Examining the Sunk Cost Fallacy
Are sunk costs relevant to rational decision making?
As a software developer, it is very common to try to incorporate new technologies into your project, but after investing lots of time, you realize that it doesn’t make sense, would require too much effort, or simply doesn’t work. However, it is hard to step away from something you spent so much time on. To cut your losses now feels inefficient, and confirms that you just wasted the last few days, weeks, or even months. This is known as the sunk cost fallacy, and is an irrational assumption made by anyone that sinks a lot of time or resources into something, then discovers it isn’t going to work out how you thought.
The first issue that came up was trying to implement the lightning network. Before I get into that, I should first explain the nature and idea of our project. The final part of Fullstack Academy’s full time immersive web development bootcamp is to build a project, and we are given about three and a half weeks to do this. My group is focused on blockchain technology, and we decided to build a tipping extension and web application for some of the most popular social media sites. As a somewhat frequent user of Reddit, Youtube, and Twitch, among others, I realized that these sites lack a common way to send payments to your favorite creators. Twitch streamers have donation links, but there are many different ways that they implement them, and using those have their own issues, such as having viewers that donate money then, immediately demand a refund. With Reddit, a user can award a post with icons, but these icons are worthless to the post creator and have no actual monetary value. Some popular Reddit posts have icons that, grouped together, cost upwards of $50 or $100, but that doesn’t matter at all to the creator because they can’t be monetized. Popular Youtube creators commonly have Patreon accounts, but again, those are hard to access, have their own downsides, and make it difficult to support individual videos or do one-time donations. Anyhow, you get the point. Topper fixes all of that.
As a soon-to-be graduate of the September cohort at Fullstack Academy, my peers and I have been working around the clock to finish our final project, and blockchain technology is a relatively new industry. As a result, the technical support and documentation isn’t always what you want it to be. Youtube usually has all the tutorials you might need to learn something, and your questions have often already been answered one way or another on Stack Overflow, but as we have been working with this technology, we have discovered that this isn’t the case anywhere near as often as we would like. As we have been working on Topper, our project (and soon to be product) over the last few weeks, we spent days trying to debug or make a certain technology work, only to decide that it would be too difficult to implement. We realized this early enough each time it happened to be able to move forward and not waste too much effort, but that isn’t always the case.
The first obstacle that we encountered with Topper was whether to try to use the bitcoin lightning network. The lightning network enables very, very fast transactions with nonexistent transfer fees, and this made sense for our project. We expected that users would be sending many small transactions, so therefore the lighting network made a lot of sense because that is its exact use case. However, when we tried to actually implement that, it was a different story. After looking into a myriad of different solutions and articles, we could not figure out how to even create a lightning node, and sunk two days into trying to get it implemented. It was hard to realize that we had to let it go, but it was the most rational choice. Moving forward, it would have been irrational to justify that we should have kept trying to make it work simply because we had already spent a lot of time in the past on this issue. It is important to realize this so that anyone, not just computer software engineers, waste as little time as possible finding the most efficient solutions to whatever they want to do.
Another fascinating experience our group had with the sunk cost fallacy while building Topper was the transaction process. Trying to get the blockchain browser technologies, Fortmatic and Metamask, to communicate in the way we wanted was a huge challenge. Metamask injected variables and methods into the chrome browser window, but we needed the information from those processes in the extension window so that we could relay that information back into our application. After agonizing over this for multiple days, the solution was frustratingly easy; all we needed to do was create a file that injected another file onto the page. That page had access to what we needed, and was able to set the variables in local storage, and from there, that could be sent over to Chrome storage. This solution was so much more simple than the things we had been trying previously, but ultimately it was better code and worked. This is a case where the sunk cost fallacy did not apply because we would not have been able to find the solution without putting in the time to find the other work-arounds.
We often hear people say, ‘cut your losses while you can’, and ‘let bygones be bygones’. It is hard to accept that the time or money that you spent cannot be regained, even if you try to push forward and make it work. It’s difficult to know when to push forward verses when your attempts are futile. Doing this will lead to better time management, an intuition that applies to the process, and more success in your professional career, whatever that may be. If you spend money to go to a sporting event, but find that it is boring and you could be spending your time doing something more enjoyable, don’t sit through it just because you already spent the money. My opinion is that one should accept that it is in the past now, cannot be changed, and can only be improved by moving forward and trying to learn from it. Thank you.
Checkout out Topper: https://topper-fsa.herokuapp.com/