A rental property spreadsheet costs nothing on the day you build it. Twelve columns, a tab per property, a SUM at the bottom of the rent column, and every dollar has a cell to live in. For the first year or two it genuinely works, and every month it works feels like evidence that landlord software is a solution looking for a problem. I ran my own books this way, and I will not pretend the spreadsheet was a mistake. It was the right tool, until the portfolio outgrew what a grid can be trusted to remember.
The catch is the billing schedule. Software bills you monthly; a spreadsheet bills you once a year, in February, when Schedule E has to be assembled from twelve months of categories you assigned in ten seconds each. The bill arrives as hours of reconstruction, a depreciation number nobody computed, a deposit that quietly became income, and deductions that fell through the grid. This page prices that bill on both sides, names the five specific ways spreadsheets fail at tax time, and ends with an honest list of reasons to keep yours.
The five failure modes
Spreadsheets do not fail loudly. A wrong total renders in the same confident font as a right one, so the breakdowns surface in February instead of the month they happened. After enough tax seasons they sort into five repeatable modes:
- Miscategorized Schedule E lines. In October, a furnace repair, a furnace replacement, and a mortgage payment all feel like “expenses.” Schedule E disagrees: the repair is a line 14 expense, the replacement is a capital improvement recovered through depreciation, and only the interest slice of the mortgage payment is deductible, with principal and escrow excluded entirely. A one-column expense list flattens those distinctions, and the line-by-line Schedule E guide shows how many there are to flatten.
- No depreciation schedule. The largest deduction on most small landlords' returns is the one no spreadsheet computes on its own. It gets its own section below, because the dollars deserve it.
- Deposit liability mixed into income. A security deposit is not income when it arrives; it is the tenant's money you are holding. A sheet with one income column books it as rent, which overstates this year's income and leaves nothing to itemize against at move-out. A kept deposit becomes taxable income only in the year you apply it to damage or unpaid rent, a timing rule a single column cannot express.
- Lost receipts. The grid stores numbers, never evidence. The water heater invoice lives in an inbox, the paint receipt in a photo roll, the contractor bill in a glove box, and an expense you cannot substantiate is an expense you may not get to keep when someone asks. Numbers without paper are opinions.
- Formula drift. Sort a column and the SUM range silently keeps its old boundaries. Insert row 19 beneath =SUM(B2:B18) and the new expense exists everywhere except the total. Drag a fill handle one cell too far and a formula becomes a constant. The sheet keeps rendering a number either way, which is exactly the problem.
Line 18 is the expensive one
Of the five, depreciation costs the most and gets skipped the most often. Say you buy a duplex for $340,000 and the assessment puts the land at $68,000. Your building basis is $272,000, and residential rental property depreciates straight-line over 27.5 years:
annual depreciation = (cost basis − land value) ÷ 27.5
That is $272,000 ÷ 27.5, or $9,891 a year against ordinary income, worth about $2,374 in federal tax in a 24% bracket, every year you own the building. The rule that should worry a spreadsheet keeper is “allowed or allowable”: when you sell, the IRS computes depreciation recapture on the depreciation you were entitled to take whether or not you claimed it. Skipping line 18 means paying tax at sale on a deduction you never used. IRS Publication 527 governs the rules, and the depreciation calculator will run your own building, including the mid-month first-year factor.
The worked figures on this page are simplified examples, meant to organize your year for your CPA, not tax advice. Basis, land value, and bracket all change the math; have a professional confirm yours.
Price the error, not the subscription
The honest comparison was never $0 against a monthly fee. It is the expected cost of one quiet error against the fee. Landlord software for a small portfolio generally runs $100 to $300 a year. On the other side of the scale:
- One year of unclaimed depreciation on the duplex above: about $2,374 in federal tax, with the recapture exposure at sale accruing either way.
- One deposit dispute lost for want of an itemized trail: a few hundred dollars up to the full deposit, and in many states a statutory multiple of it. Read your state's statute for the exact exposure.
- One February spent rebuilding categories: eight to twelve hours at whatever your hour is worth, repeated every year the underlying habit survives.
Run that math with your own numbers and the spreadsheet stops being the cheap option somewhere well before unit three. A free tool with a four-figure failure mode is not free. It is unpriced.
Keep the spreadsheet if
Software vendors will not write this list, so here it is. The spreadsheet remains the right call when:
- You hold one or two units with stable tenants. A dozen transactions a month sits well inside what a disciplined grid can carry, and at that volume discipline matters more than tooling.
- Your CPA owns the depreciation schedule. If a professional maintains line 18 outside your sheet, the most expensive failure mode is already covered.
- You reconcile monthly, on a date. My own books close on the 5th of each month, and the 10-minute monthly close works the same in a grid as anywhere else. A reconciled spreadsheet beats neglected software.
- Your categories match Schedule E from day one. Start from the free spreadsheet template with Schedule E categories rather than a blank workbook, and February becomes an export instead of an excavation.
If all four hold, keep the spreadsheet and put the subscription money toward the water heater fund.
The switch points
The portfolio usually announces the switch before the landlord admits it. The signals worth trusting:
- The third unit arrives. Transactions scale linearly; cross-checks scale worse. Three units means three deposit balances, three lease dates, and three depreciation schedules sharing one grid.
- The first move-out with deductions. The deposit math needs an itemized, dated, documented trail, and a single cell labeled “deposit” is none of those.
- You stop living near the books. I manage my own small portfolio from two time zones away, and distance is unforgiving of anything that depends on memory.
- February costs a weekend. When the cleanup takes longer than the year's bookkeeping should have, the free tool has set its price.
When you do shop, shop by job rather than by feature count; the guide to landlord software for small portfolios covers how the pricing models differ and who each one is actually built for.
Moving without re-keying
The real reason landlords stay in a failing spreadsheet a year too long is the imagined cost of moving the history. It is an export, not a project. I built rents.ai because my own spreadsheets dropped things, and the migration path is the CSV you already have: export the sheet, import it in Settings, and the transactions land in a ledger organized to Schedule E line order, with line 18 depreciation computed from your building basis and in-service date instead of living in a tab you remember to update. The limits are worth stating in the same breath: there are no bank feeds, so new transactions still arrive by hand or by CSV, and it does not collect rent, screen tenants, or post listings. It keeps the books. For the cell-by-cell version of this argument, the spreadsheet comparison goes feature by feature.
However you decide, the tool is the smallest part of running rentals well. The checklists, dates, and document habits around it carry more of the load, and the complete self-management system for 1-10 units lays out the whole machine. Routines run rentals. The tool only decides how long the routine takes.