Why Shein's Price Changes at Checkout: 554 Reviews (2026)

Shein is one of the highest-rated shopping apps on Google Play. On 17 August 2026 it held 4.76 stars from 11,230,667 ratings with more than a billion installs, and 4.72 from 3,372,321 App Store ratings. Whatever the app does, most people who use it are not upset about it.

But the minority who leave 1-3 stars have a favourite subject, and it is not sizing, shipping or quality. It is arithmetic. We pulled 6,600 unique reviews across 18 Google Play storefront queries in 15 languages, kept the 554 that were 1-3 stars with at least 80 characters of text, and tagged every one. Price and discount mechanics are the single largest complaint family in the set at 17.5%, ahead of delivery at 16.8% and app behaviour at 15.9%.

Inside that family sits one very specific claim, repeated in six languages by people who have never met each other: the number on the product page is not the number you pay, and the change happens in the cart.

What We Read, and What It Cannot Show

Look at that last column before you look at any count in this article.

The windows are wildly different, so raw counts across languages mean nothing. A thousand English reviews cover five days. Four hundred Turkish reviews cover a year. Shein's review volume is enormous, and the store returns newest first with a hard cap, so a busy storefront gives you a narrow window and a quiet one gives you a wide window. Comparing "94 Turkish complaints against 96 English complaints" would be meaningless. Every comparison below is a share of that storefront's own negative reviews.

Apple's public review feed returned nothing for this app, in every storefront, ten pages deep, which is a known per-app behaviour rather than an outage. So the App Store rating above comes from Apple's metadata and every quoted review here is from Google Play.

We cannot verify a single price. We did not place orders, and we cannot see any reviewer's cart, currency setting, coupon state or shipping address. What follows is what shoppers report, plus what those reports do and do not add up to.

The Complaint Families

Reviews often belong to more than one family, so the shares do not sum to 100.

The price family is also the least evenly distributed. It accounts for 26.2% of Korean negative reviews and 24.5% of Turkish ones, against 5.3% of French and 4.5% of Polish. That spread is the first hint that this is not one bug.

Finding 1: The Listing Price and the Cart Price

Twenty-two reviews describe the same sequence: a price on the product page, a different and higher price once the item is in the cart or at payment. We hand-checked every match and removed two that turned out to be about stock and card errors. What is left spans six languages: nine Turkish, seven English, two Korean, two Mexican Spanish, one Brazilian Portuguese, one Arabic.

most of the time the price is advertised as one thing, but at checkout its higher.

the prices seem low and when you add them to your cart the price gets higher

they are showing price is low and when you will add in the cart it will be double

Spanish

en el total de compras, que aparece en carrito de compra es uno y al momento de pagar, aparece otro monto. Más elevado.

Portuguese

o valor muda quando vai pro carrinho. poluição visual.

Arabic

الاسعار صارت سيئه جدااا والخصومات سيئه بس اتمم الاغراض بالسلة برتفع سعر المنتجات

Korean

5천원짜리 옷. 그러나 선택하면 18,000원 결제가로 바뀌는 마법이 일어남. 사기도 아니고 뭐하는 것임.

The Turkish reviewers are the ones who write down both numbers, and they are the reason this cluster is measurable rather than anecdotal:

Turkish

ürüne bakarken 250 tl sepete gidince 500tl oluyor. hadi dedim memlekette de aynı fiyat alayım dedim. 2500 tl lik ürün attım. ödemeye geçtim 6000 tl.

Turkish

elbise fiyatı 230 tl sepete ekleyince 760 tl siz dalga mı geçiyorsunuz? uzak durun

Turkish

358 tl yazıor üründe tek renk seceneği var sepete atıyorum 2 katına cıkıyor uygulamayı sildim

Turkish

sepette bin tl gözüküyor ödemeye geçince bin tl ikibin civarı çıkıyor

Six reviews in this cluster name two figures. The reported jumps: 250 to 500 lira, 2,500 to 6,000 lira, 230 to 760 lira, 358 to double, 1,000 to about 2,000 lira, and in Korean 5,000 won to 18,000 won. One English reviewer describes a basket that moved while she was out of the app:

Yesterday I had 5 items in my basket for £15.28 and was about to pay. I left the app and came back shortly after, and the price had jumped to £52. I understand prices can change, but increasing by nearly £37 is ridiculous.

What Is Probably Happening, and What We Cannot Confirm

This is the part most articles skip. A price that rises between the listing and the cart has several boring explanations, and at least one of them is documented by the shoppers themselves.

The most detailed account in the entire set comes from a Korean reviewer of two years, writing a three-star review. She explains that coupons carry per-item and per-basket limits, that items stop being coupon-eligible after a period, and that the practical effect is a cart total that climbs as you add things. Then she gives a workaround:

Korean

쿠폰이 제대로 적용되는 상품들을 먼저 구매하고 되지 않는 상품을 따로 결제하는 게 낫습니다.

Buy the coupon-eligible items first, pay for the rest separately. That is a shopper reverse-engineering a discount engine, and it fits the rest of the cluster: a listed price that assumes a promotion, a promotion with conditions, and conditions that resolve only once the basket is assembled.

An Arabic reviewer describes the same logic from the opposite end. Completing the basket is what makes prices rise, not any single item.

Other explanations we cannot rule out from review text alone: import tax and shipping added at payment rather than on the listing, a currency or region setting that differs from the shopper's location, quantity or variant defaults, and time-limited flash prices that expire between the two screens. The £15.28 to £52 report above is consistent with an expiring promotion, and the reviewer says as much before objecting to the size of the jump.

What the review set does establish is narrower but still real: shoppers in six languages consistently experience the listing price as a promise and the cart price as a surprise. Whether the engine is behaving as designed is a separate question from whether the interface communicates it.

Finding 2: Coupons That Ask for More

Twenty-two reviews, 4.0% of the set, are about coupons and discount codes that did not apply. They come from ten different storefronts, which makes this the most geographically spread complaint in the price family.

I'm really tired of the fake coupons that say spend 10.00 and get 20% off. When you try to use the coupon, it tells you, you have to spend more. The free prizes are never awarded.

Polish

irytujące wyskakujące 50 okien z kuponami nagrodami, kołami fortuny. aplikacja dla 3 letnich dzieci a nie dla dorosłych co chcą jasno i przejrzyście cis kupić. robienie w bambuko, kup teraz, dostaniesz kupon na xxx zł który zq 2 dni albo 12h traci ważność.

Dutch

heel jammer als je een actie krijgt dat je dan je punten niet kan inleveren. je spaart ze niet voor niets.

Korean

쿠폰40%,50%,60%,30%,25%×2 적용하면취소됨 상담사한테 연락해도모름. 쿠폰이 만료됨.

That Korean report is worth reading twice. The order was cancelled after stacking discounts, support could not explain it, the coupons expired in the meantime, and the shopper was told she had been flagged as high risk. Fraud controls on discount stacking are entirely normal for a retailer at this scale. Being told nothing while your coupons expire is a product decision rather than a fraud decision.

A Russian reviewer reports something smaller and easier to check: the check-in rewards display one figure and credit another.

Russian

Тот же обман с поощрительными баллами за посещение в течение 10 дней без перерыва. Пишут: вы получаете 8 баллов, а когда жмёшь на кнопку, получаешь всего 1 балл.

Finding 3: The Prize Games

Seventeen reviews are about the free-gift and referral mechanics rather than about anything a shopper bought. Seven of those seventeen are Japanese and four are Korean, which is the sharpest national concentration in this study: 23.3% of Japanese negative reviews are about giveaways, against 4.2% of English ones.

Japanese

フリップという無料で4万円もらえるやつを15回くらいしたのに結局99%→99.1%とかってなって結局もらえらなくて15人招待するのに苦労したのに結局何もしてくれませんでした

Japanese

無料プレゼントやルーレット等、あげるあげる詐欺。商品を購入したら無料のプレゼントがもらえるような表示をしているが、これは全くの嘘で、カスタマーセンターいわく、無料のプレゼントのチャンスをGetしただけとの事。

Japanese

フラッシュグラブというiPhone17promaxとApple WatchとiPadAirプレゼントイベント条件あわせて今回渋々買いましたがそれだけでは貰えず後だしの条件追加で尚且つ抽選という形になり時間がきたら早い者勝ちという訳のわからない仕組みにかわり解放時間１分前位から押しまちして０になったと同時に押しましたが残念ということになり3点とも全て同じ結果になりました絶対受け取れない仕組み

The recurring structure in these reports is a progress bar that stops just short, or conditions that appear after the shopper has already met the ones that were shown. The reviewer who reached 99.1% after fifteen attempts and fifteen invitations is describing time spent, not money lost, which is exactly why it produces one-star reviews from people whose orders arrived fine.

A Korean reviewer describes the same shape in a referral event: a completion screen, then an invalidation with no explanation of which condition failed.

Finding 4: When the Compensation Is Not Money

Fifty reviews mention refunds and returns. A subset describes a pattern worth separating out, because it turns a service failure into a second purchase: the remedy offered is store credit, points or a coupon rather than the money back.

A Japanese reviewer who received the wrong item reports being offered a coupon worth up to 1,500 yen on a 1,500 yen minimum spend, or free shipping, instead of a refund, and being told to buy the item again. An American reviewer describes refunds paid in points, with a cap on how many points can be spent per order. And a Russian reviewer describes what happens to a discount when part of an order goes back:

Russian

вы же типа по скидкам брали, а при возврате скидка анулируется на все, даже на те товары которые у вас остаются.

If a basket discount is calculated across the basket, unwinding it on a partial return is defensible arithmetic. The complaint is that nobody sees that arithmetic until the refund arrives short. The same reviewer adds the part that stings more: wallet balance cannot be spent on discounted items, which is most of the catalogue.

Finding 5: The Tax Reviews Are Not About the App

Fourteen reviews are about customs duty and import tax, and eight of those fourteen are Turkish. They are among the angriest in the set, and they are not Shein app defects.

Turkish

65 TL lik ürün gümrükle 800 küsür TL ye denk geliyor. saçma sapan pahalı.

Turkish

Ürünlerden gümrük vergisi istiyor bu normal ürünler yurtdışından geldiği için ama birde gümrük vergisi kadarda nakliye ücreti istiyor. Bu yüzden ürünlerde normal fiyatının 2-3 katına çıkıyor

Turkey tightened its low-value import regime, and the resulting duty applies to any cross-border parcel regardless of which app ordered it. An Italian reviewer reports the same category of change with a different figure, around 20 euros more per order, after taxes that shoppers previously did not pay started being collected.

We are flagging this deliberately. A naive reading of the price family would fold these fourteen reviews into "Shein raised prices", and that would be wrong. It is also why the Turkish share of the price family, 24.5%, is the highest of any large storefront: part of it is a national customs rule wearing a one-star review.

The Number That Nearly Fooled Us

The Arabic storefront produced 65 negative reviews, the third largest bucket in the study. Read individually, 49 of those 65, or 75.4%, are the same request: Shein does not ship directly to Iraq, and Iraqi shoppers want it to.

Arabic

نعم هناك حلول لكن ليست رخيصه و هو الوسيط من بلاد اخرى ولكن على اي قطعه يتم زيادة سعرها بشكل مبالغ بسبب انتقالها من دولة الى دولة وذلك قد يستغرق وقت طويل جدا مثل شهر كامل و ايضا لا يمكننا استخدام اي كودات متوفره ولا يتم حسم الخصومات

That reviewer explains the cost of the workaround: a forwarding agent in another country, a per-item markup, delivery that can take a month, and discount codes that cannot be used at all.

This is not a defect report. It is a feature request filed as a one-star review, and it is three quarters of one storefront's negative volume. If we had counted "Arabic negative reviews about delivery" as evidence of a delivery problem in the app, we would have published a false finding. Star ratings are a blunt instrument, and this is what that looks like at scale.

What This Set Does Not Prove

It is not a sample of Shein shoppers. It is a sample of complaints, and at 4.76 stars the complaining minority is small. Nothing here says the average order goes wrong.

Fifteen languages is not fifteen countries. People write in the language they are comfortable with, storefront and residence are not the same thing, and the Iraqi cluster above is the proof.

No claim here is about a specific product or a specific price. We cannot see anyone's basket, and we did not test the checkout ourselves.

Older quotes are labelled by date for a reason. The Turkish and Korean windows reach back into 2025 because those storefronts return fewer reviews per day. A 2025 report describes 2025.

What to Do With It

If your Shein total keeps landing higher than you expected, the review set suggests three checks that cost nothing. Compare the product page price against the cart line item rather than against the order total, because tax and shipping legitimately enter later. Check whether the listed price depends on a coupon, a quantity threshold or a countdown, since that is what the Korean reviewer above worked out the hard way. And if you are ordering across a border, look up your own country's import threshold before blaming the app, because in this data that is a meaningful share of the anger.

Live complaint feeds for both platforms: Shein on Android and Shein on iPhone. We also ran the trust question separately in is Shein legit and safe.

For the same treatment of the nearest competitors, see is Temu legit and is AliExpress legit. The category-wide view is in cheap shopping apps ranked by 1-star reviews and what shoppers complain about in ecommerce apps.