The Best CEOs Read Their Business Backwards

Executives have never had more information about their companies.

Dashboards track revenue, margin, market share, customer satisfaction, employee productivity and dozens of other measures in near real time. Artificial intelligence promises to make the executive dashboard even more sophisticated, identifying patterns and producing

analysis before a leadership team has finished its morning coffee.

Yet more visibility does not necessarily produce more understanding.

Revenue can tell you what happened. It cannot tell you why. Customer churn can reveal that something went wrong. It cannot tell you when the problem began. Margin compression can appear neatly in a financial statement months after the decisions that caused it have already been made.

This creates one of the more consequential differences between reporting and leadership.

Most businesses are read forward: leaders establish a plan, execute against it and measure the resulting performance. The best executives also learn to read the business backwards. They begin with the outcome and trace it through the chain of decisions, behaviours and assumptions that produced it.

That sounds simple. In practice, organizations are surprisingly bad at it.

Consider a company that misses its revenue target. The obvious diagnosis is a sales problem. Perhaps the sales team needs more leads, better incentives or additional training. But work backwards and the picture may change. Revenue fell because conversion declined. Conversion declined because customers resisted the price increase. Customers resisted because the product's perceived value had weakened. Value had weakened because service levels deteriorated. Service deteriorated because an earlier cost-reduction initiative removed capacity from the operation.

The revenue miss appeared in sales.

The cause began somewhere else.

This distinction matters because businesses are increasingly managed through outcomes. Revenue must grow. Margins must improve. Customer satisfaction must rise. Productivity must increase. Those are legitimate objectives, but they are also the final outputs of complicated systems.

Managing the output without understanding the system can produce a dangerous illusion of control.

Economists have a useful concept for this. Goodhart's Law is commonly summarized as: when a measure becomes a target, it tends to stop being a useful measure. The problem is not measurement itself. The problem arises when organizations become so focused on moving an indicator that people begin optimizing the number rather than the underlying condition the number was supposed to represent.

Few corporate examples illustrate the danger more clearly than Wells Fargo.

For years, the bank emphasized cross-selling as evidence that it was building deeper customer relationships. The number of products per household became an important performance measure and part of the company's investor narrative. But intense pressure around sales goals contributed to employees opening accounts customers did not need or, in some cases, had not authorized. Eventually, the bank eliminated retail product sales goals and redesigned incentives around customer service, retention and long-term relationships.

The metric had been intended to represent something valuable: stronger customer relationships.

Eventually, improving the metric became more important than improving the relationship.

That is the danger of reading a business only from the dashboard forward.

The opposite approach begins by asking what must be true underneath a desirable result.

Amazon has practiced a version of this thinking for decades. Its early shareholder philosophy emphasized customer and revenue growth, repeat purchasing and brand strength, but it also committed the company to measuring investments analytically, abandoning initiatives that failed to produce acceptable returns and increasing investment in those that worked.

The distinction is subtle but important. Financial outcomes matter, but management attention belongs disproportionately on the inputs and decisions capable of changing those outcomes.

Revenue is an outcome.

Customers returning is a behaviour.

Why they return is a management question.

That final question is where much of the strategy lives.

Domino's provides another useful example. By the late 2000s, the company had built a formidable delivery operation, but consumer research exposed a more fundamental problem: taste had become a major reason customers were leaving the brand. Rather than treating weak performance primarily as an advertising problem, Domino's worked backwards from consumer behaviour to the product itself. It reformulated its core pizza and then built the famous "Pizza Turnaround" campaign around acknowledging the criticism and explaining what it had changed.

The marketing worked because the diagnosis came first.

Domino's did not merely ask how to sell more pizza.

It asked why people did not want the pizza it was already selling.

There is an important lesson here for executive teams.

When a result disappoints, organizations naturally look for the department whose name appears closest to the metric. Revenue becomes sales' problem. Margin becomes finance's problem. Retention becomes a marketing problem. Productivity becomes operations' problem.

But results travel through systems.

A pricing decision can become a retention problem. A hiring decision can become a customer-experience problem. A procurement decision can become a brand problem. A marketing campaign can become a capacity problem. A cost reduction can eventually become a revenue problem.

The executive's job is therefore not merely to ask who owns the number.

It is to understand what produced it.

This is not an argument against dashboards. Executives need outcomes. Lagging indicators provide an essential record of whether a strategy is working.

But a dashboard should be the beginning of the conversation, not the end of it.

For every critical outcome, leadership should be able to identify a small number of underlying behaviours or conditions that plausibly produce it. Then those assumptions should be tested continuously. If repeat purchases are expected to drive lifetime value, watch repeat behaviour. If faster service is supposed to improve retention, establish whether it actually does. If a new technology investment is expected to increase productivity, determine where time or cost is genuinely disappearing rather than simply recording that the technology was deployed.

This matters even more as Artificial Intelligence (AI) gives executives the ability to measure almost everything.

The risk of the next generation of management may not be insufficient data. It may be mistaking increasingly sophisticated measurement for increasingly sophisticated understanding.

A company can have perfect visibility into the wrong things.

The best CEOs know the difference.

They look at revenue and see the customers, prices, capacity and decisions underneath it. They look at margin and see the operating choices that created it. They look at customer churn and ask what happened months before the cancellation arrived.

They do not ignore the result.

They read backwards from it.

Because by the time a problem reaches the dashboard, the business has often been telling the story for months.

Read more by Janita Pannu on Forbes Communication Council