How We Measure What We Publish

Every number on this site falls into one of two buckets. Product capability facts are things you can reproduce on the free plan in minutes. Outcome numbers are ranges reported by early users, with the method and the limits stated next to each one.

We do not publish third-party benchmark studies we have not run, and we do not invent review counts or ratings. When a number changes, this page is updated and dated.

Product Capability Facts

200+

Bank and payment-platform formats

What it means
CSV and Excel exports from banks and payment platforms worldwide that our AI parser recognizes automatically, including Stripe, PayPal, Wise, Revolut, and major banks across the US, Europe, and Asia.
How we measure it
Format coverage grows continuously as users upload files. Each new layout that parses correctly is added to the supported set. If a file fails, the user can retry after a correction, and the pattern is folded back in.
How you can verify it
Export a statement from your own bank in whatever format it produces and upload it. If the columns are detected and transactions are categorized, that format is covered for you.
30+

Currencies

What it means
Currencies supported across transactions, invoices, time entries, and reports, with exchange rates refreshed daily.
How we measure it
A currency is supported when rates are available and math checks out in every view: transaction totals, invoice totals, and reports.
How you can verify it
Create an invoice in any supported currency and confirm totals match your own calculation.
50+

OCR languages

What it means
Languages our receipt scanner can read from payment screenshots and receipt images, covering Latin, Cyrillic, CJK, Arabic, and Indic scripts.
How we measure it
Language coverage is tested per release against a sample set of receipts and payment screenshots in each supported language.
How you can verify it
Upload a receipt in your language and check that the amount, date, and merchant are extracted.
<2 min

Setup time

What it means
Time from creating an account to importing your first transactions and seeing categorized results.
How we measure it
The happy-path walkthrough is timed with a stopwatch on a fresh account: sign-up, one CSV upload, first AI categorization.
How you can verify it
Create a free account and time yourself. If your first import takes longer than two minutes, tell us what got stuck.
6

Competitors honestly compared

What it means
QuickBooks, Xero, FreshBooks, Wave, Zoho Books, and Expensify, each with a dedicated page listing where we lose, not just where we win.
How we measure it
Each comparison lists both Tally Assistant strengths and competitor strengths. Pricing is checked against published plans at the time of writing and dated on the page.
How you can verify it
Open any comparison page. The competitor's strengths are listed in full before any recommendation is made.

Outcome Numbers

These numbers describe what early users report, with the method and the caveats printed next to each claim. They are ranges, not controlled studies.

91%

Fewer reconciliation errors

What it means
The typical reduction in transaction-level errors users report when switching from manual spreadsheet entry to AI-parsed imports. This is a user-reported range from early testing, not a third-party study.
How we measure it
Early users were asked to compare their spreadsheet reconciliation results against the same month processed through Tally Assistant, counting mismatched or mis-categorized transactions on each side.
How you can verify it
Run one month of transactions through the free plan, then spot-check the categories against your bank statement. Count the mistakes. Most users find the AI pass is close to clean, where manual entry of the same volume usually has several slips.
Limits
Individual results vary with transaction volume, file quality, and how carefully categories are reviewed. We do not claim this number as a controlled study.
~10 min

Saved per invoice

What it means
The typical time difference between building an invoice line by line in a template and generating one from a one-sentence description.
How we measure it
Based on timed comparisons by early users: opening a template, typing line items, calculating totals, and attaching a PDF against typing one sentence and reviewing the AI result.
How you can verify it
Time your next invoice both ways. If your invoices are more complex than a few line items, the gap can be larger.
Limits
Complex invoices with many line items take longer to review than simple ones. The number is a midpoint, not a guarantee.
3-4 hrs

Saved monthly per user

What it means
The monthly time range early users report saving on CSV processing, categorization, and invoice follow-up combined.
How we measure it
Self-reported by early users comparing their prior monthly bookkeeping time against their time in Tally Assistant, including payment reminders running automatically.
How you can verify it
Track your bookkeeping time this month, then compare with the same volume on a manual spreadsheet. Most users land in this range.
Limits
The biggest driver is monthly transaction volume. Fewer than 30 transactions a month will land at the low end or below.

What We Don't Publish

Last updated: August 11, 2026. Questions about a specific number? Contact [email protected] and we will tell you exactly where it came from.

See what we test, publish, and deliberately do not publish on the benchmark page →