NetSuite top customers by revenue: building a report that holds up

"Who are our top ten customers?" sounds like a two-minute request. It becomes a forty-minute request the moment two people produce two different lists, which is the normal outcome, because NetSuite can answer the question in at least three defensible ways. Here is the same customer for the first half of 2026, measured three ways:

None of those is wrong. The $164,580 gap between invoiced and revenue is sales tax, freight recharges and a credit memo; the gap between revenue and collected is just terms. But if the sales director's deck uses invoiced and the CFO's uses posted revenue, the ranking will differ and you will spend the meeting on arithmetic instead of customers.

Revenue, invoiced or collected: pick one and label it

The native report, and what it assumes

NetSuite's Sales report category includes a Sales by Customer pair — summary and detail — which is the fastest route to a ranked list. Set the date range, sort descending on amount, and you have something usable in under a minute. Before you send it, check three things.

• What the amount column contains. Sales reports are built around transaction amounts, which can include shipping and tax depending on how your items and reports are configured. Your income statement contains only what posted to income accounts. If the two totals differ by a suspiciously round percentage, that's usually tax.
• Whether the date range means transaction dates or posting periods. There is an accounting preference that controls whether reports are driven by posting period or by transaction date, and it can apply to financial reports only. When it does, your sales report uses transaction dates while your income statement uses periods — which is precisely how one invoice dated 31 March and posted in April makes two reports disagree.
• Which subsidiaries are in scope. In OneWorld, the subsidiary context you're in and the consolidation you display both change the answer. A consolidated ranking translates child-subsidiary amounts at a rate, so a customer trading in two currencies can move several places purely on FX.

The saved search version

When you need customer fields the report doesn't carry — segment, region, account manager, credit terms — build it as a transaction saved search. The recipe is short:

1. 1.Criteria: Posting is true. Type is any of Invoice, Cash Sale, Credit Memo, Cash Refund. Date is within your range. Main Line is true if you are summing the transaction total, and false if you are summing line amounts — mixing the two double-counts everything.
2. 2.Results: Customer (group), Amount (sum), and Amount grouped by period if you want the trend beside the total.
3. 3.Sort: descending on the summed amount. Saved searches don't take a TOP 10 — you sort and read the first ten rows, or add a criterion on summed amount if you genuinely need to cut the list short.
4. 4.Then add the fields the report couldn't give you: customer category, sales rep, terms, or a custom field on the customer record. This is the entire reason to be in a saved search rather than a report.

The SuiteQL version

For the definition that ties to the income statement, go through the accounting lines rather than the transaction totals. This is the query worth keeping, because it can't accidentally include tax, freight or a non-posting order:

SELECT * FROM (
  SELECT
    c.entityid                                   AS customer,
    ROUND(SUM(tal.credit) - SUM(tal.debit), 2)   AS net_revenue
  FROM transactionaccountingline tal
  JOIN transaction t ON t.id = tal.transaction
  JOIN account a     ON a.id = tal.account
  JOIN customer c    ON c.id = t.entity
  WHERE a.accttype  = 'Income'
    AND tal.posting = 'T'
    AND t.trandate BETWEEN TO_DATE('2026-01-01', 'YYYY-MM-DD')
                       AND TO_DATE('2026-06-30', 'YYYY-MM-DD')
  GROUP BY c.entityid
  ORDER BY net_revenue DESC
) WHERE ROWNUM <= 10

Three notes on that query. Income is a credit balance in the ledger, so revenue is credits minus debits — get the sign backwards and your worst-returning customer tops the list. If you run multiple accounting books, add a filter on the accounting book column or you will sum every book together. And swapping the trandate range for a posting-period filter is what makes this agree with the balance-sheet-driven reports; the accountingperiod table holds quarter and year rollup rows as well as months, so filter to base periods when you build that list. If your saved queries have degenerated into single-line strings, the SuiteQL formatter will make them readable again.

Parent and child customers

This is the trap that produces a genuinely wrong answer rather than a differently-defined one. If your account uses customer hierarchies — one parent with a subsidiary or store per child record — transactions post against the child. Group by the transaction's entity and a $4.2M group appears as eleven separate $380,000 customers, none of which makes the top ten. Roll up through the customer record's parent field:

SELECT
  COALESCE(BUILTIN.DF(c.parent), c.entityid)    AS customer_group,
  ROUND(SUM(tal.credit) - SUM(tal.debit), 2)    AS net_revenue
FROM transactionaccountingline tal
JOIN transaction t ON t.id = tal.transaction
JOIN account a     ON a.id = tal.account
JOIN customer c    ON c.id = t.entity
WHERE a.accttype = 'Income'
  AND tal.posting = 'T'
GROUP BY COALESCE(BUILTIN.DF(c.parent), c.entityid)
ORDER BY net_revenue DESC

Reading the ranking

A ranked list on its own is trivia. Three things make it a decision.

• Concentration. Add a cumulative percentage column. If your top 5 customers are more than half of revenue, that number belongs in the board pack and in your bank covenant conversations. The same Pareto arithmetic applied to items instead of customers is what the ABC analysis calculator does.
• Movement, not level. Rank this period against the same period last year, in the same query. The interesting rows are the ones that moved four places, in either direction. A customer sliding from 3rd to 9th is a churn conversation you would otherwise have next quarter.
• Why revenue moved. "Revenue up 6%, margin down 2 points" is almost always mix. Decomposing it into price, volume and mix effects takes ten minutes with the price volume mix calculator, and price volume mix analysis walks through a worked example.
• Margin per customer, not revenue per customer. Once you have income and COGS lines from the same transactions, the ranking changes shape. The mechanics of getting cost onto those lines properly are in the NetSuite item profitability report.

Or skip the build

The report is not hard. Agreeing on the definition, remembering the credit memos, rolling up the hierarchy and re-running it for a different period next week is what actually costs time. Ask ERPray "top 10 customers by posted revenue for H1, rolled up to parent, with margin" and you get the number computed live from your own NetSuite account with the SuiteQL shown underneath — so the definition is on screen instead of in someone's head.