The number was $812. Not $300, which is what the budget said — the tidy line item Priya and Matt had agreed on two years ago and never looked at again. Eight months of actual credit card statements, actually categorized, said their real restaurant spending averaged $812 a month. More than their car payment and groceries-for-one combined. Almost triple the number they’d been telling each other, and themselves.
Priya and Matt are a composite Ottawa couple — more on that at the bottom — but their numbers are painfully ordinary. Both around $70,000 a year: she’s a policy analyst with the federal government, he’s at a Kanata software firm. Combined take-home of roughly $8,600 a month. Rent in Westboro, one car, a cat, no kids yet. On paper, comfortable. In the “Wants” cell of a spreadsheet they made in 2024: Restaurants — $300.
The night of the confession
The categorization happened the way these things usually do: not out of discipline, but out of a fight that wasn’t really about money. Matt made a comment about a $94 Uber Eats order. Priya pointed out his daily shawarma runs. Somebody said “fine, let’s actually look,” and suddenly it was 11 p.m. and they were sitting at the kitchen table with eight months of statements, November through June, doing the thing they’d both quietly avoided for two years.
Here’s what eight months actually looked like: November $743, December $1,109, January $655 (the annual austerity mirage), February $702, March $818, April $779, May $861, June $829. Total: $6,496. Average: $812 a month.
And here’s where it went, on average:
| Where it went | Monthly avg |
|---|---|
| Lunches near their offices | $217 |
| Sit-down dinners (ByWard Market, Elgin, Wellington West) | $198 |
| Delivery apps (Skip, Uber Eats) | $164 |
| Coffee shops | $126 |
| ”Too tired to cook” weeknight takeout | $107 |
| Total | $812 |
No single line is scandalous. $217 in lunches is two people buying lunch twice a week each. $126 in coffee is a latte habit, not a crisis. That’s precisely why the total stayed invisible: it arrived in $16 and $23 increments, across two cards, and never once felt like $812.
The fight that night wasn’t actually about the money. They could afford $812 — the math below proves it. The fight was about the shame of not knowing. Two competent adults who negotiate salaries and file their own taxes had been carrying a number in their heads that was off by 170%, and each privately suspected it and said nothing. “I felt like we’d been lying,” Priya said, “except nobody had lied. We just never looked.”
You don’t have a spending problem. You have a visibility problem.
This is the thesis, so let’s say it plainly: most people don’t have a spending problem — they have a visibility problem. You cannot decide whether $812 a month on restaurants is fine, reckless, or actually too low for two working adults who love food, until you can see the number. Priya and Matt weren’t overspending against their values. They had no idea what they were spending, which meant they had no values position at all — just a fictional $300 that made the spreadsheet feel finished.
They resisted fixing it for the classic reason: “we’re not spreadsheet people.” Fair. Categorizing eight months by hand took them most of a night and one argument. It’s exactly the job the Smart Budget Importer exists to eliminate. The flow, if you’ve never seen it: a drop zone that takes bank statements as CSV, TSV, or PDF — one file or several at once — auto-detects statement formats from 18 banks, and maps any other CSV with a quick column-matching step. There’s a “try with sample data” button if you want to see it work before trusting it with your own statements, and a paste-your-text fallback. Everything runs in your browser; nothing is uploaded. One honest caveat: the AI-categorization layer is opt-in and off by default — the deterministic rule-based categorizer does the bulk of the work, and you only need the AI reader for things like scanned photo statements. You’ll still correct a few categories by hand. It’s minutes, not a night.
We ran their real numbers through the budget tool
We rebuilt Priya and Matt’s post-confession budget in our own Budget & Surplus Router — live, real outputs, no rounding for drama.
Inputs: $8,600 monthly net income. Needs: rent $2,150, groceries $780, utilities $310, transit and car costs $540, insurance $260, minimum debt payments $310 — $4,350 total. Wants: restaurants and delivery $812, shopping $250, entertainment $130, subscriptions $95, gym $110 — $1,397 total. Savings: $500 in workplace RRSP contributions.
What the tool said: Needs came in at 50.6% against the 50% target — “on track.” Wants: 16.2% against the 30% target — “under.” Savings: 5.8% against the 20% target — “under.” Monthly surplus: $2,353.
Read that middle line again. With $812 a month going to restaurants, their wants are still fourteen points under the classic 50/30/20 band. The tool never flags the restaurant line. The uncomfortable number was never the problem — the not-knowing was, and so is that 5.8% savings rate.
The surplus router is where it gets useful. With their roughly $10,000 emergency-fund gap entered, the router’s first step reads: “Liquidity first: your emergency fund has a $10,000 gap. At $2,353/mo you’ll be fully funded in 5 months.” Once that gap closes, it reroutes: $583/mo to the TFSA (filling this year’s $7,000 room) and the remaining $1,770/mo to non-registered investing. Five months of visibility buys them a funded safety net; a year buys a maxed TFSA. None of it requires cancelling a single dinner on Elgin.
An opinion you’re welcome to fight about
Here’s ours: budget apps mostly fail because they demand daily attention, and daily attention is the first thing real life takes back. You categorize diligently for three weeks, miss a busy stretch, the backlog becomes a guilt pile, and by February the app is abandoned. A once-a-quarter statement autopsy — drop the files in, correct a few categories, stare at the real totals for twenty minutes — beats a daily app you’ll quit, because the goal isn’t monitoring. It’s visibility, a few times a year, honestly.
And the counter-note this article owes you: sometimes the number is fine. $812 on restaurants for two working adults who genuinely love eating out can be a completely values-aligned choice — it’s under 10% of their take-home, and their budget absorbs it with $2,353 to spare. The point was never that $812 is shameful. The point is choosing it on purpose instead of discovering it at 11 p.m., mid-argument.
As of this writing, Priya and Matt still haven’t decided what their restaurant number should be. They’ve floated $600. Matt argued for keeping $800 and cutting delivery specifically. Priya wonders if the real answer is “whatever it is, as long as we look quarterly.” They know the surplus, they know the trade-offs, and they still can’t name the right number.
Maybe that’s fine. Maybe that’s the whole win: they’re finally arguing about a real number instead of a fictional one.
Priya and Matt are a composite, built from common patterns in Canadian household spending — the figures, statement math, and tool outputs above are real and were run live through our calculators, but no actual couple’s statements were used. If their kitchen table feels familiar, that’s the point.