Feedback loops everywhere
Feedback loops are a part of your everyday life whether you know it or not. Some examples are thermostats, blood sugar regulation or financial markets. Understanding them can help you see their good and bad sides. It can help you decide whether to use them in your next project as a designer. It can also help you make better decisions in your day to day life.
Before exploring what the feedback loop is, let’s take a step back and start with some basics.
Why loops?
When you put your laundry in your washing machine and set the duration, you expect to get clean laundry when the machine’s finished. There’s a simple input (duration of washing) and output (laundry).
You just made an open loop.
It’s called open because the output (laundry, or its cleanliness in this case) has no say in what the future behavior of the system (washing machine) will be.
Now let’s fix that and add some mechanism that could tell the washing machine when the laundry is clean, like some kind of a cleanliness sensor.
We took an open loop and closed it. Meaning we took the output (cleanliness of laundry) and altered the input with it (washing time). The washing time will now be dictated by the state of the laundry itself. We can now call this mechanism a feedback loop.
This means that the washing machine can now keep running until the laundry is actually clean and not just for some predefined duration of time.
I’m sure that by now you can come up with other examples of feedback loops like the one with the washing machine.
Thermostats, fights, and games
A very similar example is thermostat and heating. We call these balancing feedback loops since their job is to balance the system to the desired state.
To add heat when there’s too little, to shut the heating down when there’s too much. To keep washing when the laundry is still dirty and stop the washing machine when the laundry is clean.
These balancing loops are sometimes called negative feedback loops since they’re acting in reverse to forces imposed on the system (if you open a window and the temperature drops, the thermostat causes the heating to turn on — that’s the opposite to what would otherwise happen).
Next, there are positive feedback loops which are often called reinforcing loops (because they’re reinforcing the original input).
Positive in this context means that the output change affects the input in a way that the future output is even more amplified. However, it doesn’t necessarily mean, that it’s a good thing.
Systems
Feedback loops can make huge and often unforeseen changes to large systems. Climate, financial market or human body are great examples of such systems.
Let’s look at some interesting cases:
- A study that describes how desertification supports further desertification.
- A series of papers suggesting that climate change is fueled by itself in a series of reinforcing loops — CO2 from burning fossil fuels causes warming that allows for more water in the atmosphere (water vapor is an even more potent greenhouse gas). The warming also leads to the melting of permafrost that contains a lot of greenhouse gases as well as further amplifying the original effect of CO2 input from burning fossil fuels.
- Paul Watzlawick has a great example in his book “Pragmatics of Human Communication”. People often interpret their behavior as a response. The thing is, it’s also a cause for a future response of the other person in the relationship. That often leads to misinterpretation of what’s the cause and what’s a response. And that makes communication even worse.
Arguably the most prominent example is an escalation. You hit me, I hit you harder — The thing is that in an escalation, the goal of the system isn’t absolute, like in the washing machine example (clean laundry), but relative (get ahead of the other side). Surely you can imagine how this can go wrong when we’re not talking about a school fight but two nations arms racing instead.
Another common example is a phenomenon called success to the successful — when you win a match, you get prize money that enables you to get better equipment or hire a better coach which improves your odds to win even more matches. You can apply this to money itself — the more you invest, the more you can reinvest.
In general, unregulated positive feedback loops lead to the collapse of a system since they’re forever amplifying the original input. That is not sustainable in most real-life scenarios.
Escalation and success to the successful are examples of system traps. There are many more and they can be caused by phenomenons other than feedback loops as well. The book Thinking in Systems by Donella H. Meadows explains them very well. I highly recommend reading this book if you’re into this topic.
There’s also a great video from Mark Brown’s Game Maker’s Toolkit about Feedback loops in games — it explains well how feedback loops can be used for curating gaming experience. From encouraging weaker players by offering them powerups to making the game more difficult for players that do too well. And many more. Check out the video.
Once you begin to see the feedback loops in everyday life, you can try to leverage them into your advantage as a designer.
Feedback loops in user experience
In the UX field, as designers, we often work hard to design the best possible solution for our users’ problems only to find out that they get used way less than we thought.
There’s a way to greatly support a feature, better yet, a certain behavior, by using reinforcing feedback loops we talked about earlier.
For example, let’s imagine you’re designing a simple note-taking app. You added the possibility to comment on some parts of the notes. Now you could be done right there and call it a day with the desired job to be done satisfied.
However, you could further support this behavior by showing the information about the comments in the note listing or e-mailing this info to relevant people.
This will lead to more people being engaged by the feature further supporting its future use.
We call this an engagement loop, however, it’s nothing more than an application of a positive feedback loop from earlier, remember? Having said that, some nuances come with this particular application like it oftentimes does with psychology-related disciplines.
Nir Eyal in his book Hooked took a close look at this phenomenon and described a hook model, a simplified process of getting your users hooked on some behavior by carefully crafting incentives as they use your product.
Like we explored earlier, unregulated loops can be very powerful. Sometimes, they can have fatal consequences. That applies to the digital world as well. In fact, these feedback loops can be so reinforcing when used effectively that they can create an addiction.
Wait… is this ethical?
Recently, patterns like infinite scrolling are in the public eye. They can be seen as an application of the hook model and can be harmful to the user’s well-being.
This proposed bill to ban patterns like autoscrolling or autoplay is evidence that addiction to digital products is a prominent topic in society.
As designers, we have the power to influence human behavior by crafting the interactions. We should do so knowingly and weigh in possible consequences.
Ethics in design is a topic for at least a book, and there’s a lot of articles online covering many angles. There's a great interview with Harry Brignull from darkpatterns.org that's definitely worth reading.
I believe UX is much more like a language in terms of what’s considered normal.
The more something is used, the more it becomes normal and accepted. That’s why every designer should be aware of the change he or she supports by choosing certain patterns to influence users’ behavior.
You can read a great article that says it's imperative that designers take responsibility for their design decisions.
So, how are you going to use feedback loops in your next project?
Sources
- Global desertification: Drivers and feedbacks — https://www.sciencedirect.com/science/article/abs/pii/S0309170812000231
- https://iac.ethz.ch/group/climate-physics/knutti/publications.html
- Pragmatics of Human Communication, Paul Watzlawick — https://en.wikipedia.org/wiki/Paul_Watzlawick
- https://en.wikipedia.org/wiki/Nuclear_arms_race
- Thinking in Systems: A Primer, Donella H. Meadows — https://en.wikipedia.org/wiki/Donella_Meadows
- Mark Brown’s Game Maker’s Toolkit — https://youtu.be/H4kbJObhcHw
- Jobs to be done framework — https://jtbd.info/
- Hooked, Nir Eyal — https://en.wikipedia.org/wiki/Nir_Eyal
- https://www.theverge.com/2019/7/30/20746878/josh-hawley-dark-patterns-platform-design-autoplay-youtube-videos-scrolling-snapstreaks-illegal
- https://medium.com/r/?url=https%3A%2F%2Fuxdesign.cc%2Fethics-in-design-an-interview-with-harry-brignull-from-darkpatterns-org-95489057c440
- https://uxdesign.cc/nazis-designers-dark-patterns-4c381b014b10
