Every budgeting app opens with the same ask: connect your bank. And a lot of perfectly rational people stop right there — because they don't want a third party holding a connection to their accounts, because their credit union isn't supported, because a chunk of their month runs on cash, or because handing over a bank login is a strange thing to do in the first five minutes of trying an app.
Here is the part the sign-up screens undersell: none of that stops you from budgeting. People ran exact budgets for decades before bank feeds existed. What replaces the feed is a small system — four numbers to set up, a logging habit with a trigger, and a five-minute weekly check — and this post walks through the whole thing, including what it honestly costs per week.
A linked budget collects itself: transactions appear, categories fill in, balances stay current without you. Going manual means you become the feed — and that trade has a real shape:
If that trade reads as acceptable, the system below makes it cheap. Each step exists to kill a specific way hand-tracking usually dies.
List the accounts that touch your month
Checking, savings, the cash in your wallet, and every credit card you actually use — each becomes an account in your budget app or ledger, with its balance as of today. Get the numbers from your bank app or last statement; this is the one time you have to copy anything over. Debts go in as what you owe, and they subtract: the point of the list is your real net position, not a feel-good subtotal.
Start with a handful of categories, not thirty
Housing, food, transport, fun, everything-else is a real budget; a 30-line category tree is a data-entry job you will quit by the 12th. Every transaction you log has to land somewhere in two seconds, without a taxonomy decision. Split a category later when its number gets big enough to ask questions about — that is the only signal you need.
Log on a trigger, not from memory
The system lives or dies here, and the fix is mechanical: attach logging to a moment that already happens. Tap it in while the card terminal is printing, while the receipt is in your hand, while you are still in the parking lot. Batch the predictable stuff instead — rent, subscriptions, insurance get logged on payday in one sitting, so daily logging is only the variable purchases. Memory is not a tool; end-of-day recall reliably loses the small stuff.
Reconcile against the bank once a week
Once a week, put your tracked balance next to what the bank app says. Match? Trust confirmed, done in a minute. Gap? The difference is a specific number, and a specific number is findable — it is one or two unlogged purchases, almost always from the day the routine broke. Log them, balances agree again, and the ledger goes another week being a thing you can act on. This five-minute ritual is the entire difference between hand-tracking that works and hand-tracking that quietly rots.
Treat card payments and transfers as moves, not spending
Paying the credit card from checking is not an expense — the spending already happened at the store, and it is already in a category. The payment is a transfer: checking goes down, the card debt goes down, and no budget line is touched. Log it as spending instead and the month double-counts; it is the classic hand-tracking error and the usual reason a manual budget "doesn't add up".
Close the loop with a monthly review
Once the logging and the weekly reconcile are habits, the numbers are trustworthy enough to act on: once a month, read the net, find the biggest gaps between plan and reality, and change one thing. Twenty minutes with a fixed agenda — and because your data is hand-fed, step one of that review is already done.
Step six's agenda is its own post — the 20-minute monthly money review — and step two's starter categories can be sketched with the free envelope budget planner before anything touches an app.
Here is the whole setup for a typical month — four accounts, four balances copied from the bank app, one honest net position:
| Account | Type | Balance today | Counts as |
|---|---|---|---|
| Checking | Bank Account | $1,240 | You have it |
| Savings | Bank Account | $3,800 | You have it |
| Wallet cash | Bank Account | $60 | You have it |
| Credit card | Credit Card | $420 | You owe it |
| Net position | $4,680 | ||
That $4,680 — what you have minus what you owe — is the number the whole system protects. Ten minutes of copying balances, and it exists for the first time.
One week later, the checking account's ledger looks like this — starting balance $1,240, five entries logged:
| Logged this week | Amount |
|---|---|
| Paycheck | +$1,120 |
| Rent (first half) | −$850 |
| Groceries — two trips | −$137.40 |
| Gas | −$42 |
| Streaming renewal | −$15.99 |
| Tracked balance | $1,314.61 |
| Bank app says | $1,296.16 |
| Drift | $18.45 |
The ledger thinks there is $18.45 more than the bank does — so $18.45 of spending went unlogged. Scroll the bank app's week and there they are: Tuesday's coffee ($4.60) and Saturday's parking ($13.85), together exactly $18.45. Log both, the balances agree, done — the whole check took five minutes, and the error lived for six days instead of six months. That containment is the entire trick of hand-tracking: you are allowed to be imperfect, because the reconcile keeps imperfect small.
You can run this system in a notebook. An app mostly buys you arithmetic — and in Spense, manual is a first-class path rather than the free tier's consolation prize:
And if the typing ever stops being worth it, connecting a bank takes two minutes — Spense asks what to do with the manual accounts you built, and even then your bank password goes to the bank's own login through Plaid, never to Spense. Whichever direction you pick, the thinking behind the budget itself is the same — it's laid out in the Spense Method, and not one of its four rules requires a linked bank.
For most people, no — reputable apps connect through aggregators like Plaid using read-only access, and the budget app itself never sees your bank password. "I would rather not" is still a complete reason: you might not want a third party holding a connection to your bank at all, your bank or credit union might not be supported, or you might simply not want to hand over a login on day one of trying an app. Manual tracking means that choice costs you the typing, not the budget.
Logging a purchase takes seconds on a phone — call it 20 seconds to be safe. At 5 variable purchases a day that is about 12 minutes a week, plus a 5-minute weekly reconcile: roughly 17 minutes a week all-in, with recurring bills batched on payday. People who quit manual tracking rarely quit over the total time — they quit because they tried to log from memory at 10pm instead of on a trigger.
Nothing dramatic — this is what the weekly reconcile is for. Do not reconstruct the lost days from memory; open the bank app, log the transactions it shows that you missed, and move on. The gap between your tracked balance and the bank's number tells you exactly how much went unlogged, so the error cannot silently grow.
Yes. A cash-based budget is just the manual system with one account called cash and stricter envelopes. Tracking apps that support manual accounts handle this fine — the "account" is the drawer, the balance is what is in it, and the weekly reconcile is counting it.
You lose the sync features — automatic transaction import and hands-off categorization — which is a real cost, honestly counted in the steps above as the logging habit and the weekly reconcile. In Spense specifically, everything downstream of the data works identically either way: budgets, spending limits, monthly reports, goals, transfers. Manual accounts are a first-class path, not a limited tier.
Yes, and it is a sensible sequence: hand-track for a month to learn your patterns, then link when the typing stops being worth it. In Spense, the first time you connect a bank it asks what to do with the manual accounts you already have — keep them alongside the synced ones, or start fresh. Nothing forces the choice early.
Spense opens straight into a working budget — no email, no card, no bank login. Add a manual account and go; 14-day free trial, and linking stays optional.
Keep exploring envelope budgeting and financial wellness
Plaid typically checks a bank one to four times a day, and purchases take days to post. Why synced budgets lag, why pending rows move, and what a new bank adds.
Start by measuring last month, not by cutting back. A first week in seven short jobs, one month sorted into nine categories, and four rules for transactions.
Let the bank collect, let the app sort, and read the month once. What a useful monthly summary shows, and the by-hand route if you'd rather not link a bank.