Three months ago, I committed to something that seemed simple but turned out to be transformative: tracking every single penny I spent for 90 consecutive days in Beancount. As someone pursuing FIRE (Financial Independence, Retire Early), I knew that I couldn’t optimize what I didn’t measure—and I definitely couldn’t calculate my FIRE number (25x annual expenses) without knowing what I actually spent.
Why Measurement Matters
The FIRE community constantly preaches about savings rate and the 4% rule, but here’s what nobody tells you upfront: your estimated spending is probably wrong. Dead wrong. I thought I had a good handle on my finances. I used credit cards for everything, reviewed statements monthly, felt like I was “aware” of my spending. But awareness and measurement are completely different disciplines.
Before I started tracking in Beancount, I estimated my annual expenses at around $45,000 ($3,750/month). Seemed reasonable. I had a mental model: $1,500 rent, $400 groceries, $300 utilities/subscriptions, $500 dining out, $200 transportation, $850 discretionary. Clean math. Comfortable assumptions.
The 90-Day Tracking Challenge
I set up a simple Beancount structure: main expense categories aligned to my budget, a daily ritual of entering transactions before bed (5-10 minutes), and a weekly Fava review every Sunday morning with coffee. No complicated plugins. No over-engineering. Just disciplined, granular transaction recording.
The rules were simple:
- Every transaction gets recorded the same day
- Every expense gets a specific category (no lazy “Miscellaneous” dumping)
- Balance assertions every week to catch mistakes early
- No judgment, no restriction—just measurement
What 90 Days of Data Revealed
By day 30, patterns started emerging. By day 60, I was questioning my entire financial self-image. By day 90, I had undeniable evidence that my spending baseline was $6,550/month—not $3,750.
The blind spots were brutal:
- Subscription creep: $187/month across 14 services I “needed” (Netflix, Spotify, LinkedIn Premium, Dropbox, Adobe, Substack subscriptions, NYT, Audible, gym membership, meal kit trial I forgot to cancel…)
- Micro-transactions that compound: Coffee shops and lunch near work added up to $420/month. Each purchase felt like $8-12, but the monthly aggregate was shocking.
- “Hidden” monthly costs: Car insurance, renters insurance, Amazon Prime, quarterly software renewals, annual memberships—all averaged out to $340/month I hadn’t factored in properly.
- Lifestyle inflation: My “reasonable” dining out budget of $500/month was actually $890. Dinner with friends twice a week, weekend brunch, date nights… It adds up fast.
- The true cost of hobbies: Photography gear, hiking equipment, ski trips—I spent $2,200 over 90 days on “occasional” hobby purchases. That’s $733/month, not the $100/month I assumed.
The most painful discovery? My estimated FIRE number was based on spending $45K/year. My actual spending was trending toward $78,600/year—a 75% error. That changes my target nest egg from $1.125M to $1.965M. That’s an extra 3-4 years of work if I don’t course-correct.
Behavioral Impact: Awareness vs Restriction
Here’s what surprised me most: I didn’t change my behavior during the tracking period, yet I still spent 8% less than baseline by week 10. Just the act of recording the $14 lunch made me question it. Not every time—but enough times that I brought lunch from home twice a week instead of never.
I wasn’t budgeting. I wasn’t restricting. I was just aware. The cognitive load of entering “Expenses:Dining:Lunch $14.00” created a micro-moment of reflection that occasionally shifted my decision.
But awareness alone isn’t enough. Now that I have 90 days of clean data, I know exactly which categories to optimize:
- Cut subscriptions from 14 to 5: saves $120/month
- Pack lunch 3x/week: saves $150/month
- Reduce dining out from 8x to 4x/month: saves $200/month
- Pause hobby spending for 6 months: saves $730/month
Those four changes alone bring me down to $5,350/month ($64,200/year), much closer to my original target. But I only know this because I measured.
The Beancount Advantage
Plain text accounting made this sustainable. My workflow:
2026-03-27 * "Blue Bottle Coffee" "Morning coffee"
Expenses:Food:Coffee 8.50 USD
Liabilities:CreditCard:Chase
Five seconds to type. Completely transparent. I can grep my entire financial life. I can see trends over time. I have version control via Git (yes, I git commit my finances). When I’m 6 months in, I’ll have statistically meaningful data to model scenarios: “What if I move to a lower cost-of-living area?” or “What if I freelance part-time?”
The Big Question
Does exhaustive tracking drive behavior change, or does it just drive awareness?
After 90 days, my answer is: it drives awareness, which creates the opportunity for intentional behavior change. You can’t fix what you can’t see. Tracking makes the invisible visible. But change still requires decision and discipline—it’s not automatic.
The real power is this: once you measure for 90 days, you have your true baseline. No more guessing. No more self-deception. Just data. And from data, you can build a real plan toward financial independence.
For those of you who’ve completed your first 90 days (or are in the middle of it): What did tracking reveal that shocked you? Did awareness alone change your behavior, or did you need active intervention?
For the Beancount beginners: Yes, it’s worth it. Start today. Track everything. Commit to 90 days. You’ll thank yourself when you have real numbers instead of comfortable lies.