Auto-Categorization

Auto-Categorization fills in missing categories on your transactions by looking at how you’ve categorized similar transactions in the past. No AI, no external services, no rules to configure — just your own history, working for you.

How it works

When you click “Auto-Categorize” in the add-on sidebar, here’s what happens:

1. The add-on scans all transactions in your Transactions sheet that already have a category assigned.
2. It builds a lookup of merchant names to categories, weighted by frequency. If you’ve categorized “COSTCO” as “Groceries” 15 times and “Home” twice, it picks “Groceries.”
3. It finds any uncategorized transactions and checks if their merchant/memo matches something in the lookup.
4. Matching transactions get their category filled in automatically.

You’ll see a toast notification telling you how many transactions were categorized — or if no matches were found.

No AI, no external services

Auto-Categorization runs entirely inside your Google Sheet. Your transaction data is never sent to any external server, API, or machine learning service. The matching is simple frequency-based lookup: same merchant name → same category you’ve used most often for that merchant.

This means:

• Your data stays private — nothing leaves your spreadsheet
• It gets better over time — the more transactions you categorize manually, the more the add-on can match automatically
• It’s predictable — no black-box decisions, just “you categorized this merchant as X before, so it does it again”

Merchant normalization

Bank memos are messy. “COSTCO WHSE #1234 DURHAM NC” and “COSTCO WHOLESALE RALEIGH” are the same merchant, but the raw text is different.

Auto-Categorization normalizes merchant names before matching by stripping:

• Store numbers and location codes
• City names and state abbreviations
• Phone numbers and URLs
• Transaction type prefixes (e.g., “DEBIT CARD PURCHASE”)
• POS system prefixes (SQ*, TST*, SP)
• Common noise words (Inc, LLC, Online, etc.)

This means a transaction from “SQ *BLUE BOTTLE COFFEE SF CA” will match against your previous “SQ *BLUE BOTTLE COFFEE OAKLAND CA” — because after normalization, they’re both just “blue bottle coffee.”

When it doesn’t match

Auto-Categorization won’t assign a category if:

• The transaction memo is completely new (never seen before)
• The memo is too short or generic after normalization
• You haven’t categorized any similar transactions yet

In these cases, just categorize the transaction manually. The next time a similar one appears, Auto-Categorize will handle it.

Best used with CSV Import

The typical workflow is:

1. Import a batch of transactions via CSV Import
2. Click “Auto-Categorize” — most transactions get categorized instantly
3. Manually categorize the handful of new merchants
4. Next month, even fewer will need manual attention

The more months you use Aspire Budgeting, the more effective Auto-Categorization becomes.