Measuring And Predicting Daily Satisfaction

From Personal Science Wiki
Jump to navigation Jump to search
Project Infobox Question-icon.png
Self researcher(s) John Cottongim
Related tools Fitbit, Google sheet
Related topics Productivity, Sleep, Activity tracking, Food tracking

Builds on project(s)
Has inspired Projects (0)
Show and Tell Talk Infobox
Featured image Measuring-and-predicting-daily-satisfaction.jpg
Date 2017/06/17
Event name 2017 QS Global Conference
Slides Measuring-and-predicting-daily-satisfaction.pdf
UI icon information.png This content was automatically imported. See here how to improve it if any information is missing or out outdated.

Measuring And Predicting Daily Satisfaction is a Show & Tell talk by John Cottongim that has been imported from the Quantified Self Show & Tell library.The talk was given on 2017/06/17 and is about Productivity, Sleep, Activity tracking, and Food tracking.

Description[edit | edit source]

A description of this project as introduced by Quantified Self follows:

John Cottingim is 35 years old. He lives in New York City and manages a process automation team. In this video, John discusses how he learned to track, measure, and predict his daily satisfaction by combining automation, visualization and machine learning without any coding.

Video and transcript[edit | edit source]

A transcript of this talk is below:

John Cottongim - Measuring and Predicting Daily Satisfaction

…About me, John Cottongim, 35 year old New Yorker. My day job is process automation at a big financial institution. That sound really technical but I assure I am absolutely not a coder. So everything I’ll talk to you today was done and can be done without writing a line of code. What I was able to do was create a system, a series of tools put together to track my personal data using these no-code solutions, that combine my device data, surveys and into a visual tool that was able to affect my outcome. I was also able to utilize machine learning to predict my primary endpoint, which for me, was improving my personal daily satisfaction. A little bit of background is probably helpful here. So early on, probably five or six years ago before I got into QS, I was a big skeptic. I really didn’t understand why people were wearing these trackers. I have several now, but I just didn’t get it. But I think it was because I was healthy and fit and I didn’t see the need to change. So definitely a skeptic, but you’ll see that I’ve gone quiet far the other way since then. So what happened? I really think that life happened. I became a bit overworked for the first time and I wasn’t taking the time to stay in shape that I wanted to and got a bit stressed. Ultimately, what I felt was I wasn’t as satisfied with the days like I wanted to be. So what did I do? First things first, I got one of these. Fitbit and started tracking and I got back what most of us get back probably as we first get into this journey, which was primarily steps. But I quickly realized that the data points that I was getting were really the inputs to what I wanted to improve upon. And I was receiving some advice, but at first it was really kind of vague and impersonal, like I walk more on the weekend which wasn’t really helpful. I knew that that was half the story, so what did that leave me? It left me still more or less in the same state that I was at before, and not spending the time I wanted to on myself. But now I have data. So of course I didn’t stop there. I did what I do best. I started to build spreadsheets and automate. How did I do that? With Google Drive and If Its This Then That which as we know is a fairly common combination. Sweeping my data into my Google Sheets, I knew I was still missing something. I now had these inputs, but what I was missing was my outputs. So did I feel healthy? Was I satisfied with my day? I needed to add those, so a combination of Google Forms and AppSheet allowed me to actually track those outputs. So what this solution looked like was on the left side of the screen, simple mobile application, a pretty basic design that swept my data over to the Google Spreadsheet, which eventually made its way up to about 103 inputs and outputs. Simple piece of advice here that I learned from myself, automating the reminders to track as just as important as tracking because it helped me add compliance to that data, and really reduce the mental load of me remembering to do this. Now that I had all this data I had to crunch it, so over a holiday weekend with very patient wife nearby, I taught myself enough machine learning to predict my primary endpoint, which is my satisfaction ahead of time. Visualizing all of this is absolutely critical. Use the tool to be able to create a dashboard or an update overnight with all of this data that I had sweeping into it, so I could see all my trends and habits in a single spot. Here’s a detailed view of my satisfaction. The grey bubbles are actually the machine learning prediction, hoped my satisfaction would be that day, and you actually see it diverging when I started to take action off of that. The last mile probably for me was getting this to a place where I would see it every day. And with the combination of tools on the screen, I was able to get this into application and it’s now living on my fridge where I make my coffee in the morning. It comes on when I show up, so it’s absolutely unavoidable. So what did I learn? Besides data visualization, and machine learning, and developing no-code apps which I didn’t have before and they are awesome tools to have. I learned that making it easy is absolutely critical. It now takes less that a minute a day for me to input this data as well as look at the outputs, and that automating the end to end process though difficult at first was definitely well worth it. I also learned that getting detail really pays off here, that analyzing my raw inputs like water intake and transforming those inputs to more predictive measures. Things like averages and threshold over a certain level. Those are really important and ultimately led to better recommendations for myself. Over time life does change so my inputs and the model, they all have to be up kept to a certain degree. But managing that keeps it relevant to me so without it, it would really slip. What actually drove my daily satisfaction? As you would expect, steps were a weak predictor for me. Managing my sleep better as well as taking time for hobbies which was interesting learning. Those are the things that really impact my daily satisfaction. I also learned because of the detail that I have some interesting things about my sleep and weight going up and down together as well as again back to that hydration. One day of hydration helps but actually the most correlation to satisfaction is my streak over the course of three days. I learned that in the community today it’s still hard to go beyond steps. I have nine different tools that came together over a year period. And frankly, the user experience is still not what I would have wanted. I looked out there for the single source solution for these things and there’s some out there, but the advice I think is still a bit vague and impersonal. Being either a glutton for punishment or for a good project, I’ve taken it on myself to try to build that platform. I call that Wejourn, and that combines a combination of crowd data, as well as allows personal journeys. Things like daily satisfaction or improving your health or lowering your stress. The app uses both tracking data as well as self-surveys and combines that crowd data. It’s out there in both marketplaces and you can see me outside at a booth here if you want to talk about that. Also, if you want to learn more about how I built my personal dashboard if you’re interested, maybe for yourself, I have a workshop going tomorrow afternoon so please stop by to that. Love to talk about the tips and tricks and some of the pitfalls that I’ve gone through over the past from building that.

As I mentioned outside of the booth to talk about Wejourn if you’re interested, and please feel free to reach out to my email on the screen there. And with that I’ll pause and take any questions.

About the presenter[edit | edit source]

John Cottongim gave this talk.