Fix These Tutor Payroll CSV Errors First
· 5 min read
Your payroll system rejects the file, or worse, accepts it and pays someone for 40 hours they did not work. Both come from the same place: a few predictable errors in the CSV you exported for your tutors. Here are the ones that actually cause problems at a tutoring center, ordered from the ones you will hit almost every pay period down to the rare ones, each with the fix.
Tutor names that don't match your payroll records
This is the error you will hit first and most often. Your export says "Sam R." or "Samantha Reyes-Lin," and your payroll system has "Samantha Reyes." The upload either fails on that row or creates a duplicate worker, and now one tutor's hours are split across two people who both look underpaid.
The fix is to make the name field in your time tracking match your payroll record exactly, before you export. Same spelling, same hyphen, same order. Decide on one canonical name per tutor (the one payroll uses) and use it everywhere. If two tutors share a first name, that is exactly when a nickname creates a silent mismatch, so spell both out. Fixing the name once at the source beats find-and-replace in a spreadsheet every two weeks.
Half-hour sessions rounded the wrong way
Tutoring runs in 30 and 45 minute blocks, so rounding errors hit you harder than a restaurant that works in full shifts. A tutor who worked 2:45 gets logged as 3:00, or 2:30, depending on who rounded and how. Across ten tutors and two weeks, that drift is real money, and it is money your grant report cannot explain.
Pick one rule and apply it to every row: either pay exact minutes, or round every session to the same increment in the same direction. Do not round some up and some down by feel. The other half of this is the format. Payroll usually wants decimal hours (2.75), not clock time (2:45), and mixing the two in one column is how a 45 minute session becomes 45 hundredths of an hour. A decimal and HH:MM hours converter settles that before the numbers go in the file.
Missing or renamed columns
Payroll uploads map columns by header. If your export calls it "Worker" and the importer wants "Employee," or a column you relied on last month is gone this month because a setting changed, the whole file bounces. Tutoring centers that also report to a grant feel this twice, because the grant wants a distinct-tutor count and a total-hours figure in a shape that does not change.
Use an export with a stable CSV column schema that keeps the same headers every single time. Kangaroo Clock holds its columns steady and includes a distinct-worker count, so a template you built in January still maps in December. If your current tool renames or drops columns between exports, that is the thing to fix, because no amount of careful cleanup survives a header that moves.
A forgotten clock-out that inflates the hours
A tutor clocks in for a 4pm block and forgets to clock out. If the row stays open until you run payroll Friday, a naive export can show a 90 hour session. Upload that and you have either paid someone a fortune or triggered a payroll error that takes an afternoon to trace.
Look for one open entry with an absurd duration before you export, every time. The real fix is how auto-close handles forgotten clock-outs: a stale open entry gets closed at its start time plus a cutoff, never at the current time, so a tutor who forgot once shows a reasonable session instead of three days. Check your cutoff matches your longest real block so a genuine long session is not cut short.
Two programs merged into one line
A tutor who worked both the after-school reading program and a Saturday SAT session should not land as one undifferentiated block of hours if those get paid or funded differently. When the programs are merged, you cannot split the report, and the grant that funds reading cannot see its own hours.
Tag entries by program and group tutors by role before you export, then filter. The point is that when the CSV lands in front of payroll or a funder, each line already belongs to the right bucket. If you are still deciding how to get tutor hours into payroll, the walkthrough on moving tutor hours into payroll by CSV covers the export-to-upload path end to end.
Timezone drift on sessions near midnight
This one is rare but nasty. A late session logged at 9:45pm shows up on the wrong day, or an hour off, because the file stored one timezone and your payroll read another. The result is a session that lands in the wrong pay period, which is the kind of error nobody catches until a tutor asks why a week is missing.
Store times in UTC and display them in the viewer's local time, which is what Kangaroo Clock does, so the stored value is unambiguous and what you see is your own clock. If you are stitching together exports from two different tools, confirm they agree on the zone before you trust the dates.
The quick pre-upload check
Before any file goes to payroll, scan for five things: every name matches a payroll record, hours are in one consistent format, the expected columns are all present, no single row has an impossible duration, and the dates fall inside the pay period you meant. Thirty seconds of that saves the afternoon you would otherwise spend untangling a bounced upload.
If you are setting up hour tracking for a tutoring center from scratch, you can start a free workspace and export a real CSV in a few minutes to see exactly what your payroll system will get.
Tags: tutoring, payroll, csv, time tracking
See it in your own setup
No signup needed. Add a few names, share a kiosk URL, watch hours land.