We’re all aware of the consequences of paying too much. No one wants to set up a long-term supplier agreement only to find another supplier would have done it for less. This sort of thing is so abhorrent, companies have a whole crew of experts who buy things for a living according to infinitely detailed business processes. We may end up buying nothing, but we will not pay too much.

The consequences of paying too much are clear, quantifiable and visible: The price you paid, minus the price you could have paid, times volume. But what about paying too little?

Some time ago, I worked at a big company that is much admired for its profit-driven approach. One day my boss called the group into his office for a spontaneous conference call. On the line was a corporate officer of our company who explained a big problem: The motors business wasn’t shipping enough motors to the appliance business, and more refrigerators needed to leave the factory—fast.

After a flight that night, I showed up at the motor plant the next morning to find a sea of exceptionally expensive experts crawling all over the place. Their verdict: Yield at final test was embarrassingly low, and no one knew why. The burn rate to fix it was going to be something special.

The unproven motor design with its attractive price tag was stacking up way too many impatient refrigerators. The motor was far too inexpensive to be profitable.

As any engineer would do, I asked for the process documentation, which was fully compliant with ISO 9000, but too thin to be helpful. Turns out it was too costly to define the details of the process. Again, we spent too little. To figure out what was going on, I had to create the documentation from scratch. As a manufacturing scientist from corporate, my hourly rate was far higher than that of the manufacturing engineer who should have been allowed to create it in the first place. Now, we would be spending far more than we should have spent at the start.

After digging in, I realized the automated assembly process required force control, but instead was using position control. When I asked why, I was told force control was too expensive. In the end, because the motor didn’t perform, the less expensive control approach was too expensive. To save money, we should have spent more.

In a rare moment of humanity, the executive VP said publicly his engineers talked him into using the new design. Later, he owned the decision. “In hindsight, I approved a motor design that achieved its performance by improving a second-order effect at the expense of a first-order effect.” He had not spent enough time thinking about the design.

In the moment, it’s difficult to be aware of the consequences of spending too little. The trouble with spending too little is you find out you spent too little after you spend too little. At decision time, there are guaranteed (and quantifiable) consequences of choosing the higher priced option, but there’s only a possibility of negative consequences for choosing the lower priced option. In my experience, these negative consequences can show their heads in far-removed refrigerator plants, so it makes sense to think through the many degrees of separation.

 You’ll get funny looks when you ask, but next time the choice comes down to cost, ask “Is it possible the lower cost option is more expensive?” There’s a real cost to spending too little, but only if you think there is.