There's one dimension Media Traders often overlook when optimising their Display/Video campaigns: the SSP seat level.
Want to see why? Visit practically any publisher's website and append /ads.txt to the domain. You'll find the publicly declared seller accounts authorised to offer its advertising inventory. 🔍
Why do major French publishers have hundreds of ads.txt entries?
When we looked at three of France's large websites using DSP planners as a reference, our original analysis recorded these figures:
Methodology: the figures above reproduce the examples in the original GreyFox Ads article; they are not a live audit as of November 2026. Ads.txt files change over time. The original post informally called these entries “SSP seats”; technically, an ads.txt record declares an advertising system, seller account ID and relationship, and does not prove that an active seat or unique selling path was used in an auction.
“But why so many seats? How is that even possible?” 🤯
The IAB Tech Lab's ads.txt documentation distinguishes two main relationships:
| Field | What it means | What it does not prove |
|---|---|---|
DIRECT | The publisher controls the seller account listed on the advertising system. | That the route is necessarily the cheapest or most effective. |
RESELLER | The publisher authorises another entity to control the listed account and resell its inventory. | That the reseller adds no value, or that its account was used for this impression. |
For a full explanation, see our English guide: “Programmatic: WTF is ADS.TXT?!”.
Why are SSP seats and reseller connections multiplying?
In programmatic advertising, more available Bid Requests (BRs) can mean more opportunities to compete for a budget. This helps explain the commercial temptation to add connections (a point raised in the Alliance Digitale research cited in the original post). But spend is not mechanically proportional to BR volume: bid eligibility, demand, floors, win rates and throttling also affect delivery.
Even if a publisher opened one or two accounts at every major SSP, that would typically add up to a few dozen accounts—not hundreds.
“So what's the trick?” 🤔
In our analysis, more than 80% of the declared connections were associated with resellers 👀 (an indicative observation, not a verified market-wide average).
With server-side header bidding wrappers, connecting additional demand partners becomes easier without putting every call directly into the visitor's browser. A reseller can be integrated by the publisher and may itself aggregate demand from other resellers. That opens the door to multiple routes to the same underlying impression.
DirectReseller A
SSP accountReseller B → partner
Additional route
These are examples of possible supply paths, not a representation of the exact technical setup of a named website.
That's what we call Header Bidding power strips! 🔌🔌 Our earlier article explains why server-side wrappers can produce complex reseller chains (without implying unlimited technical capacity or zero latency).
When do reseller SSP seats actually add value?
🍀 Not all resellers are bad. Some make genuinely useful contributions, such as:
- Video-player integration: access to inventory inside a player, potentially with competitive floor prices.
- Exclusive ad formats: opportunities that the publisher's direct seats may not offer in the same way.
- Incremental demand or technical capabilities: a route that genuinely improves eligible access, competition or operational delivery (rather than reselling an identical impression at a higher cost).
Those are reasons to evaluate a reseller, not to exclude it simply because of the label in ads.txt.
And when is it just duplication?
⚠️ In many cases observed on the French Open Web, reseller SSP seats compete for exactly the same placements as the publisher's own SSP seats. The Trader may then end up paying additional intermediary fees without gaining a materially different opportunity. 💸💸
| What the reseller claims | What the buyer should verify |
|---|---|
| “More inventory” | Is it actually incremental, or the same placement available through a publisher-owned account? |
| “Better video access” | Is there a real player integration, distinctive format and legitimate in-stream classification? |
| “Better performance” | Do quality and outcomes improve after accounting for fees, auction price, exposure and measurement? |
| “Better reach” | Does it add eligible people or inventory that the buyer could not already reach through another path? |
What are the risks of excessive reselling for Media Traders?
And sometimes you come across some rather dodgy resellers… 😈
- Tricky formats: suspiciously high click-through rates (15% CTR was an example from our original post), or video banners mislabelled as “preroll” when they are not genuine pre-roll placements.
- Bid Request multiplication: splitting auction calls across variants (such as ad size) or using aggressive ad rotation, which may increase the number of opportunities sent without creating equivalent incremental user reach.
(These are warning signs worth investigating, not proof that every high-CTR placement or multi-size request is fraudulent.)
More duplicated routes can also mean greater infrastructure load, less transparent fees and a more complicated supply chain. Ultimately, the advertiser should know whether each extra intermediary is worth paying for.
How can you stop unnecessary reseller fees with Supply Path Optimisation (SPO)?
That's why a robust SPO strategy matters: you want to avoid burning money on intermediaries that add no value. 🔥💸
- Read the publisher's
ads.txtfile to identify authorised account IDs and theDIRECT/RESELLERrelationship. - Check
sellers.jsonto help identify the business behind seller IDs and understand whether it is a publisher or intermediary. - Inspect the OpenRTB
SupplyChainobject (schain), where available, to examine the entities participating in a specific bid request. - Compare real delivery and total costs across routes: spend, valid impressions, effective CPM, quality, win rate, fees and measurable incremental value (subject to what your technology exposes).
- Keep publisher-owned seats and value-adding resellers; challenge duplicate, unexplained or consistently underperforming paths rather than blacklisting every reseller indiscriminately.
These are complementary tools: ads.txt says who is authorised; sellers.json helps identify who they are; schain can show the supply-chain participants for a bid request. See the IAB Tech Lab explanation.
For a related angle, read our analysis of reseller-generated Bid Requests—but don't confuse share of BRs, share of authorised ads.txt entries and share of actual spend. They measure different things.
How did GreyFox Ads split FinTech campaign spend between publishers and resellers?
It's precisely one of our big battles at GreyFox Ads. 🐎⚔️
We have just completed a six-figure campaign for a FinTech advertiser, with this split in our programmatic supply paths:
ONE GREYFOX ADS FINTECH CAMPAIGN · SPEND ALLOCATION
The remaining 17% went to a handful of resellers rigorously selected for their added value—not to every reseller listed in the inventory chain.
This is the spend split reported for one six-figure GreyFox Ads campaign, not a universal target or a claim that the same percentages optimise every advertiser's results. The original article does not provide a counterfactual performance test.
The objective isn't “zero resellers”. It's zero unnecessary resellers. That's how you make the French Open Web more efficient and transparent for buyers. 💪
Sources, methodology and further reading
- GreyFox Ads — original LinkedIn post on the proliferation of resellers.
- GreyFox Ads — Header Bidding power strips (original post linked in the article).
- GreyFox Ads — Supply Path Optimisation (original post linked in the article).
- GreyFox Ads — original French article.
- IAB Tech Lab — ads.txt authorised digital sellers specification.
- IAB Tech Lab — how ads.txt and seller account IDs work.
- IAB Tech Lab — sellers.json and the OpenRTB SupplyChain object.
- GreyFox Ads — Header Bidding power strips.
- GreyFox Ads — the limits of Meta-Curation.
Editorial note: the historical website-entry counts, the >80% reseller observation and the 83%/17% campaign split originate from GreyFox Ads' own analysis. They are not industry-wide benchmarks. The IAB Tech Lab specifications substantiate how authorisation and transparency standards work, not the numerical observations in the original post.