Everybody says we should have goals. And yes, it's true that goals can focus our attention and motivate us to achieve more. SMART goals (simple, measurable, achievable, results-focused, time-bound) make sense.

Goals can also be a disaster.

 The risks of setting goals

One group of American researchers argues that goals should come with warning labels and be treated like prescription medicine, rather than over-the-counter solutions.

Here's the classic case. In the 1960s, Ford set a goal to build a car that would cost less than $2,000 and weigh less than 2,000 pounds. They needed it in two years instead of the usual three and a half.

The result: The Pinto. But the Pinto’s petrol tank was in a dangerous position. The engineers knew that, but they were focused on the company's goal. Drivers died in fireballs. Ford eventually paid millions in damages, had to recall 1.5 million cars and damaged its reputation.

We can develop blinkers when we concentrate on any specific goal. When researchers ask proofreaders to focus on either punctuation or factual errors, most people will stick to their assigned goal – and no more. Those with a less specific goal will spot both the punctuation lapses and the factual errors.

An alternative to narrow goals and KPIs

In our team, we don't have narrow goals, or even a list of KPIs. Our administrative assistant is our ‘Manager of Efficiency’. Another colleague is ‘a safe pair of hands for our finances’. Those are their informal titles and sum up the purpose of their work. That's enough.

Losing focus on other worthwhile goals is just one of the risks when goals are badly designed.

How about stretch goals?

Streth goals are a good idea, but goals that stretch people too far encourage cheating. Big firms have uncovered invoices for work that was never done. In an American bank 5,000 employees committed fraud to meet sales targets.

Goals that seem unattainable also sap motivation and encourage selfish behaviour, such as hoarding sales leads or professional knowledge.

The evidence suggests those American researchers who want goals to have warning labels have a point. So sure, let's have goals, but let's be cautious too.