of flight instruments and altimeters, we use metrics: numbers, formulas, and statistics designed to give us a simplified understanding of the world around us.

Investors track the economy by checking the Dow Jones Industrial Average or the 10 year treasury yield. These two numbers don't come close to describing the overall health of the economy, but they do tend to correlate with it, meaning that positive increases in either metric usually indicate increased confidence in domestic financial markets.

Some people don’t just use metrics to understand the world around them. They’re looking for an edge. Data scientists comb through data, mining for insights and looking for the formulas or measurements that help their clients maximize their competitive edge. In 2019, companies paid focused on big data and data science generated $189B in revenue. Each investor is looking for the next hot stock or investment opportunity. Metrics are what they use to find them.

Many organizations use metrics as performance targets. Companies regularly incentivize their salespeople in ways that reward positive metrics. They reward salespeople based on revenue targets, conversion rates (how many client conversations result in a sale) and profit goals. Metrics used to track or incentivize performance are called Key Performance Indicators (KPIs).

The goal of KPIs is straightforward: find metrics that correlate with success, and measure performance against them. It makes perfect sense, but there’s a catch.

In 1975, economist Charles Goodhart stated what would become Goodhart's Law:

"When a measure becomes a target, it ceases to be a good measure." - Goodhart, C. A. E. (1975). "Problems of Monetary Management: The U.K. Experience". Papers in Monetary Economics. I. Reserve Bank of Australia.

Goodhart laid out a complicated economic case for how and why this happens, but his thinking was simple: When people know they’re being evaluated on a metric, they model their behavior to maximize it, even when this behavior isn’t necessarily good for the organization.

Many companies use KPIs. But there are three common mistakes companies make when choosing their KPIs and designing their dashboards:

Metrics Mistake 1: Measuring the Wrong Thing

The most common mistake companies make with metrics is picking the wrong things to measure. Not every instrument on the pilot’s dashboard is useful in every situation. Too many instruments can be distracting. Tracking the wrong KPIs is easy to do, and can have dire consequences.

In the early 1980s, IBM famously measured programmer productivity by how many lines of code they wrote. A programmer who wrote 1,000 lines was considered more productive than one who wrote 100. In the days of punch-card programming this made sense. But as smaller personal computers began to take over, this KPI tended to reward bloated software that was slow to run and difficult to debug. An upstart Seattle-based company called Microsoft had a better idea. They held competitions for who could rewrite IBM’s software using as few lines as possible. This resulted in leaner software that required less memory and ran faster. By the end of the decade Microsoft was using this approach to push the limits of personal computers with the Windows Operating System. IBM was still struggling to jumpstart their mainframe business.

IBM’s mistake was leaving KPIs in place for too long. In an era where personal computers shipped with limited memory and speed, IBM’s total-lines-of-code KPI rewarded programmers for writing bloated software. It’s likely many programmers learned to “game” the KPI, writing needlessly long programs to trigger performance bonuses. Moreover, IBM’s emphasis on the KPI likely meant some of their more talented programmers went underappreciated. This is a classic example of a KPI measuring the wrong thing, in this case inadvertently rewarding inefficiency.

Metrics Mistake 2: Destructive Incentives

Wells Fargo is one of the oldest names in banking. Founded in 1962, they’ve played a role in financing and underwriting some of the most important investments in the nation’s history. In the late 1990s, Wells Fargo implemented an incentive plan that almost destroyed them.

In a massive cultural shift, Wells Fargo decided they no longer saw themselves as a bank, but as a consumer products company. They decided to see bank clients as customers, and bank services, including credit cards, home equity loans, and savings accounts, as consumer products. With this shift came a change in employee expectations, and one important new KPI.

Each bank employee was now expected to be a salesperson. Everyone at Wells Fargo was encouraged to sign up new customers whenever possible, and sell as many new products as possible to existing customers.

Early reviews were strong. In 2011 the Wall Street Journal ran stories on how Wells Fargo’s new direction was helping revitalize an industry still ravaged by the 2008 financial crisis. To many observers, Wells Fargo was an innovator.

This success was short lived. The addressable market for new bank accounts and credit cards wasn’t as big as Wells Fargo had hoped. Employees struggled to meet expectations. Before long, would-be-salespeople started opening new accounts without client permission. Many Wells Fargo customers suddenly had new credit cards or home equity credit lines they didn’t know about. Between 2011 and 2016 Wells Fargo employees opened more than 1,534,280 unauthorized deposit accounts and 565,433 credit card accounts. When the other shoe dropped, it dropped hard. Wells Fargo paid more than $3B in fines and fired over 5,300 employees.

For Wells Fargo, the cost of poorly chosen KPIs was severe. The bank’s overemphasis on customer and account growth, left in place when the market saturated, encouraged fraud. What’s more, the inflated metrics and KPIs stopped being useful as a measure of progress. While new customers and new accounts normally help a bank, fake accounts and inactive credit mean little. Wells Fargo’s fake customers didn’t pay fees, accrue interest, or buy houses. As Charles Goodheart predicted, the moment Wells Fargo decided to measure employees on account growth, account growth stopped being a useful metric.

Metrics Mistake 3: Tunnel Vision

When KPIs are poorly chosen, misapplied, or left in place too long, organizations can suffer. Not only are employees measured on the wrong criteria, but leaders can lose sight of their organization’s greater vision, chasing a number or statistic that might not correlate with long-term success.

One side-effect of poor KPI management is tunnel vision, a term borrowed from the medical condition when one loses peripheral eyesight. Overemphasis on one metric can prevent leaders from thinking beyond it, missing when other metrics signal trouble, or overlooking opportunities that might help the business but not improve the KPIs being measured.

Invariably, tunnel vision tends to promote short-term thinking at the expense of long-term vision. When a company is singularly focused on increasing profitability, they may overlook capital or operating investments that advance long-term success but reduce short-term profits. In the internet retail explosion of the early 2000s, many traditional retailers struggled with outdated KPIs like new store openings and sales-per-square-foot. These metrics were increasingly irrelevant in the era of online retail. Organizations unable to accept or understand shifts in their industry can fall victim to “fighting the last war”, chasing outdated metrics and missing their windows for transformation.

Leaders who fall victim to tunnel vision are like pilots staring at the dashboard. They may have a comprehensive understanding of airspeed, altitude, and heading, but miss out on the bigger picture. Sometimes the bigger picture reveals new opportunities for innovation, reinvention, or transformation. Sometimes it’s an oncoming storm.

The Lesson: Dashboard Thinking and Windshield Thinking

Good leaders pay attention to business metrics while keeping their minds open to the bigger picture. This kind of thinking requires practice, diligence, and an open mind. It also requires a commitment to training the practice of unstructured thinking, dedicating time to exploring ideas, observations, and instincts in a forum where traditional metrics, conventions, and assumptions are cast aside. Leaders who achieve in building this culture follow the example of Neil Armstrong. They take their dashboards seriously while never forgetting to look through the windshield.

There is no sure-fire art or science for designing metrics. Some leaders spend their careers searching for the magical formula, KPI, or incentive that will ensure excellence and guard against failure. It doesn’t exist.

A best practice with metrics is to change them often. Don’t let a KPI stay in place long enough to become institutionalized or “gamed” by bad actors. Maintain a wide set of performance metrics and, to the degree you use them as incentives, mix things up. Experiment with new rewards. Keep it fresh. Diversify. This is the key to quality dashboards that inform good decisions and provide useful information.

But the greater lesson when it comes to metrics - the true answer to the problems of false incentives and tunnel vision, is to think about your business in two fundamentally different ways simultaneously. This is the all important balance of Dashboard Thinking and Windshield Thinking.

For many leaders, Windshield Thinking doesn’t come naturally. So make time for it. Reserve hours or even days for unstructured, “blue sky” thinking with your teams. Encourage bold, unconventional ideas and explore what might make them work. Experiment with turning old habits and assumptions on their head. What are you missing? What are your competitors doing that you aren’t? What elements of your culture or processes are overdue for change?

Windshield Thinking can be powerful. Use it as a channel for brainstorming, imagining new possibilities, and experimenting with ideas that disrupt or challenge conventional ways of doing business. If your dashboards tell you everything is going according to plan, look out the windshield. Assess your competition. Reflect on how your industry is changing. If your dashboards tell you your teams are high-performing, reflect on your company’s vision. Are your teams incentivized to work collaboratively, to find better ways of doing things?

Just like Neil Armstrong used his dashboards and his windshield to land on the moon, modern leaders need the objective and the subjective, the quantitative and the qualitative. They need Dashboard Thinking and Windshield Thinking.