What's the right formula for weekly revenue trends in Google Sheets?

Don't use `WEEKNUM()` inside a QUERY formula - it's not supported by Google Visualization Query Language and throws a PARSE_ERROR. Instead, create a helper column with `=ARRAYFORMULA(TEXT(C2:C,"YYYY")&"-W"&TEXT(WEEKNUM(C2:C,2),"00"))` to generate a sortable week label like "2025-W03", then run your QUERY against that column using GROUP BY. This approach handles year boundaries correctly and scales to 50,000+ rows without issues.

How much should a freelancer set aside for taxes?

For most freelancers earning $50,000-$120,000 in net income (as of July 2026), a 28% blended reserve covers self-employment tax (15.3%) plus federal income tax. Set the reserve against collected revenue, not invoiced revenue - money you haven't received yet isn't taxable. If your expected annual tax liability exceeds $1,000, the IRS requires quarterly estimated payments due in April, June, September, and January. Under-payment penalties apply if you miss them.

Why does my receivables aging formula count blank due-date rows as overdue?

Google Sheets treats blank cells as zero in date math, and zero evaluates to December 30, 1899 - which is always in the past, so every row with a missing DueDate gets flagged as overdue. Fix it by adding `ISNUMBER(DueDate range)*` inside your SUMPRODUCT. That check skips blank cells, text strings, and any "PARSE ERROR" values left by your date normalization formula, so your AR total only counts rows with legitimate dates.

How do I stop my Status column from collecting freeform variations like "paid", "PAID", and "Paid - check"?

Apply a dropdown data validation rule (Data → Data validation) that restricts entries to exactly "Paid", "Unpaid", and "Disputed". For rows imported via CSV that bypass validation, add a helper column with `=IF(NOT(ISNUMBER(MATCH(G2,{"Paid","Unpaid","Disputed"},0))),"INVALID STATUS","")` and conditional-format it red. You'll catch bad values on import before they corrupt your pivot totals.

At what row count should I switch from ARRAYFORMULA to QUERY in my freelancer dashboard?

ARRAYFORMULA formulas run cleanly below roughly 5,000 rows and get noticeably slow approaching 50,000 rows. SUMPRODUCT handles up to about 15,000 rows without issues. For datasets above 15,000 rows, QUERY is the better choice for aggregations - it processes large ranges efficiently and keeps the sheet responsive. If your invoice log grows past 50,000 rows (unlikely for a solo freelancer, but possible over several years), a one-time Apps Script parse pass to store static calculated values is faster than any formula approach.

Freelancer Finance Dashboard in Google Sheets (2026)

Most tutorials show you a clean 10-row example. Your real data is 400-1,400 invoice rows per year from three different billing tools, each exporting dates in a different format. This article is built around that reality.

What Data a Freelancer Finance Dashboard Needs

Your raw invoice log is where everything breaks first. Before you build any metric, you need a stable schema that survives a new CSV export without collapsing every downstream formula.

Column	What Goes Here	Why It Breaks
`InvoiceID`	Unique ID from billing tool	Duplicates appear when you re-export a date range
`Client`	Client name, raw	Capitalization varies ("Acme Corp" vs "acme corp")
`IssueDate`	Invoice date	Three date formats from three tools
`DueDate`	Payment due date	Often blank on older exports
`Amount`	Invoice total	Sometimes includes currency symbols ("$1,200.00")
`PaidDate`	Date payment received	Blank until paid - nulls cause overdue miscounts
`Status`	Paid / Unpaid / Disputed	Freeform text unless you lock it down

That Status column will haunt you. Without a dropdown validation rule locking it to exactly three values, you'll get "paid", "PAID", "Paid - check", and "piad" all in the same column by quarter two. Add a helper column next to it:

=IF(ISBLANK(G2),"",IF(NOT(ISNUMBER(MATCH(G2,{"Paid","Unpaid","Disputed"},0))),"INVALID STATUS",""))

Conditional-format that helper column red on any non-blank value. You'll catch import drift on Monday morning before it corrupts your pivot.

Core Metrics for Your Freelancer Finance Dashboard

Four numbers deserve a permanent spot at the top of your dashboard. Every other view feeds from these.

Monthly collected revenue - what actually hit your account, not what you invoiced.

Outstanding AR - the sum of all unpaid invoices past their due date. This is different from total unpaid (which includes invoices not yet due).

Tax reserve - 25-30% of collected revenue, set aside automatically. As of July 2026, the self-employment tax rate sits at 15.3% on top of federal income tax. A 28% blended estimate covers most freelancers in the $50k-$120k net income range. If you're outside that band, adjust the percentage.

Client concentration - the share of revenue from your top 3 clients. If that number is above 60%, your income stability is fragile in a way a revenue total doesn't show.

These four fit in a single row of scorecards above your data. Directors reading over your shoulder at a standup will understand them in 5 seconds.

The Date Problem You'll Hit on Every Import

Mixed date formats in one column are genuinely the most common reason freelancer dashboards break. A year of invoices from FreshBooks, QuickBooks, and a client's own PO system will give you "2024-01-15", "1/15/24", and "15 Jan 2024" coexisting in the same column.

The parsing formula that handles all three without silently eating bad rows:

=IFERROR(
  DATEVALUE(B2),
  IFERROR(
    DATEVALUE(TEXT(B2,"MM/DD/YYYY")),
    "PARSE ERROR"
  )
)

The key is that last fallback: return the string "PARSE ERROR" rather than a blank or zero. A blank zero looks fine in a SUM. A red "PARSE ERROR" cell is impossible to miss. Add a conditional format rule: if cell contains "PARSE ERROR", fill red. You'll catch 15-40 bad rows per thousand records on a typical multi-system export, and you'll see them immediately instead of wondering why your Q3 total is $2,000 short.

Run this formula through an ARRAYFORMULA on the entire column so new rows process automatically. Below 5,000 rows this is instant. Above 50,000 rows, it starts to drag - at that scale, run a one-time Apps Script parse pass and store static values.

Receivables Aging

The overdue calculation has a trap that burns people: blank DueDate cells evaluate as zero (which Google Sheets reads as December 30, 1899), so every row with a missing due date gets counted as overdue. Guard against it:

=SUMPRODUCT(
  (G2:G="Unpaid")*
  ISNUMBER(D2:D)*
  (TODAY()-D2:D>0)*
  E2:E
)

The ISNUMBER(D2:D) check skips any row where DueDate is blank, a string, or a PARSE ERROR result. Without it, you're overstating AR by whatever the sum of undated invoices happens to be.

For an aging breakdown (current / 1-30 days / 31-60 days / 60+ days), use four SUMPRODUCT formulas with different day-range conditions. They're verbose but they're fast and transparent. SUMPRODUCT handles up to 15,000 rows without slowdown; above that, switch to QUERY.

Weekly Income Trend

This is where a common mistake wrecks otherwise solid dashboards. The Google Visualization Query Language - what powers =QUERY() in Sheets - does not support WEEKNUM() as a function inside the SELECT or GROUP BY clause. If you write SELECT WEEKNUM(C), SUM(E) GROUP BY WEEKNUM(C), you get a PARSE_ERROR. Not a subtly wrong result - an actual error that breaks your entire trend view.

The fix is a helper column. In column H, add a week-bucket label:

=ARRAYFORMULA(
  IF(C2:C="","",
    TEXT(C2:C,"YYYY")&"-W"&TEXT(WEEKNUM(C2:C,2),"00")
  )
)

This produces labels like "2025-W03" for every invoice row. Then QUERY against the helper column:

=QUERY(A:H,
  "SELECT H, SUM(E)
   WHERE A<>'' AND G='Paid'
   GROUP BY H
   ORDER BY H",
1)

The result is a clean weekly revenue table you can drop a line chart on. It handles year boundaries correctly (week 52 of 2024 sorts before week 01 of 2025) because you're sorting on a text string that leads with the year.

For row counts: QUERY handles 50,000+ rows cleanly. ARRAYFORMULA on the week-bucket helper column is fine below 50,000 rows - above that, calculate it once with Apps Script and store the values as static text.

Tax Reserve: Automating the Calculation

The reserve formula is simple - what makes it useful is tying it to collected revenue, not invoiced revenue.

=SUMPRODUCT(
  (G2:G="Paid")*
  (YEAR(F2:F)=YEAR(TODAY()))*
  (MONTH(F2:F)<=MONTH(TODAY()))*
  E2:E
) * 0.28

This sums only rows where Status is "Paid", PaidDate is in the current year, and the month is on or before the current month. Multiply by 0.28 (or your blended rate). The result is what you should have in a separate savings account right now.

The IRS requires quarterly estimated payments - due dates fall in April, June, September, and January. If your annual tax liability is expected to be $1,000 or more, you owe estimated payments or face underpayment penalties. A simple formula flag:

=IF(annual_tax_estimate>=1000,"⚠ Quarterly payments required","")

Tie annual_tax_estimate to your YTD collected revenue extrapolated to 12 months. It's a rough signal, not a tax opinion - but it's the difference between a surprise April bill and a planned one.

Automating the Data Pull

The manual part of this workflow - exporting CSVs from your billing tool, pasting them into your log sheet, fixing the date column - takes 20-30 minutes every week. That's fine at first. After six months, it's the thing you skip, and then your AR aging is stale and your tax reserve is under-counted.

ModelMonkey can replace that manual pull. Connect it to your billing source, describe what you want ("pull all paid invoices from the last 30 days and append to my Invoice Log sheet"), and it builds a refreshable connection that runs automatically. No Apps Script, no manual exports. The dashboard stays current without the weekly maintenance cost.

Try ModelMonkey free for 14 days - it works in both Google Sheets and Excel.