How the eMAG Buy Button Works: Factors Beyond Price and the Margin Maths (2026)

On eMAG several sellers can offer the same product page, and the “Add to cart” button goes to exactly one of them. This article does not promise to win you that button — nobody can, because eMAG does not publish its ranking formula. What it does cover is more useful: what your data actually tells you, how to detect that a non-price factor is at work, and what cutting your price really does to your margin.

What does the eMAG API tell you about the Buy Button, and what does it never tell you?

It tells you three things — your rank, the winning price and the number of offers — and never the fourth, the second-place price. In the official eMAG Marketplace API documentation, the offer read response (product_offer/read) carries these fields:

• buy_button_rank — your position in the “Add to cart” race. 1 means the button is yours.
• best_offer_sale_price — the winning price on that product.
• number_of_offers — how many sellers currently have an active offer.
• part_number_key — eMAG's product identifier, and the real unit of competition. Your rivals are not other listings; they are the other offers under the same part_number_key.

A small but practical architectural difference: on Trendyol the buybox data comes from a separate endpoint, while on eMAG it arrives inside the offer read. Sweeping your catalogue once therefore also brings the competitive data — a real advantage under a tight request limit. We covered the Trendyol side in the Trendyol buybox guide.

What it never tells you is exactly the number you need most: the second-place price. Trendyol exposes an equivalent; eMAG does not. The consequence is concrete: once you have won the button at a low price, the question “how far can I raise it?” has no ready answer. You are left with two bad options — sit indefinitely at the low winning price (a silent margin leak), or raise blindly and lose the button.

What matters besides price? The honest answer

The honest answer: which factors decide the ranking is unknown; the only thing your own data can prove is that a non-price factor exists. Most answers you will find online are invented. eMAG does not publish its ranking criteria or their weights, and the API returns no reason for your rank. So any table claiming “delivery score counts this much, reviews that much” has no source. What we do have is two kinds of knowledge: hard mechanical limits, and what you can measure from your own data.

Two mechanical limits that are certain

1. A price outside the band is never saved. Every eMAG offer has a minimum and a maximum sale price, and a price outside that band is rejected. So a price you considered “competitive” on paper may never have entered the race at all. On a day your rank drops, the band is the first thing to check.

2. Prices are compared on the VAT-exclusive plane. On eMAG the price you type excludes VAT — the API field is defined literally as “sale price without VAT”. A seller who types a VAT-inclusive figure out of Trendyol habit is really listing that figure as their net price: in Romania, at the 21% standard VAT rate, typing 100 RON means the customer pays 121 RON, while typing 121 RON means the customer pays 146 RON. That seller does not lose the button for being uncompetitive; they lose it for pricing on the wrong plane. We explained that difference in detail in selling on eMAG: commission, VAT and the Buy Button.

How to detect a non-price factor

There is exactly one rigorous method: if your rank moved while your own price did not, the thing that moved it was not your price. That proves a non-price factor exists; it does not identify it. Still, the three fields you have are enough to make the distinction:

• Rank dropped and the winning price dropped too → this is price competition; decide against your floor price.
• Rank dropped, winning price unchanged, number of offers up → a new seller arrived; before cutting, see whether they stay.
• Rank dropped and no number moved → what changed is a non-price factor. The right move here is not a price cut but a check of stock and fulfilment, plus a dated record of the observation.

You can do this by hand for one product; not for five hundred. That is where measurement earns its keep: making the pricing call from that day's winning price and offer count rather than from a hunch that “someone probably undercut me”.

How much does a price cut erode your margin?

In the example below a 5% cut wipes out roughly 28% of the unit margin — five times the size of the discount. “Keep going down until you win” is not a strategy; it is the controlled destruction of your margin. To see why, take a single unit on the eMAG plane. The commission and cost figures below are examples: there is no single “eMAG commission” rate — it varies by category and can drop on campaign sales. Read your own rate from the seller panel and your commission invoice.

You cut the price by 5%; your margin fell by 28%. To collect the same profit you now need roughly 40% more units — and every one of those units brings extra packing, extra shipping and extra return risk. The thinner the margin, the more brutal the leverage: starting from an 8 RON margin instead of 15 RON, the same 5% cut would wipe out more than half of it.

The one solid rule that follows: where the discounting stops is decided by your floor price, not by your rival. And on eMAG the floor formula differs from Trendyol's, because the price contains no VAT:

eMAG (ex-VAT plane): floor = (cost + shipping + target margin) ÷ (1 − commission rate).
Trendyol (VAT-inclusive plane): floor = (cost + shipping + target margin) ÷ (1 − commission rate − VAT share).
Apply the Trendyol formula on eMAG and you are deducting a tax the price never contained: your floor comes out too high and you price yourself out of the competition without noticing.

State the base openly: both the table and the formula above charge commission on the ex-VAT price, because on eMAG the commission is taken on the ex-VAT amount. eMAG's price field is defined literally as VAT-exclusive in the official documentation (§2.4): VAT is collected from the customer and passed on to the state, it never enters your pocket — so it is not part of the commission base either. What does vary is the rate itself: read your own rate from your commission invoice, it varies by category.

To try it on your own product, use the eMAG commission calculator: enter your ex-VAT price, commission rate and country, and see what is left for you.

How far can you raise your price without losing the button?

With no second price you cannot read the ceiling — but you can measure it. Every observation is a piece of feedback:

• Rank 1 means you were winning at that price → the “highest price at which I won” moves up.
• Rank above 1 means you were losing at that price → the “lowest price at which I lost” moves down.

Those two numbers bracket the ceiling from both sides and converge on the real limit within a few rounds. The critical detail: measurements go stale. Because competitor prices move, a bracket older than 48 hours no longer describes today's market and should not drive a decision. Ceiling discovery is therefore not a one-off task but a continuous observation loop.

And the limit of the promise, stated plainly: this mechanism measures the ceiling; it does not guarantee the button. When non-price factors are in play, even your best-margin price may not bring it back.

“I changed the price — why didn't it show up immediately?”

The answer is not slow software; it is the marketplace contract. The limits written in the eMAG API documentation are 1 request every 3 seconds and 20 requests per minute — and that limit is shared across all resources, so pulling orders and writing prices spend the same budget. Reads return at most 100 records per page; bulk writes carry at most 50 offers per request.

The arithmetic: a catalogue of 3,587 products takes about 36 requests, roughly 2 minutes, to sweep. That is the same whichever tool you use — the limit belongs to the marketplace. Invalid requests are additionally capped at 3 per minute and 180 per hour, so answering an error with aggressive retries only slows your whole account down.

One more trap: by the documentation's own warning, an offer may have been saved even when the response comes back with isError=true. So writing a price does not end when the request leaves; it ends when the next offer sweep confirms the observed price.

What happens when your floor price rises above the upper limit?

Short answer: writing a profitable price for that product becomes impossible, and the right move is not to force the price but to update the band in the panel. A frequent scenario most sellers miss: your cost went up, your floor price rose, but the offer's upper limit is still at its old value. Writing a profitable price for that product is now technically impossible — no engine, however clever, can write outside the band. The right response is not to force the price but to update the band in the panel. And this must not be skipped silently; it should be reported as a visible reason: “the eMAG maximum price is below your floor price”.

What does Verimle do on eMAG — and what does it not do?

The honest framing: Verimle does not win the Buy Button for you. What it does is measure and record. Once your eMAG account is connected, each offer sweep records your rank, the winning price and the number of offers; you follow a drop in rank in the panel; it keeps price history; and when it computes a margin-protected repricing target it never goes below your floor price and clips the target into the mandatory price band.

The scope, item by item:

• You follow your Buy Button rank in the panel. Every offer sweep measures your rank, the winning price and the offer count and writes them into the panel, so you see a drop per product together with its history. The chain that produces an automatic button-loss alert is wired on the Trendyol and Amazon side.
• Profit on eMAG is computed from your order data and labelled “estimated”. The reason is technical and clear-cut: the eMAG commission invoice returns a document total, not an order/SKU breakdown. Profit therefore comes from your orders and the commission rate you entered, and the label states honestly what that figure is.
• The eMAG API has no endpoint for filing a deduction dispute; invoices are read-only.
• Seller score and star-review tracking are unavailable on eMAG — the API has no equivalent.

A setup built around those limits does not have to stay on one channel: we looked at how the same product leaves different profit on Trendyol and eMAG in the multi-marketplace selling strategy. If you want to see which eMAG data we pull, the eMAG integration page lists the scope item by item. For official seller resources, see eMAG Marketplace.

Three frequently asked questions

What decides your eMAG Buy Button rank?

eMAG does not publish the formula and the API returns no reason: you see only your position, the winning price and the number of offers. The only hard limits are on the price side — a price outside the mandatory band is never saved, and prices are compared on the VAT-exclusive plane. You can infer a non-price factor only when your rank moves while your own price does not.

Do I have to cut prices to win the Buy Button?

Price is the one lever you control, but the cost is disproportionate: in the example above a 5% cut erases 28% of the margin and demands roughly 40% more units for the same profit. That is why your floor price, not your rival, should decide where discounting stops.

How far can I raise my price after winning the Buy Button?

eMAG never returns the second-place price, so there is no ready answer; the ceiling is measured. The highest price at which you won and the lowest at which you lost bracket it from both sides, and because competitor prices move, an observation older than 48 hours should be treated as stale.