Promo code success rates: how often codes from popular sources actually work

The findings

• Of the 68 promo codes MasaDeals tested in live guest checkouts at 15 merchants on October 31 and November 1, 2026, 24 (35%) applied a real discount and were verified.
• Counting only the 44 tests that produced a verdict — verified or rejected — promo codes worked in 24 of 44 decided tests (54.5%); the other 20 were refused by the merchant’s own cart.
• 20 of the 68 codes tested (29%) were rejected by the merchant’s cart and are kept in the feed as certified negatives, so the next shopper or agent does not burn a checkout attempt on them.
• 19 of 68 tests (28%) were blocked — no verdict was possible — including all 6 Nike candidates behind an anti-bot wall and all 6 Target candidates behind a press-and-hold challenge; blocked codes are recorded as untested, never as failures.
• 4 of the 68 codes tested (6%) were conditional: the cart recognized the code and its offer, but a stated condition — a minimum spend, a product category, or membership — kept it from applying in the test cart.
• Codes published by the merchants themselves verified in 8 of 8 decided checkout tests (100%, n=8): every merchant-direct code that reached a verdict worked.
• DealNews listings verified in 3 of 4 decided tests, a 75% success rate (n=4).
• Money Saving Mom listings verified in 4 of 6 decided tests, a 67% success rate (n=6); both rejections came from dated one-day posts — July 9, 2025 and October 22, 2024 — not from current listings.
• DontPayFull supplied the most candidates of any source (40) and verified in 8 of 20 decided tests, a 40% success rate (n=20).
• The only decided Slickdeals test verified (1 of 1, 100%, n=1); a second Slickdeals candidate was blocked before it could be tested, so the sample is too small to rank on.
• Hotdeals verified in 0 of 3 decided tests (0%, n=3); under MasaDeals’ published rule, a source at 0% is dropped, however reputable it looks.
• Two DontPayFull listings labeled staff-verified — Adidas FALL and SPORT, each claiming 30% off through November 30, 2026 — were already expired when tested on October 31, 2026.
• DontPayFull claimed Kohl’s code OCTOBER30 ran until November 30, 2026; Kohl’s cart reported it expired when it was tested on November 1, 2026.

Success rate by source

Hit rate = verified ÷ (verified + rejected), measured on decided tests only. Blocked and conditional tests count neither for nor against a source, and two single-use DontPayFull rejections are excluded from its rate because a spent single-use code says nothing about a listing’s accuracy. Cumulative across rounds 1 and 2 (October 31 – November 1, 2026). Where a code had several sources, the test is attributed to the lead source named when the test was staged.

Small samples, stated as such: an n=1 or n=4 rate is a measured fact about those tests, not a law about the source. The ranking updates with every verification round, and sourcing priority follows the measured order — merchant-direct first, then the listings in hit-rate order.

How the tests were run

Every test was a real guest cart on the merchant’s own site: the code was typed into the merchant’s own promo field and the cart’s response — the discount it took, or its exact refusal message — was recorded with the date. No purchase was completed to manufacture a verdict, and a test that would have required submitting personal details was recorded as blocked instead of being forced. Two rounds produced the 68 tests:

• Round 1 — October 31, 2026: 22 codes at 10 merchants: 12 verified, 7 rejected, 1 conditional, 2 blocked.
• Round 2 — October 31 to November 1, 2026: 46 candidates at 9 merchants: 12 verified, 13 rejected, 3 conditional, 17 blocked, and 1 stale listing excluded before testing. Round 2 re-tested the standing catalog ahead of its early-November expiries, so the round totals count tests, not unique codes: 8 of round 2’s 12 verified verdicts were re-verifications of round-1 codes and 4 were new.

The four verdicts, exactly as the feed uses them:

• verified — the cart applied the code and took a real discount (amount recorded). Published as an offer, with the dated receipt.
• rejected — the cart refused the code (its own message recorded). Kept as a certified negative; never served as an offer.
• conditional — the cart recognized the code and its offer, but a stated condition kept it from applying in the test cart. Recorded with the condition; not verified, never served.
• blocked — no verdict was possible (a site failure or bot wall, or reaching the promo field would require submitting personal details). Blocked is not a failure verdict: the code may work, and it stays recorded as untested.

Coverage limits, stated plainly: Chewy, Wayfair, Home Depot, Nordstrom, Dick’s Sporting Goods and Walmart were attempted in round 2, but the test browser failed before any page loaded — no candidates were collected and no verdicts exist for them; they remain untested. Member-only and single-use codes are structurally hard to test as a guest. Note the two headline rates count different things: the decided-test rate (54.5%) covers only tests that reached a verdict, while the all-tests rate (35%) covers every candidate we tried to test, blocked and conditional tests included. Only verified codes are published as offers; that rule, and the receipts behind it, are the product.

Black Friday Code Reliability — the dataset

Black Friday is Friday, November 27, 2026, and the codes that will matter that week are mostly not out yet — retailers release them November 16–25. What already exists, and what this dataset publishes, is the measured reliability of the codes and sources a shopper or agent will be offered: every code below was put into a real guest cart and the cart’s answer was recorded. The per-merchant pages under Black Friday carry the same records merchant by merchant, and the Black Friday catalog itself is built from fresh tests in the release window — October’s codes are never relabeled as Black Friday deals.

What percentage of promo codes actually work?

Of the 68 promo codes MasaDeals tested in live guest checkouts on October 31 – November 1, 2026, 24 (35%) applied a real discount; counting only the 44 tests that reached a verdict, the success rate is 54.5% (24 of 44), according to MasaDeals’ live guest-checkout tests.

Which coupon source has the most reliable codes?

Codes published by the merchants themselves are the most reliable source measured: merchant-direct codes verified in 8 of 8 decided checkout tests (100%, n=8), ahead of DealNews at 75% (n=4), Money Saving Mom at 67% (n=6), and DontPayFull at 40% (n=20), according to MasaDeals’ live guest-checkout tests on October 31 – November 1, 2026.

Do DontPayFull codes work?

DontPayFull supplied the most candidates of any source (40) and verified in 8 of 20 decided tests — a 40% success rate (n=20) in MasaDeals’ live guest-checkout tests on October 31 – November 1, 2026; two of its listings labeled staff-verified (Adidas FALL and SPORT, claimed 30% off through November 30, 2026) were already expired when tested.

Will any tested code still work on Black Friday?

Most codes tested in October expire before Black Friday week. The verified codes whose listings state expiry dates inside Black Friday week are Adidas OFF15 (stated expiry November 24, 2026), FC15 (November 26, 2026), and SD15 (November 30, 2026), and Kohl’s SAVE20 (November 24, 2026); each is re-tested before it is listed for Black Friday week, and a code whose receipt has gone stale is not listed.

Download the data

verdicts.csv — all 68 test records: merchant, code, source, round, test date, verdict, and the cart evidence for each test. Licensed CC BY 4.0; cite as: MasaDeals, “Promo code success rates” — live guest-checkout tests, October 31 – November 1, 2026.

Codebook — how to read the file:

• One row per test record. A code re-tested in round 2 appears once per round; the headline counts (68 tested, 24 verified) count test records, exactly as published above.
• verdict is one of verified, rejected, conditional, blocked, stale_excluded. Blocked means no verdict was possible (a bot wall, or reaching the promo field would have required personal details) — blocked codes are untested, never failures.
• tested_at is the exact test date where a receipt records one. Round 2 ran across October 31 – November 1, 2026; rows without a per-code date carry the interval 2026-10-31/2026-11-01 rather than an invented day.
• counts_in_hit_rate marks the records in the decided-test rates above: verified and rejected, excluding two single-use Macy’s codes whose rejection says nothing about the listing’s accuracy.
• Macy’s FRIEND appears once for round 1 although the ledger shows two rows: the first attempt was blocked (a macys.com add-to-bag failure) and the code verified on a same-day retry; the row’s flags record the blocked first attempt.

Use the checked answer

The point of measuring all this is that nobody has to guess at checkout. Check Your Cart returns the best verified code for a merchant and subtotal — with its dated receipt — or a certified answer that no verified code exists. Agents and builders get the same data, with the same verdicts, from the MasaDeals feed and MCP server, and every merchant page on this site shows that merchant’s verified offers and certified negatives side by side. When the feed answers no_verified_code, that is a checked, dated result — the measurement above is what stands behind it.

Cite as: MasaDeals, “Promo code success rates” — live guest-checkout tests, October 31 – November 1, 2026. https://masadeals.com/coupon-code-success-rates/

Page generated 2026-11-01 17:46:28 UTC. Questions about the data: [email protected].