Revenue Recognition

How companies decide when a sale counts as revenue, the five-step rule behind it, and where the judgment in that rule can be stretched.

Lesson 1 of 8 · 8 min read · Advanced · Free · Last reviewed October 2026

The short version

Revenue is recorded when a company hands over what it promised, in the amount it expects to be paid, not when the order is signed or the cash arrives. US rules set this out in five steps: find the contract, list the separate promises in it, set the price, split the price between the promises, and record revenue as each promise is kept. The steps need estimates, such as how many products will be returned, so two honest companies can report the same sale differently. Most revenue problems come from recording sales too early. Receivables growing faster than sales, and deferred revenue shrinking while sales grow, are the usual places it shows.

By the end of this lesson, you'll be able to

  • Apply the five steps of revenue recognition to a sale with more than one promise
  • Explain how estimates such as returns and discounts change the revenue a company reports
  • Recognize the signs that revenue may have been recorded too early

Why this matters

Every profit figure starts with . Margins, earnings per share, and most valuation multiples are built on top of it, so a mistake at the top line runs through everything below. The Income Statement treated revenue as a given. This lesson looks at how a company decides the number, because that decision involves more judgment than most investors expect.

The rule in one sentence

Under US accounting rules, a company records revenue to show the transfer of promised goods or services to customers, in the amount it expects to be entitled to in exchange.1 Two things follow from that sentence. Revenue depends on handing something over, so signing a contract or sending an invoice doesn't create revenue by itself. And the amount is what the company expects to keep, which can be less than the price on the invoice.

The standard turns that principle into five steps:1

  1. Identify the contract with the customer.
  2. Identify the separate promises in the contract. The standard calls each one a .
  3. Determine the price the company expects to be entitled to.
  4. Allocate the price across the separate promises.
  5. Record revenue when, or as, each promise is kept.

Worked example: one sale, two promises

STU Corp, the fictional water filter company from Part 3, sells a filter system for $100 and a one-year cartridge plan for $160. The plan sends a replacement cartridge every three months. Suppose STU Corp offers both together for $234, paid on the day of the sale. All numbers here are hypothetical.

There are two promises: the system, and four cartridge deliveries over the year. The $234 has to be split between them in proportion to what each sells for separately.

Step 4: splitting the price
PromisePrice if sold separatelyShare of the totalShare of the $234
Filter system$10038.5%$90
One-year cartridge plan$16061.5%$144
Total$260100%$234

The $26 discount is spread across both promises. The plan's $144 covers four deliveries, so each delivery carries $36.

Step 5: recording revenue as each promise is kept
WhenWhat happensRevenue recordedDeferred revenue left
Day of sale$234 received, system delivered$90$144
Month 3First cartridge delivered$36$108
Month 6Second cartridge delivered$36$72
Month 9Third cartridge delivered$36$36
Month 12Fourth cartridge delivered$36$0

STU Corp collected all $234 on day one but recorded only $90 of revenue that day. The other $144 sat on the balance sheet as , a liability, because the company still owed the customer four cartridges. If the sale happened in the last month of the year, this year's income statement would show $90 from it and next year's would show $144.

At a point in time, or over time

A promise is kept when the customer gets control of the good or service. For some promises that happens at a single moment, such as the delivery of a filter. For others it happens gradually. The standard says revenue is recorded over time when, for example, the customer receives and uses up the benefit as the company performs.1 A year of software access or a cleaning contract works that way, so its revenue is spread across the period.

The distinction matters most for long projects. A builder paid $30 million for a three-year project may record revenue as the work progresses, which requires estimating how far along the work is. If that estimate is too optimistic, revenue arrives too early.

Where the estimates are

Step 3 asks for the amount the company expects to be entitled to. When the final amount can change because of returns, rebates, volume discounts, or bonuses, the company has to estimate it. It may include an uncertain amount in revenue only to the extent that a significant reversal later is unlikely.1

Worked example: a sale with a right of return
STU Corp ships to a retailerAmount
Systems shipped10,000
Wholesale price each$80
Amount invoiced$800,000
Returns expected, from past experience4%, or 400 systems
Revenue recorded$768,000
Refund liability$32,000

The invoice says $800,000, but STU Corp records $768,000, because it expects to refund the rest. The 4% is an estimate. If management assumed 2% instead, revenue would be $16,000 higher with no change in what was shipped. Across a year of sales, small changes in assumptions like this can move reported revenue by a noticeable amount.

Another judgment is whether the company is selling its own product or arranging a sale for someone else. A marketplace that passes $100 from a buyer to a seller and keeps a $15 fee reports $15 of revenue if it's acting as an agent, and $100 if it controls the product before the customer gets it.1 Profit is the same either way, but revenue, and every ratio built on it, is very different. When two competitors look far apart on sales, check whether they report the same way.

How revenue gets pulled forward

Most revenue problems involve timing: sales recorded in an earlier period than they belong to. Two classic methods:

  • . Persuading distributors or retailers to take more product than they need before the period ends, using discounts, longer payment terms, or generous return rights.
  • Bill-and-hold sales. Billing a customer for goods that stay in the seller's warehouse. These can be legitimate under strict conditions, but they've also been used to book orders that customers hadn't really asked for yet.

The SEC's case against Sunbeam, an appliance maker, involved both. The SEC found that at least $62 million of the $189 million of income Sunbeam reported for 1997 came from accounting fraud, including sales pulled forward from later periods with discounts and other incentives, and bill-and-hold sales it described as little more than projected orders disguised as sales.2 Pulling sales forward has a built-in problem: next period starts with a hole where those sales would have been, so the company has to do it again, and more of it, to keep growing.

What to check

Companies must explain their revenue policies and give detail about them, including revenue broken into meaningful categories, the balances of receivables and deferred revenue from contracts, and the amount of the price allocated to promises not yet kept.1 In a 10-K this is in the notes to the financial statements, usually in the first two or three notes. Things worth looking for:

  • Receivables growing faster than revenue. Customers are taking longer to pay, or sales were recorded before customers were ready to pay for them.
  • Deferred revenue falling while revenue rises. For a company paid in advance, that can mean it's using up its backlog faster than it's refilling it.
  • A large share of sales in the last days of each quarter. Companies sometimes disclose this, or analysts ask about it on earnings calls.
  • A change in revenue policy or in an estimate such as expected returns, especially one that lifts revenue in a weak year.
  • Revenue growing much faster than at similar companies without a clear reason in the business.

None of these proves anything by itself. Each is a reason to read further, which is the subject of Accounting Red Flags.

Key takeaways

  • Revenue is recorded when a company hands over what it promised, in the amount it expects to keep.
  • A sale with several promises is split between them, and each part becomes revenue when that promise is kept.
  • Estimates for returns, discounts, and progress on long projects all change the revenue a company reports.
  • Whether a company reports the full price or only its fee changes revenue without changing profit.
  • Receivables outrunning sales, and deferred revenue shrinking while sales grow, are common signs of revenue recorded early.

Check your understanding

In December, a company collects $1,200 for a year of software access that starts on December 1. Its financial year ends December 31. How much revenue does it record this year?

A company's revenue rose 12% this year, its receivables rose 45%, and it mentions new, longer payment terms for distributors. What's the most reasonable reading?

What's next

Revenue is one of several places where accounting records something before or after the cash moves. Accruals and Working Capital covers the rest, and shows how to measure how much cash a company's day-to-day operations tie up.

Go deeper

Learn more

Notes and sources

  1. Financial Accounting Standards Board. Accounting Standards Update No. 2014-09, Revenue from Contracts with Customers (Topic 606). Checked 2026-10-11. Back to text
  2. U.S. Securities and Exchange Commission. In the Matter of Sunbeam Corporation: Order Instituting Public Administrative Proceedings (Securities Act Release No. 7976, May 15, 2001). Checked 2026-10-11. Back to text

Saved to this browser in the prototype. On the live site, progress saves to your account.

This lesson is general education, not personalized investment, tax, or legal advice. Examples use fictional people and companies, and hypothetical numbers.