How much does Rakuten pay?

Rakuten is the seventh answer this format has found, and on paper it should have been the shortest page here. The other six all hid something: a currency with no published value, a ceiling stated to one decimal place with no dollar figure behind it, a rate set per user by an algorithm. Rakuten hides none of that. It pays dollars, and it prints a Cash Back rate on the tile of every one of the brands it works with, before you sign in. The question looks answered before it is asked.

Then you read two Rakuten pages at the same time and get two different rates for the same store.

The rate is published for every store, and that is genuinely rare

Start with the part Rakuten does better than anything else in this format. Its home page says it works with “3,500+ brands with new Cash Back rates every day”, and every store page carries a headline rate, a Terms & Exclusions block listing what the rate does not apply to, and a Posting Time line saying when the Cash Back will appear. None of that requires an account. Nothing else on this site's competitor list publishes the equivalent.

The exclusions are not boilerplate either. NordVPN's block runs to six sentences and states that Cash Back is limited to new NordVPN customers, to the initial sign-up, to one per Rakuten member, and “only on the NordVPN Basic plan portion of a bundled subscription” where you buy Plus, Complete or Prime. It also says that orders placed through any country's site other than the US one void the Cash Back, and that using a coupon code Rakuten does not list will void it too. Everything a reader needs in order to be disappointed is printed in advance, on the same page as the number.

The front page and the store directory disagree

Rakuten's home page, signed out, scrolls a wall of brand tiles. Reading the rates off it on 15 September 2026, the common value is 12%, with a long run of 11%, 12% and 13% either side of it and a headline tile reading 108% Cash Back.

Rakuten's own store directory, at the All Stores link in its navigation, has a section headed “Featured stores with Cash Back”. Twenty-four brands are named there. These are the rates it gives them:

Nineteen of the twenty-four are at 3% or below. That is our arithmetic on Rakuten's list, not a Rakuten figure: count the rows, 19 ÷ 24 is 79%. The median is 3%.

The second section on that page, “More stores to love”, names six more and does not soften it: Target 1%, Lowe's 1%, Macy's 2%, Bloomingdale's 2%, Expedia up to 3%, Best Buy up to 10%. The household names are the low ones.

Both pages are Rakuten's, both were open in the same browser within a few minutes of each other, and neither is lying. The front page is a promotional carousel and the directory is the catalog. But a reader who wants to know what Rakuten pays and looks at the page Rakuten puts in front of them gets a number four times the median of the page one click further in.

The most useful word on a Rakuten store tile is “was”

Six of those twenty-four featured tiles carry a second, smaller figure. Rakuten prints the previous rate beside the current one. Here is every instance on that page:

• Hotels.com — up to 3%, was 1%
• Vrbo — 3%, was 1%
• VistaPrint — 6%, was 2%
• Carter's — 10%, was 2%
• Bergdorf Goodman — up to 15%, was 2%
• AliExpress — 16%, was 2%

Every raised rate on that page reverts to 1% or 2%. Not one of the six has a baseline above 2%. The raises run from three times the old rate to eight times it, and Rakuten tells you the old rate itself, in smaller type, immediately to the right.

This is worth being precise about, because it is the finding and it cuts both ways. It is a disclosure nobody else in this format makes, and it is the single most useful thing on the site for answering the question in this page's title: the elevated rate is the exception and Rakuten says what the rule is. A shopper who reads the small figure rather than the big one knows what the store will pay in a month. A shopper who does not, and who builds a habit around 16% at AliExpress, is building it on 2%.

The home page states the same thing about itself in a different register: “3,500+ brands with new Cash Back rates every day.” The rate is not a property of the store. It is a price that changes daily, and both of those sentences are Rakuten's.

108% is a real number, and the store page says 98%

The headline tile on the home page belongs to NordVPN and reads 108% Cash Back. It is not a typographical error and it is not impossible: affiliate commission on a first-year VPN subscription can exceed the discounted price of the subscription, so a service that shares its commission can pay out more than the sale was worth.

NordVPN's own Rakuten store page, opened straight afterward, says 98% Cash Back, was 20%. The browser tab agrees with the page rather than the home page: “NordVPN Coupons, Promo Codes & 98% Cash Back - September 2026”.

The ten-point gap is explained at the foot of the home page, in gray type under the last row of tiles: “Eligible rates include an extra 10% Cash Back as your welcome reward. Earn up to an extra $50!” So the arithmetic is ours and trivial: 108 − 10 = 98. Every rate on that carousel is the store's rate plus a one-time signing bonus capped across the account at $50, and the tiles do not say which part is which.

Applied to the 12% that dominates the carousel, the same subtraction gives 2% — which is exactly where the store directory says the catalog sits. The two pages stop disagreeing the moment you subtract the welcome reward. They only disagree because one of them adds it silently.

What the biggest name on the site pays

Amazon's tile says $5 Cash Back, in dollars rather than a percentage. Its Terms & Exclusions explain why. The $5 is not a rate at all:

• It is limited to members who have not yet made a first purchase through Rakuten.
• It pays $5 on each of the first three eligible Amazon orders of $25 or more, and only through the Rakuten browser extension or app.
• Purchases must be made within 60 days of signing up.
• It excludes Amazon Pharmacy, Amazon Health and the Amazon FSA/HSA store.
• Amazon orders do not count toward any Refer-A-Friend or Welcome bonus.

So the ceiling is $15, which is ours: 3 × $5, and it requires at least $75 of spending, which is 3 × $25. At the ceiling that is an effective 20% and it is the best rate on the site by a distance. It is also available once, to a new account, for sixty days.

The number that is not published anywhere on that page is the one a returning member needs: what Amazon pays after the third order. No ongoing rate is stated. This is the missing denominator, and on this page it belongs to the single brand most people would check first.

What it takes to reach the minimum

Rakuten's help center sets the floor at $5.01 in confirmed Cash Back before the next payment date. A balance of $5.00 or less carries over. Put the catalog rate against that floor and the spending requirement falls out. All three of these are ours, dividing Rakuten's threshold by Rakuten's published rates:

• At 1% — Target, Lowe's, Etsy, Instacart — $5.01 ÷ 0.01 = $501 of spending.
• At 2% — Macy's, Walgreens, Staples, TEMU — $5.01 ÷ 0.02 = $250.50.
• At 3%, the median of the featured list — $5.01 ÷ 0.03 = $167.

None of those is a large amount of shopping over a year. All of them are several hundred dollars of shopping before the first payment, which is the thing the 12% carousel does not prepare anyone for.

Four dates a year, and the wait is not the headline

Rakuten pays on February 15, May 15, August 15 and November 15, and only Cash Back confirmed by the payment date is included. Confirmation has its own clock, and the help center publishes three different ones:

• Online and in-store: 3 to 14 weeks.
• Dining: 3 to 5 days after you dine.
• Travel: 45 to 120 days after completing the booking.

Stack the slow end of the travel window against the schedule and the worst case is the number worth knowing. Our arithmetic on Rakuten's two published figures: a booking made on 16 May, the day after a payment date, can take 120 days to confirm, which is 13 September. That misses the 15 August date, so it pays on 15 November — 183 days, or about six months, between the booking and the money.

There is a faster route and it is easy to miss. The same help center page says a confirmed balance can be redeemed for a gift card “within minutes”, and that this route needs $5.00 rather than $5.01. The quick option has the lower threshold. That is unusual enough to be worth saying plainly, because in every other app on this list the fast exit costs more, not less.

The largest published payment is not for shopping

Rakuten's Refer-A-Friend page offers $50 when someone joins through your link and makes qualifying purchases totaling $50 within 90 days of joining. The bonuses are stated to be unlimited.

Set that beside the rate. This division is ours: that same $50 of qualifying spending, at the 1% to 3% the featured directory actually publishes, earns the friend $0.50 to $1.50 in Cash Back. The referral bonus is therefore between 33 and 100 times what the purchase itself pays.

That is not evidence of anything improper, and a referral bonus funded out of marketing budget is an ordinary thing for a company to run. It is a statement about where the money is. The largest single number Rakuten publishes for an individual member is paid for bringing in another member, and it is a hundred times the largest number it publishes for the shopping itself. Anyone deciding how to spend an hour on this app should know which of the two the company has priced highest.

It also explains the shape of almost every other page answering this question. A post that persuades you to click a Rakuten link is worth $50 to whoever wrote it, once you spend $50. Ours is worth nothing, which is the only reason to believe the paragraph above it.

The whole sum, side by side

Same rows as the other pages in this format, so all seven compare. Every Rakuten figure is published or derived from a published one, read on 15 September 2026.

The first row is the one no other page in this format can fill, and the third row is the reason the first one still does not settle the question.

What Rakuten does better than we do

This section is on every page we write about a competitor. On this one it is the easiest it has been.

• It publishes the rate, for every store, to anyone. No account, no install, no support ticket. Six of the seven apps in this format require you to be inside the app before you can see what an action pays, and two of them will not tell you then.
• It publishes what the rate used to be. Printing “was 2%” next to “16%” is a disclosure that costs the company conversions and helps the reader, and nothing else on this site's competitor list does anything comparable. It is the best single practice this format has found anywhere, and it deserves to be copied rather than merely noted.
• The exclusions are on the same page as the number. In full, in plain sentences, before you buy. The disappointing details about bundled plans and country-specific sites are printed in advance rather than discovered afterward.
• It pays money. Dollars, by check or PayPal, with no intermediate currency whose value Rakuten sets. Four of the six other apps here put one in the way.
• It states how it makes money, on the front page. “Brands partner with us to send them shoppers. They pay us and we share that money with you. That's how we're free.” That is the whole business model in three sentences, in large type, above the fold.
• It is on Android, and it has a browser extension, an app and an in-store program. Three of our four apps are iPhone and iPad only.

The fair reading of the two-page discrepancy, too: a promotional carousel is what a home page is for, the welcome reward it adds is real money, and the footnote explaining it is present rather than absent. The criticism is narrower than it looks. It is that the footnote is in gray type at the foot of the wall, the tiles carry no marker of their own, and the arithmetic needed to recover the actual rate is a subtraction the reader has to know to perform.

Seven ways this question gets answered

Seven pages in, the pattern is still about the shape of the disclosure rather than the app, and Rakuten adds the case this format had not met: the company that publishes everything and still leaves the question open.

• Both halves published. Swagbucks. Read the same reward at every denomination and the answer is a number you can act on.
• The rate published, the price hidden. Sweatcoin. A ceiling to one decimal place and still no dollar figure.
• Neither published, and the rate is personal. Mistplay, whose terms say different users may be paid differently for identical actions.
• Both published for the capped part, neither for the rest. Receipt Hog, where the guaranteed half finishes at fifty cents a week.
• The price denied in the terms and stated in the prize rules. Fetch, priced exactly in the one document type where somebody outside the company decides whether a value has to be printed.
• No conversion to do, so the arithmetic moved to the exit. Ibotta, where the numbers that decide what you receive are the $20 floor and the $3.99 monthly fee below it.
• Everything published, twice, at two different values. Rakuten, and this page. The rate is on the tile, the previous rate is beside it and the exclusions are underneath — and the front page still shows 12% where the directory shows 2%, because one of them adds a one-time bonus without marking it.

The instruction to carry forward is the one this page had to learn late. When a company publishes a number in more than one place, read every place before quoting any of them. Rakuten's home page, its store page and its own browser tab title gave three figures for NordVPN within a minute, and the reconcilable difference between them was a footnote two screens down. The same habit is what the review of this app applies to the payment schedule, and it is the only reason the 12% on the carousel does not end up in an article as the rate Rakuten pays.