top of page

The Sunk Cost Effect in BI Projects

1 July 2010
Keren Harel

Imagine the following scenario (based on a true story): It's a sunny Friday morning in March. You decide to walk around Tel Aviv and have breakfast at one of the famous cafes in the city. When you arrive, you see a long line of people stretching down the street, waiting to enter. The hostess tells you you'll have to wait about half an hour. You decide to wait (after all, it's a popular place). Half an hour passes, then forty minutes. After fifty minutes of waiting (remember, you're standing in the middle of the street under the scorching sun), you approach the hostess and ask how long you'll have to wait. The smiling hostess estimates you'll have to wait another fifteen minutes. You debate whether to wait or not and decide to wait - after all, you've already waited fifty minutes; what's another 15 minutes compared to the fifty you've already waited? Fifteen minutes pass, and still no table is available, and again, you're debating whether to stay or leave. After all, you've now waited for an hour and five minutes, you're at the top of the list, and indeed, a table will be accessible in just a few minutes. What will you do now?

Such situations happen to us every day, and we don't notice them - in line at the bank, waiting for the doctor, adding to a renovation budget (we've already invested 10,000 NIS in renovations, what's another 1000 NIS compared to that?). This phenomenon is called the Sunk Cost Effect (Arkes & Blumer 1985).

Before we understand the sunk cost effect and how it affects our lives, it's essential to understand its costs. The term sunk costs comes from the world of economics. Sunk costs are also known as stranded costs; these are costs that have already been incurred and cannot be "recovered," so they do not affect the future cash flow* of the project and decisions about the future. A classic example of sunk costs in a technology project such as a BI project is the investment of resources in research and development. When a company is deciding whether to implement one technological system or another, the feasibility study of the project will consider the value obtained from implementing the system (cost savings, response time reduction, process efficiency, etc.). The development costs that "created" that technology have already been spent, whether the project is realized or not, so they are sunk costs, and rationally, we should ignore them when dealing with the question of project feasibility.

So, what is the Sunk Cost Effect?

The sunk cost effect is our difficulty ignoring those sunk costs, which we should rationally overlook, as they are irrelevant to future decision-making. If we return to the cafe waiting dilemma, the time you've already waited is unrelated to the decision to continue waiting (you can't get the time back, and you won't be compensated for the long wait). What's relevant is how much longer you'll have to wait. It has been proven that we are not rational creatures, so it's hard to ignore that we've stood for 50 minutes under the scorching sun. So we tell ourselves: "We've already waited so long, we might as well wait a few more minutes..."

Where do we encounter the sunk cost effect in a BI project?

In checking the feasibility of the project

The feasibility check of a project is done based on the projected cash flow. Without going into calculations, there may be a situation where if we consider sunk costs in the feasibility calculation, the project will not be feasible to implement, even though this is incorrect. In this case, we will neither profit from executing the project nor recover the sunk costs. When the costs are a fait accompli, and the decision on the project neither adds nor detracts from them, they should not be included in the cash flow.

In project planning

Sometimes an organization will prefer to upgrade existing systems instead of purchasing new and more advanced systems to "save costs" ("resources have already been invested in existing systems, it's a shame to invest in new systems"). In practice, upgrading existing systems may cost more over time than purchasing new systems, and it's possible that after investing resources in upgrading, the existing systems still won't meet organizational needs, and it will be decided to purchase the new systems.

During the project

Sometimes, a project starts, and failures and problems arise that could not have been predicted during its course. The sunk cost effect will manifest in the tendency to continue with a failing project after money, effort, or time have already been invested and to try to justify its investment. The tendency to continue with losing and failing projects stems from the desire not to appear wasteful and mistaken, and it takes great courage to admit a mistake and stop a failing project. Another name for the sunk cost effect is the Concorde effect (Arkes & Ayton, 1999), named after the expensive airplane. It was clear from the start of its production that the project would end in a significant loss, but they continued to invest considerable sums in it and eventually grounded it. Something similar happened in Israel with the "Lavi" project.

In conclusion, the sunk cost effect meets us in every aspect of our lives and usually causes losses. In the best case, we'll lose half an hour of our lives in line for a doctor or waiting for a service representative, and in the less good case, we'll lose significant amounts of money (and if it's not our own money, then we might also lose our job...). It's essential to be aware of the existence of this effect, especially when we're leading large projects such as a BI project, and to understand that admitting a mistake may not be pleasant for a moment but, over time, can significantly minimize damage.

Good luck,

*Cash flow is an economic entity's cash movement in an accounting period. Generally, it is divided into three parts: cash flow from operating activities (for example, cash flow generated from sales to customers or cash flow used to pay suppliers), investment activities (for example, cash flow used to buy shares for long-term holding or cash flow used to buy a building) and financing activities (for example, cash flow generated from receiving a long-term loan).

A treasure chest under water
bottom of page