The Only Ad a Marketplace Can Run That Its Sellers Cannot

• The one claim no seller can make
• If you are the seller, the same fact reads differently
• Why marketplace brand ads look weak in a performance dashboard
• Retail media is not brand advertising
• Four conditions that break this argument
• The barrier is not risk
• Private label
• Land-grab
• You have to be able to keep the promise
• The substitution test
• What to run against last quarter's creative
• What this evidence does not prove
• Conclusion: claim only what your sellers cannot

Savis, a personal care and cosmetics brand, sells on Khanoumi, a leading beauty and personal care ecommerce platform in Iran. We made the pre-roll campaigns for both, and the two briefs had almost nothing in common.

For Savis we made product-focused films built for the pre-roll slot and pointed at product pages — on Savis's own site and on Khanoumi.com. They ran at $0.0118 per click and passed 50 million views in a month. For Khanoumi we made something that never put a hero product on screen: a comedy series about what happens to people who buy counterfeits.

The standard explanations do not survive contact with this. It was not budget: both campaigns were funded at brand level. It was not funnel stage: both were aimed at people who had not yet bought. It was not even the category, since the same cosmetics moved through both advertisers. What separated the two briefs is something most media plans never write down — what each advertiser was structurally permitted to claim.

That is the real question behind marketplace versus seller advertising, and the answer is structural rather than budgetary. A marketplace and its sellers cannot advertise the same way, because they are not selling the same thing. The seller sells a product. The marketplace sells the removal of purchase risk. Any ad a marketplace runs that one of its sellers could also have run is an ad the marketplace paid for on its seller's behalf.

What follows is drawn from those two campaigns and the pattern behind them: what each advertiser could honestly claim, why the marketplace's version of the claim is worth more, why it nonetheless looks weak in a performance dashboard, and the four conditions under which the whole argument collapses. It ends with a test you can run against last quarter's creative in an afternoon.

The one claim no seller can make

Start with what a seller can say. Savis can say its product is genuine, effective, worth the price. Every one of those claims is true and worth making. But a counterfeiter can say all of them too, in the same words, with better production values if it chooses. A claim available to everyone in the category carries no information. It can still sell — product advertising works — but it cannot settle the question the hesitant buyer is actually asking.

Now the sentence a seller structurally cannot say: everything on this shelf is genuine, and if it is not, we will make it right. That requires authority over an entire assortment. No individual seller has it.

This is the part most marketplaces have backwards. They treat selection as the asset — the number of products carried. But every additional seller is another opportunity for the buyer to be wrong, which means selection creates the very anxiety the platform then has to spend money resolving. The asset is not selection. It is underwriting: the marketplace is the only party in the transaction whose promise scales across every stock-keeping unit (SKU) on the platform.

Which is why the Khanoumi work looks so strange next to the Savis work. It had no hero product because the product was not the point. It was about the consequence of being wrong — the specific, recognisable misery of having bought the fake. The promise was authenticity; the creative expression of that promise was somebody else's bad afternoon.

If you are the seller, the same fact reads differently

Most writing on this subject addresses platform operators. The more useful reading, if you are a brand selling through marketplaces in the Gulf, runs the other way.

When a marketplace spends its brand budget on product-level advertising, you are being advertised for free. That feels like a windfall. It is closer to a dependency. The exposure belongs to a budget you do not control, aimed at a category you do not own, and it can move to a competitor's products in a planning cycle without anyone telling you.

The diagnostic is uncomfortable and quick: if the marketplaces you sell through stopped running category advertising tomorrow, how much of your Gulf visibility disappears with it? Whatever number you land on is the share of your presence that is borrowed. It is also, precisely, the gap your own brand and performance budgets have to be planned as separate instruments to close.

Why marketplace brand ads look weak in a performance dashboard

Seller advertising has a short conversion path, so it can be judged on efficiency — cost per click, cost per acquisition, return on ad spend. Marketplace brand advertising is doing something slower: shrinking an objection. A buyer who was not going to risk it, and now might, does not produce a click on the day they change their mind.

Run both through the same dashboard and the marketplace ad reliably looks like the weaker campaign. It is not necessarily weaker. It is being measured with an instrument built for a different job. This is worth separating from the related but distinct question of creative built for the auction rather than adapted into it — that is about what the format does to the asset; this is about who is entitled to make the claim in the first place.

Retail media is not brand advertising

The obvious objection: marketplaces run product advertising constantly, at enormous scale, and it works. That is true, and the distinction matters. What a marketplace sells to its sellers is retail media — sponsored placements, category takeovers and search slots on its own property, funded out of seller budgets. When it does that it is not advertising; it is operating as a media owner, selling access to demand it already owns. That is a business line, and a good one. Retail media and brand advertising are different instruments with different payers, and the argument here concerns only the second: the marketplace's own brand budget.

The harder objection is the one worth taking seriously: product advertising can itself be a trust signal. We carry the real thing, here it is, from us makes the authenticity argument through the product rather than around it, and it is verifiable to the buyer in a way that a comedy series is not. A marketplace chief marketing officer (CMO) could reasonably argue that our Khanoumi approach spent brand money on an unverifiable promise.

The honest answer is that both work, and they fail differently. A product ad proves one instance. It does not generalise to the shelf, which is where the anxiety actually lives, and it has to be repurchased stock-keeping unit (SKU) by SKU. A category-level trust argument generalises but cannot be checked on the spot, so it only survives if the operation behind it is real. If a platform can verify sellers and honour returns, the second is cheaper per unit of trust. If it cannot, the first is the only defensible option — and the platform should fix the operation before it buys any more media.

Four conditions that break this argument

The barrier is not risk

Where nobody fears being cheated — everyday staples, low-value commodities — the hesitation is about price or delivery. In the Gulf specifically, delivery reliability and returns handling are often the primary trust axis rather than counterfeit. The marketplace still has a claim its sellers cannot make; it is just a different claim, and the advertising should make that one. The same distinction decides what a brand entering a new market has to change: the shelf sets the brief before the shopper does.

Private label

When the marketplace owns the product, it is a seller in that moment and should advertise like one.

Land-grab

If awareness is the binding constraint, build recognition first. A trust argument is wasted on people who do not know you exist.

You have to be able to keep the promise

An authenticity campaign from a platform that does not verify sellers or honour returns is a lie with a media budget behind it. It fails faster than running nothing, because it recruits exactly the customers who will test the promise, and their disappointment is louder than their purchase was.

The substitution test

Take your four most recent brand-level ads. Replace your logo with your largest seller's logo. If the ad still makes sense, it was a seller ad — and you paid for it.

Score out of four. The failures are rarely bad ads. They are usually good product advertising funded by the wrong budget, doing work a seller would have been glad to do itself.

If you are the seller rather than the platform, run it inverted: replace your logo with the marketplace's. If the ad still makes sense, you are competing on a claim the platform can make better than you can, and you are paying for the privilege.

What to run against last quarter's creative

Take four ads. Score them. Then, for each one that failed, answer a single question in writing: what could we have claimed here that none of our sellers could have claimed? That answer is the brief for the next campaign.

What this evidence does not prove

Two honest limitations before you act on any of this. First, the evidence above is about what each advertiser was able to claim, not a controlled measurement of incremental effect — the two campaigns ran at different times under different reporting conventions, and we are not going to reverse-engineer a lift figure out of that. Second, the campaigns are Iranian. The structure of a marketplace transaction is the same in Tehran and Riyadh; the incidence of counterfeit and the shape of the trust barrier are not. That is exactly why the argument rests on structure rather than on market sizing.

Conclusion: claim only what your sellers cannot

The model worth carrying into the next planning meeting is small. A marketplace's brand budget should only ever buy claims its sellers cannot make; everything else is a subsidy paid to the shelf. Sellers should read the same rule from the other side — anything the platform can say better, let the platform pay to say. Neither side needs a new measurement framework to act on this. Both need to look at four ads and ask who else could have run them.

We arrived at this from opposite ends of the same shelf: the product-led pre-roll work for Savis and the counterfeit comedy series for Khanoumi — made by the same team, in the same year, for advertisers with the same commercial goal. The briefs diverged because the claims available to each of them did.

If you would like the substitution test run against your last quarter of creative by someone who did not make it, that is what we do. Tell us which four ads and we will tell you which ones survive.