Conversion rate on Amazon and MercadoLibre: what number is normal on each
September 23, 2026
A seller opens the Amazon report and sees a listing converting at 11 percent. They open MercadoLibre and see 2 percent on the same product line. And there begins a story told badly in thousands of WhatsApp threads: “MercadoLibre does not work for me”. Sometimes that is true. Most of the time what happened is that they compared two numbers that do not measure the same thing, computed over different universes, and drew a business conclusion from a badly posed division.
Conversion rate is probably the most quoted and worst interpreted metric in ecommerce. Everyone knows the formula. Almost nobody checks what the numerator counts, what the denominator counts, and what kind of visitor is being talked about. And when you sell on two marketplaces and also run your own store, those three questions have different answers on every platform.
The cost of the confusion is real and gets paid in decisions. A seller who believes their Meli conversion is bad drops the price and burns margin to fix a problem that does not exist. Another who believes 11 percent on Amazon is excellent never learns that in their category it is ordinary, leaning low. Both are optimizing against the wrong reference.
This article sorts that out: what each channel actually measures, which numbers are published and where they come from, why marketplace conversion does not compare to an owned store, and how to build your own baseline when the internet averages do not apply to your case.
the same word, three different formulas
Start with what almost nobody checks: each channel’s operational definition.
On Amazon, the metric is called Unit Session Percentage and it appears in Seller Central’s Business Reports, in the detail page sales and traffic report by ASIN. It is computed as units ordered divided by sessions. A session is a unique visit to your detail page within a 24-hour window: if the same shopper enters five times that day, it counts once. Watch the numerator: it counts units, not orders. If one customer buys three pieces in a single purchase, all three add up.
On MercadoLibre, the Metrics section of the seller panel works with sales divided by visits to the listing. Visit counting follows a different logic, and sales are counted as sales, not units. MercadoLibre also compares you against your category average inside its own panel, which is by far the most useful reference you will find for that channel.
In an owned store —Shopify, for instance— the standard conversion rate is transactions divided by site sessions, not by product page sessions.
Three different numerators (units, sales, transactions) over three different denominators (detail page sessions, listing visits, site sessions). Putting those three percentages in the same table and comparing them is an arithmetic error before it is a judgment error. We have written about how these asymmetries repeat everywhere in Amazon vs MercadoLibre: the differences that complicate your data.
what conversion is normal on Amazon, and according to which data
Here there is published material, and it is worth citing with its provenance attached.
The compilations circulating in 2026 —assembled by seller-tool vendors from aggregated data across their own users— place a typical Amazon seller’s conversion between 3 and 10 percent, with high performers above 15 percent. Some of those same sources report an overall average near 10 percent and describe the 10 to 15 percent range as “good”.
By category, one of those compilations publishes ranges such as electronics at 10 to 15 percent, home and kitchen at 8 to 12, health at 7 to 11, beauty at 6 to 9, toys at 5 to 8, and clothing and footwear at 3 to 6 percent.
Two caveats you cannot skip. First: that data is explicitly from the United States marketplace, and its authors say so. There is no equivalent public benchmark for Amazon Mexico. Second: it does not come from Amazon, it comes from companies selling software to sellers who aggregate their own customer base, which is not a random sample of the marketplace. Useful as an order of magnitude. Not useful as a target.
the MercadoLibre benchmark that does not exist
If you go looking for the equivalent figure for MercadoLibre, you will find very little, and the little there is has to be handled with tongs.
What circulates is a reference around 1.5 percent published by regional consultancies, with no sample country, size or methodology specified, and with a note that new accounts usually run lower. Training material from the ecosystem tends to cite the global ecommerce range of 1 to 4 percent rather than a channel-specific figure. In other words: there is no public, verifiable conversion benchmark for MercadoLibre Mexico. Whatever number someone presents as one, ask for the source.
The good news is that for this channel you do not need an internet benchmark, because you have a better one inside the panel itself: the comparison against your category average that MercadoLibre shows in its Metrics section. It is the only reference computed over the same market, the same definition of a visit, and the same kind of buyer as yours. It is worth confirming in MercadoLibre’s help center how it is defined today, since that section changes shape fairly often.
Glossary: Buy Box, the buy box that decides who gets the sale →why a listing converts far better than an owned store
This is the part that organizes your thinking most, and the part almost nobody explains.
The 2026 measurements of owned-store ecommerce place the sector’s average conversion between 2.5 and 3 percent, and Shopify stores between 1.4 and 1.8 percent, according to aggregations by analytics vendors. Again, these are mainly US and European figures. Set against the 8 or 10 percent reported on Amazon, the gap looks enormous. It is not, really, because we are not talking about the same visitor.
Marketplace traffic arrives with intent already formed. Someone typing “heavy duty school backpack” into the Amazon or MercadoLibre search bar is one click from buying: they are already inside the platform, their payment method is already saved, they already know the shipping cost and they already trust the channel’s guarantee. All the work of generating intent, building trust and capturing the payment method is already done before your listing appears.
In an owned store, that same visitor arrived from a social network or a general search engine, often with no immediate purchase intent, and still has to decide whether to trust a domain they do not know, type in their card details, and accept a shipping cost they discover at checkout. Low conversion is not a failure of the store: it reflects a longer funnel that starts earlier.
There is an economic consequence that follows from this and is worth holding onto: that pre-manufactured marketplace intent is not free, you pay for it in commission. An owned store converts less but leaves you more per sale and, above all, leaves you the customer. Comparing conversion rates without commission and margin next to them leads to conclusions that are exactly inverted.
what actually moves a listing’s conversion
When a listing’s conversion falls, the reflex is to cut the price. Sometimes that is right. Often the problem is something else, and cutting the price only burns margin without moving the needle.
- The Buy Box. On Amazon, if you share a detail page with other sellers, sessions still count for that ASIN even when the sale goes to whoever holds the buy box. Losing the Buy Box sinks your conversion without a single thing changing in your listing. We work through it in Buy Box and price.
- Stock. A listing that is out of stock, or whose available quantity is lower than the real one, stops converting for reasons that have nothing to do with content. It is the most frequent case and the easiest to overlook.
- The delivery promise. The same product with fast free shipping converts differently than with week-long delivery. On both channels the logistics badge weighs as much as the photo.
- Reputation and reviews. One recent negative review can move a mature listing’s conversion more than a 5 percent price adjustment.
- Listing content. Photos, complete attributes, technical sheet. On MercadoLibre, unanswered questions are a silent conversion leak on top of that.
- Traffic mix. A spike in paid clicks on broad keywords raises visits and lowers conversion even when everything else is unchanged. That is why conversion should be read alongside CTR and click source.
how to set your baseline when averages do not apply
With benchmarks from another country, another channel and another formula, the only reliable number is your own. Building it is easier than it sounds.
Take three stable months, with no Buen Fin or Hot Sale inside, and compute conversion per SKU and per channel, each with its own native formula, without converting them to a common unit. The point is not to make them equal: it is to have two clean series. Also note, for each SKU, whether you had stockouts, whether you lost the Buy Box, and whether you ran advertising, because those three things explain most of the variation.
From there the reading rule is simple: compare each listing against itself over time, and against its category average within its own channel. Never against the same SKU on the other marketplace. A drop from 9 to 6 percent on Amazon is an actionable signal; Amazon showing 9 while Meli shows 2 means nothing on its own.
And when conversion moves, check in this order: available stock, Buy Box or direct competition, price against the market, recent reviews, and only then content. It runs from cheapest to most expensive to fix, and in that order most cases resolve themselves.
how this reads in iqseller
iqseller does not replace each channel’s traffic reports, nor does it pretend to invent a unified conversion rate that would not exist. What it adds is the context without which that conversion cannot be diagnosed: inventory and money, on the same screen and at the same hour.
The Inventory module shows real available stock per channel, distinguishing what sits in FBA, in Full, in your warehouse or in a 3PL. That is the first suspect when conversion falls, and it is the data point that lives in another tab in the native panels. The Alerts module warns when a SKU crosses its critical threshold, which is exactly the moment conversion starts deteriorating without anyone noticing. And the Profitability module places net margin per SKU alongside it, with the COGS you loaded, the commissions from the Amazon settlement and from MercadoLibre orders, FBA and Full fees, shipping, advertising, and the VAT and withholding breakdown.
That combination changes the question. Instead of “why is my conversion lower on Meli?”, which has no useful answer, you ask “which of my two channels leaves me more margin pesos per hundred visits?”. With the Parent → Model → SKU tree you can do it per product line rather than per loose listing, which is how assortment decisions actually get made.
Think of a backpack seller. Amazon converts at 9 percent and MercadoLibre at 2.4, and for months they assumed Meli was the weak channel. Seeing net margin and volume together, they discover Meli leaves more absolute profit pesos per month because the category commission is lower and so is the logistics cost. Conversion was never the problem; the comparison was.
Conversion rate is an excellent signal when it is read inside its own channel, against its own history, with stock and margin beside it. Taken out of context and compared across platforms that do not even calculate it the same way, it is the metric most likely to make you shut down the wrong channel.