How do I calculate expected earnings if my self-employed income is irregular?

Build a pipeline model rather than averaging past income. List every active engagement with a contract value and close probability, then compute `=contract_value * close_probability` per row. Sum by quarter using `=SUMIFS(E:E, F:F, "Q3 2026")`. This gives you a probability-weighted expected value that reflects forward-looking information — pending contracts, likely renewals, active proposals — rather than what happened to land in the last 90 days.

What close probability should I assign to a verbal yes but unsigned contract?

Use 70–85%, not 95%. Contracts stuck in "about to sign" for more than two weeks frequently fall through due to budget freezes, procurement delays, or internal reorganizations. Over-assigning close probability here is one of the most common reasons self-employed filers underpay estimated taxes — they include income that never arrived. If the contract has been verbally confirmed but unsigned for over 30 days, drop it to 60%.

How does the 110% safe harbor rule affect my quarterly tax payment?

If your prior-year AGI exceeded $150,000, your safe harbor minimum is 110% of last year's tax liability, not 100%. For someone who owed $60,000 in taxes in 2025, the 2026 safe harbor floor is $66,000 divided across four quarters — $16,500 per quarter regardless of actual 2026 income. This is encoded in the `MAX` formula on the Tax tab as `prior_year_tax * 1.1 / 4`. Per IRS Publication 505 (2025), paying this amount eliminates the underpayment penalty even if your actual 2026 liability turns out higher.

What effective tax rate should I use as a self-employed filer?

There's no single answer, but a defensible working assumption for net self-employment income between $100K–$200K is 36–40% combined. This layers self-employment tax (15.3% on the first $168,600 of net SE income as of 2025, with the employer half deductible) on top of federal marginal rates (22–24% in this range) and excludes state tax. For income above $200K, add the 0.9% Additional Medicare Tax. If you have pass-through income potentially qualifying for the QBI deduction under Section 199A, your effective rate drops — but that requires a separate calculation and isn't something you should hard-code into a planning model without running the numbers for your specific situation.

Why does my expected earnings model diverge from actual cash received?

Timing. Expected earnings reflect probability-weighted pipeline; cash received reflects invoices paid. A $30K contract that closed in September may not clear until October, shifting actual cash into Q4 while your expected value model counted it in Q3. Keep the Pipeline and Cash tabs separate — Pipeline drives tax planning estimates, Cash drives safe harbor calculations based on confirmed income. Reconcile the two at quarter-end and note any contracts that closed at materially different values than modeled, then update your close rate defaults accordingly.

Expected Earnings Spreadsheet for Self-Employed

Why Trailing Averages Fail for a Self-Employed Earnings Spreadsheet

The intuition behind using trailing income is reasonable: you take Q1–Q3, divide by 3, project Q4. The problem is that self-employed income doesn't mean-revert on a quarterly cadence — it lands in lumps when contracts close. A $45K SOW at 60% probability of closing in Q4 changes your expected income by $27K. A trailing average will miss that entirely, or worse, it'll include a one-time project from Q2 that has zero chance of recurring.

The second failure mode is symmetric. If you had a strong Q2 ($52K) and a weak Q3 ($28K), the average is $40K. But if you have two contracts actively in negotiation totaling $60K at 70% close probability, your pipeline-based expected value is $42K — and the $40K trailing figure will slightly undershoot. Neither number is catastrophically wrong, but the pipeline model is built from forward-looking information you already have.

A trailing average treats your income as if you have no information about the future. You almost always do.

Building the Expected Earnings Spreadsheet Structure

Three tabs, clearly separated:

Pipeline — one row per potential engagement. Columns: client name, contract value, close probability (%), expected quarter of payment, and computed expected value (=C2*D2). Total expected value per quarter via =SUMIFS(Pipeline!E:E, Pipeline!F:F, "Q3 2026").

Cash — confirmed income only. Invoices sent, payments received, date cleared. This feeds your safe harbor calculation. Total confirmed YTD: =SUMIFS('Cash'!C:C, 'Cash'!B:B, ">=" & Assumptions!$B$3, 'Cash'!B:B, "<=" & Assumptions!$C$3).

Tax — quarterly estimated payment logic. The formula that matters:

=MAX(
  (('Pipeline'!B42 + 'Cash'!B30) * Assumptions!$D$2) / 4,
  ('Cash'!B30 * 1.1) / 4
)

Where Assumptions!$D$2 is your blended effective rate (SE tax at 15.3% on the first $168,600 of net earnings as of 2025, plus your marginal income tax rate — combined typically 36–40% for income between $100K–$200K). The * 1.1 branch is the 110% safe harbor threshold that applies once prior-year AGI crosses $150,000, per IRS Schedule SE and Form 1040-ES instructions.

Scenario Table on the Pipeline Tab

A single expected value per quarter gives you a point estimate — the same conceptual mistake as the trailing average, just built from better inputs. The number that matters for tax planning isn't just the expected value; it's the range of plausible outcomes and where your tax payment falls within that range.

Add a scenario block directly below your pipeline summary:

Scenario	Pipeline Adjustment	Expected Q Income	Est. Quarterly Tax
Bear (close rate −20%)	$46,336	$46,336	$17,408
Base (modeled close rates)	$57,920	$57,920	$21,751
Bull (close rate +15%)	$66,608	$66,608	$25,003

The bear case applies a uniform 20% haircut to every close probability — useful for quarters where your sales cycle is longer than usual or you have a major anchor client with an uncertain renewal. The bull case applies a 15% boost, appropriate when you have warm referrals or late-stage contracts that have been verbally confirmed.

Build it with named ranges so it recalculates automatically:

Bear expected value:
=SUMPRODUCT(Pipeline!C2:C20, MIN(Pipeline!D2:D20 * 0.8, 1), (Pipeline!F2:F20 = "Q3 2026") * 1)

Bull expected value:
=SUMPRODUCT(Pipeline!C2:C20, MIN(Pipeline!D2:D20 * 1.15, 1), (Pipeline!F2:F20 = "Q3 2026") * 1)

The MIN(..., 1) wrapper prevents any adjusted probability from exceeding 100%.

Now your tax tab has three payment options instead of one. The practical move: pay the base-case tax, hold the delta between base and bull in a separate savings line, and release it at year-end if the bull scenario materialized. The bear case defines your minimum safe harbor commitment.

Close Rate Benchmarks by Stage

The model is only as good as your probability inputs. From what I've seen across agency and consulting work, these are reasonable defaults:

Stage	Description	Close Rate
Discovery call scheduled	No requirements shared	20–30%
Proposal requested	Client asked for pricing	45–55%
Proposal sent, no response	Under 2 weeks	50–65%
Proposal sent, active discussion	Back-and-forth on scope	70–85%
Verbal yes, contract not signed	They said yes, you're waiting	70–85%
Contract out for signature	DocuSign sent	90–95%

That 70–85% range for verbal commitments will feel low the first time you look at it. It shouldn't. Contracts that have been "about to be signed" for three weeks are not 95% close probability — clients get budget freezes, procurement delays, internal reorgs. Discount them. Your tax payment doesn't care about their intentions.

Where This Gets Tedious

Updating close probabilities every week across 15 active prospects is where most people abandon the model and go back to guessing. The three-tab SUMIFS logic means any change to the Pipeline tab cascades correctly to Cash and Tax, but you still have to open the sheet and make the updates. ModelMonkey can recalculate the entire pipeline and regenerate the scenario table from a chat message — you describe what changed ("moved the Acme contract to 85%, pushed the Novo project to Q4"), and it rewrites the cells. Try ModelMonkey free for 14 days — it works in both Google Sheets and Excel.