What I Learned From a Year of Daily Metrics
TL;DR: By documenting key metrics, I came to better understand the consequences of my actions. I also gained an appreciation for the power of habits and persistence.
“Commit to documenting it—no matter what.” That’s the deal I made with myself in late 2018. I agreed that I’d collect the data—be it good or bad. Equally important: I made no commitment to take any other action. I figured that by keeping this simple, I had little excuse to not stick with it. Soon, it also spurred action.
You can collect personal metrics with many tools. The following are the ones I use.
- Calorie counting on iOS and web app: myfitnesspal
- Physical data (heart rate, activity, sleep): Garmin Fenix 5X
- Weight, body fat, BMI: Fitbit Aria
- Metrics documentation: Numbers spreadsheet
These tools aren’t perfect, but they aren’t bad, either. Myfitnesspal contains a lot of inaccurate calorie counts. This requires me to compare items/sources. Wrist-based heart rate tracking also isn’t super-accurate. That said, it’s more convenient than pairing with a chest strap.
Data-syncing can be clumsy. For example, my scale doesn’t talk to Garmin Connect. Still, I want accurate weight data in there for calculating exertion. So, I manually insert my weight into Garmin Connect, each morning.
I also started collecting all my data in Google Sheets. Displaying this data got sluggish, though. This led me to switch to Numbers, which I find snappier.
I also created some workflows in Alfred to speed up the process. Now, I type “gar” into Alfred to rapidly open Garmin Connect for pulling data. I do the same with myfitnesspal, Stripe, and Google Analytics. (I know I could import this data automatically, but have my reasons for not doing so. I’ll get to these in a moment.)
What I measure
Since starting I’ve changed which metrics I collect. For the most part, though, they sort into a few key groupings. These are:
- Activity (daily pushups and situps, stretching, exercise)
- Condition (ARHR, sleep, VO2m, weight, trajectory)
- Nutrition (meal quality, fasting, calories in/out, deficit)
- Personal habits (reading to kids; alcohol intake)
- Revenue (passive, billings, 30-day averages)
- Startup activity (installs, MRR, outreach, traffic)
Each day gets one row, with a column for each of the items noted above. I also keep one cell for notes. Most times this is empty but does come in handy at times. For example, if I’m away from home, I’ll use my most recent 7-day weight average. I then write: “Away; weight estimated” in the Notes cell.
This experience might be the longest personal experiment I’ve ever undertaken. Over the year, I learned a lot from it and collected some notes. The following are ones that stand out. I’ll return to this list and add to it, should other points come to mind.
One habit leads
For me, collecting metrics is a keystone habit (Charles Duhigg talks about this in The Power of Habit). So, I document metrics near the beginning of each day. I seldom miss a day. In the odd case that I do, I’m ill-at-ease until I’ve filled in those cells.
This is telling. Metrics are in part about establishing positive habits and removing negative ones. For me, this makes measuring my habits my most important habit—and I stick to it.
My first successful metrics experience related to finances. Ours weren’t where they should be, so I set aside a weekend (~4 years ago) to create a better system. I collected every expense from the previous 6 months, sorted them into groups, and averaged them. This allowed me to see our monthly expenditures. It also helped me predict costs for upcoming months.
This resulted in a finances spreadsheet. Every day, I manually document our bank and credit card transactions in it. Every month, I copy-and-paste the monthly template for the next month. This process allows me to track our accounts for anomalies. It helps me spot potential shortfalls. It also indicates when I can set aside surplus funds for savings.
Could I automate most of this with personal finance software? Yes. But, it’d distance me from the process. That extra step of documenting expenses manually only adds a few minutes to my day. In exchange for that time I gain a sense of control that I value.
This approach also allows for customization. For example, I fill cells with positive actions (going for a run) or key milestones (a new MRR record) in green. This sounds trivial, but it’s not. These act like gold stars and provide tiny dopamine hits that affect my behavior.
I can predict the future
Much like I can predict our finances by documenting patterns, I can do so with personal matters. I once wrote that you shouldn’t step on the scale. I was wrong. A better position would have been: Don’t let today’s measurement derail your progress.
My weight fluctuates up to five pounds over 24 hours. This is in part because I practice alternate-day fasting. (I recognize that fasting seems extreme to some, but I find it manageable and productive.) Daily scale readings aren’t that informative. So, I tally the last week’s readings and divide the sum by 7. Doing so gives me my 7-day average. From this I see a pattern.
So long as I practice the same behaviors, my weight tends to follow a clear trajectory. This also allows me to calculate daily average weight loss. From this, I can extrapolate my approximate weight a week, month, or 6 months down the road. This isn’t an exact science, but it is quite useful.
Data doesn’t lie (sort of)
Last fall my weight loss hit a standstill. This frustrated me. Around the same time, a friend noted that weight loss involves plateaus. (I also found mention of this on Reddit.) So, I shrugged my shoulders and thought to myself, “I’ve plateaued. I guess this is as far as I go.”
A few weeks later, though, I took a closer look at my metrics. It turned out that I had gotten a little lax about food intake. In fact, I was up by around 500 calories a day. This wasn’t enough for me to put on weight—but it also wasn’t enough for me to drop weight.
I don’t want you to think that I’m obsessed with weight loss. Yet, it’s a useful way to examine how habits and tracking work—because it allows me to measure outcomes so easily.
We often make up reasons for why something doesn’t work. Data helps me overcome these preconceptions. It also sets me up to better assess what’s actually happening.
Measurement spurs action
I make a point of implementing a couple of small rituals aimed at creating more time with my kids. One is a nightly bedtime story. The problem is that by 9:00 PM, I’m beat. This leads me to put off reading, reasoning that we can do it the next night. The next night is often the same.
I know the boys are getting old for bedtime stories. I’m glad they take part in them now but recognize they won’t for much longer. So, I track each night I read to them. Doing so is a small act, but it serves as a daily reminder about whether I’m making time for something that matters. The longer I track such things, the more I tend to see my behavior shift.
Streaks have limits
A lot of folks believe in the power of streaks. This practice involves setting a habit and tracking its completion each day. Most times folks use a streaks tool or calendar to do this. The plan is to create a habit by not breaking the streak. I like this idea but it doesn’t work for me.
My problem with streaks is that breaking one is demoralizing. It is so much so that it stops me from starting another. I felt this most after 104 days of abstaining from alcohol. We were at a fundraiser, and I thought ”I’ve been good. Why not treat myself to a glass of wine?” The next day, I had to start my count again from 1. Bummer! Actually, more than that—I felt like I had thrown away all that discipline.
So, I adapted streaks to work better for me. Now, I track every day I don’t drink over the year. This means that if I slip one day, it’s OK. I haven’t lost my count for the year. I enjoy the break and start again the next day. This reduces the sense of failure and lowers the emotional cost of restarting.
Exercise works faster than I expected
I always figured that you needed to do a lot of exercise before you saw results. To my surprise, I found the opposite this year. For example, a week of short runs every morning can drop my average resting heart rate (ARHR) by 10 beats a minute. On a bit longer of a timeline, I saw my VO2 max increase by 15% over six months with moderate activity.
These aren’t super-impressive numbers. Yet, they make me optimistic about how much I could achieve if I applied myself. I’m also quite excited by the gains I saw by adding a small amount of interval training. This seemed to pay off within only days.
I started out thinking that my metrics collection method needed to be perfect. In fact, this idea got in the way of me starting. Turns out it’s not the case. The act of collecting metrics is the important part. Once you start you can retool as you see what works or doesn’t.
For example, I tracked daily steps for a long time. I didn’t find this information that useful (because it doesn’t accommodate my cycling or skiing.) I also tried to make a habit of cold showers and drinking matcha tea. Neither of these stuck either—so, I removed them.
I figure it’s reasonable for a document like this to evolve. I do version the file when I make significant changes, though. This allows me to roll it back if I change my mind.
Persistent action delivers
Collecting metrics taught me that small matters add up to significant outcomes. These are for both the better and for the worse. You don’t put on weight through one night of binging. You do so with just one beer a night.
Think of this way: On a sedentary day, I burn ~2,200 calories. That’s not a lot. If I manage to stay under that number and add one beer a day, I’ll quickly find the number on my scale moving upward. But, removing 200 calories from my allotment of 2,200 has the opposite effect. So, small daily actions make a difference.
After my experience last year, I came up with a personal mantra of sorts. It reads: clarity, persistence, measurement. First I need to know what I want. Then, I need to continually practice actions aimed at making that happen. Finally, I must measure these actions and results to determine if they work as desired.
Less might not be more
I tracked my personal metrics for all of 2019. In November, I realized that I no longer needed to do so, and took two weeks off.
During this time, I collected a simplified set of 5 data points. This was faster than my more comprehensive process—but it didn’t work. I lost my grounding, dropped good habits, and added bad ones. So I returned to my previous document. I then filled in the previous weeks’ missing metrics.
By late December I got excited about a new year of collecting metrics. I set aside some time and looked at what worked and what didn’t. I then created a new document for 2020. It addresses the shortcomings of the previous document.
Tweaks for 2020
I continue to track metrics every day, as part of my morning ritual (I’ll write about this ritual in an upcoming post). For 2020, I’ve simplified some of my metrics. For example, I used to apply a numeric measure for the quality of my meals. I’ve simplified this to whether it’s healthy or not.
This change is part of my effort to downplay negative actions or slips—and instead, focus on what I’m doing right. I suspect positive reinforcement will be more powerful than the alternative.
I’m also playing more with cumulative measurement aimed at changing habits. For example, I used to measure exercise mileage and duration. This didn’t do much to motivate me; it was too much information. So, I leave this data in Garmin Connect, where I can view it if necessary. For my daily metrics, I only tally how much I’ve done of each activity, for the year to date.
I learned this trick from a friend who’d completed a hundred bike rides in 2019. To me, this seemed smart: The duration of a ride isn’t as important as going for a ride. So, each time I stretch, mountain bike, nordic ski, downhill ski, run, or go to the gym I collect another point for that activity. This is surprisingly motivating.
When I first started tracking metrics I found it inconvenient and a bit silly. I reasoned that I saw so little change, I shouldn’t bother. Now, you couldn’t make me stop.
Metrics only take me 5 – 10 minutes each morning. In exchange, I get a record of past actions and a way to predict the future. It also reinforces the fact that small actions add up to significant improvements.
If you wish to make a change in your life, you might find success in starting with measurement.
I’m @karj and the above is just my opinion. Looking for more? Here’s a full list of articles and information on my books. This is what I’m doing now, and what I don’t do. I’d love it if you tried Emetti on your website!
Sign up to receive new articles by email:
There’s no charge to subscribe, no spam, and I won’t sell your contact information. You can unsubscribe from this list at any time, with one click.