The 90-Day Expense Tracking Experiment: What the Numbers Actually Reveal

Tracking every rupee you spend for 90 days sounds dramatic and slightly unhinged. It's also one of the most effective money habits we know of — not because it's sophisticated, but because it isn't.

No fancy app. No spreadsheet with pivot tables. Just a plain notes app on your phone, one note per month, and the habit of typing the amount every time money leaves your account.

Here's the setup. Picture a hypothetical salaried professional in Pune — the numbers below are illustrative, but the pattern is one we see constantly. Rent is reasonable, the salary is fine, there are no EMIs, no credit card debt, no dramatic vices. Just a normal person with a persistent, low-grade feeling that money evaporates somewhere between payday and month-end. You know that feeling. You check your balance on the 28th and think: wait, where did it all go? UPI history exists, of course, but UPI history is a crime scene with no suspect. Rows of ₹149, ₹299, ₹99 debits. You glance at them and move on.

The experiment is simple: ninety days. Write it all down. See what the numbers say.

The numbers, in our experience describing this exercise, always have a lot to say. And they're usually a little rude about it.

Why run this experiment at all

The trigger is boringly normal. It's December, and our hypothetical tracker plans to put ₹25,000 into PPF before the financial year ticks over — except a comfortable ₹25,000 isn't sitting around. The salary is fine. The rent is reasonable. No EMIs, no dramatic vices. Just someone who can't find twenty-five grand and can't explain why.

That's the part that stings. If the money were going to a new iPhone every year, at least there'd be a villain in the story. But there's no villain. Just dinner ordered on Swiggy four nights a week, paid for with a cheerful UPI ting.

The usual advice — "make a budget!" — tends to die at the setup stage. Download an expense tracker, connect the bank, categorize everything, abandon it within a week. The 90-day experiment goes the opposite direction. No accounts to link. No categories to assign. No charts. Just a note titled "JAN" and lines like:

12 Jan — Swiggy, paneer tikka dinner — ₹289

That's it. Ten seconds, max.

The food delivery number that floors everyone

Three weeks in comes the first Sunday review — done out of curiosity more than discipline — and the food delivery line is the one that makes people put the phone down and stare at the wall.

In our illustrative example, the tracker is spending roughly ₹11,400 a month on Swiggy and Zomato. The pre-experiment guess? "Maybe ₹4,000." Maybe.

Nearly three times the guess. And it's never a dramatic binge — there's no single ₹2,000 order. It's ₹249 biryani here, ₹189 momos there, a ₹399 "weekend treat" that somehow happens twice a weekend. Small at the moment of ordering; large only in aggregate, which nobody ever looks at.

This is the thing nobody warns you about with delivery apps: the unit of spending is a craving, not a bill. You never see a monthly total unless you go hunting for it. So the brain keeps the number at "some food money," which is finance-brain for "whatever amount lets me sleep at night."

Here's the illustrative damage, category by category — the guesses versus plausible monthly actuals for this kind of profile:

Category	Guessed (₹/month)	Illustrative actual (₹/month)

|---|---|---|

Food delivery (Swiggy/Zomato)	4,000	11,400
Eating out (cafes, weekend dinners)	2,500	5,100
Transport (Ola/Uber + petrol)	3,000	4,900
Shopping (Amazon, Myntra)	2,000	4,300
Subscriptions (OTT, apps, iCloud)	800	1,650
Small UPI payments (₹99–₹499)	1,500	6,200

Look at that last row. Go on, look at it. The "miscellaneous" guesses are off by more than 4x. The small UPI stuff — the ₹149 top-up, the ₹99 chai on the way to work, the ₹349 "just one game" in-app purchase at 11 pm — adds up to ₹6,200 a month. The rent is the big scary number in most people's heads, the one they negotiate and track. The UPI drips are invisible, and invisibly, they're a second rent.

The total gap between guessed spending and actual spending in this illustration: roughly ₹20,750 a month. That's nearly ₹2.5 lakh a year of phantom spending — money that leaves without ever registering as a decision.

The UPI problem nobody talks about

Here's our theory, and you're welcome to disagree. Cash hurts to spend. Cards hurt a little. UPI barely hurts at all.

When you count out ₹500 in cash, you feel it leaving. When you scan a QR and your phone goes ting, it registers somewhere between "ordering a cab" and "liking a photo." The payment completes before your brain has finished forming the thought should I spend this. Plenty of people have paid entire restaurant bills mid-conversation without a single pause in the talking. That's convenient. It's also a budgeting nightmare wearing a convenience costume.

Tracking makes the pain real again. Not in a guilt way — the experiment only works without judgment — but in a visibility way. When you type "₹349 — random app purchase" into a note, you have to look at the words. Within about a month of this kind of tracking, the impulsive ₹149–₹299 payments tend to drop on their own. No rule required. It's hard to keep pretending once the pretending is written down.

The system that actually works

Let's be specific, because vague advice is useless. The entire system:

• One notes-app note per month. Titled "JAN", "FEB", "MAR". No folders, no tags, no colour coding. The phone's default Notes app.
• Log within ten seconds of paying. UPI payment goes through → open note → type Swiggy ₹289. Done before the delivery partner gets the order. The ten-second rule matters: "I'll log it tonight" is how every tracking attempt dies. You know it. We know it.
• A Sunday five-minute review. Not analysis — just reading the week's lines like a boring diary. Sunday evening, chai in hand, scroll through. That's when the patterns hit. Wednesday dinners. The third Amazon order in a fortnight. You start noticing your own loops.
• No categories at first. This is where most attempts go wrong. Twenty minutes setting up "Food > Delivery > Swiggy" turns logging into admin work. For the first month, log raw lines only. Categories, if you even need them, can come later.
• No budget, no targets, no judgment. The only goal is observation. The moment tracking turns into scolding, your brain starts "forgetting" to log things, and then the data is useless anyway.

That's it. Total cost: zero rupees. Total setup time: about as long as it takes to create a note.

The Sunday review is the secret sauce, by the way. Daily logging builds the data; the review builds the awareness. Sit there thinking "I ordered Swiggy six times this week — six! — and cooked once," and you can't unsee it. After one review, the notes app knows you better than you know yourself, and honestly it has opinions.

Once the patterns are visible, running the numbers through a free budget planner takes five minutes and gives the leaks a structure. Nice to have. Not necessary — the notes app is the whole engine.

Why most budget apps are overkill

Hot take time. Most budget apps are overkill for most people.

They want you to link bank accounts, auto-categorize transactions, set budgets per category, get alerts when you exceed them. That's a full-time job disguised as a solution. And here's the uncomfortable bit: the people who'd diligently maintain a budget app don't need one, and the people who need one won't maintain it.

What changes spending isn't a dashboard. It's friction — the tiny, ten-second friction of writing the number down. That friction is the feature. Automation removes it, which is exactly backwards: you want more awareness of spending, not less. A notes app gives you the perfect amount of friction and zero maintenance. Two minutes a day. That's the whole religion.

None of this means apps never help anyone — if you've got investments across Zerodha and Groww and you're tracking net worth, use a real tool. But for "why is my money disappearing," a note beats a dashboard nearly every time.

Three months later: the honest tally

Let's not oversell this. Nobody becomes a reformed minimalist from a notes app. The Swiggy orders continue. The occasional unnecessary Amazon purchase continues (a Bluetooth speaker shaped like a retro radio, ₹2,199, zero regrets — hypothetically).

But the numbers move, and they move without white-knuckling. In our illustrative example, food delivery drifts from ~₹11,400 down to ~₹6,800 a month — not from a ban, but because seeing "Swiggy ₹289" four times in one week makes the fifth order feel ridiculous. The small UPI drips drop from ~₹6,200 to ~₹3,400. Two forgotten subscriptions get cancelled along the way (a music app and some "pro" tier of a notes app — the irony writes itself).

Illustrative monthly savings from awareness alone: somewhere around ₹8,000–₹10,000. That's roughly a lakh a year. From a notes app.

One important caveat: we're not financial coaches, and this isn't financial advice. This is a method, not a prescription. Your leaks will be different. Maybe it's pay-later shopping. Maybe it's weekend trips. The method doesn't care what your leak is; it just makes it visible.

And the experiment doesn't have to run forever. A lot of people do one full 90-day run, then switch to a "spot check" month every quarter — one month of full logging, just to catch new leaks. It's remarkable what a quarterly spot-check turns up: the classic find is a ₹499/month subscription for an app deleted a year ago.

Start tonight. Open a note. Type today's spending from memory, best as you can. Then log the next thing you buy within ten seconds.

Your bank statement already knows the truth. It's time you read it.