Supply-side platform
An SSP that was designed for breaks, not retrofitted from banners
Most supply-side platforms selling connected TV grew up selling display, and it shows in the places that matter: pods handled as loose impressions, timeouts tuned for a page that can reload, identity models built around a cookie that does not exist on a television.
Four things CTV does differently
The unit is a break, not an impression
Fifteen seconds inside a ninety-second pod is worth something different depending on what else is in that pod. An SSP that treats each slot independently will sell two competing car brands into the same break and call it a good day.
There is no page to reload
A failed call on the web costs an empty rectangle. A failed call in a stream costs a black screen or a stall in the middle of the show. Timeouts here are a product decision, not an infrastructure detail.
The identifier is the device
No cookies, no fingerprinting - IFA and RIDA passed as-is, with lmt and dnt respected. Frequency has to be enforced server-side because nothing on the TV will do it for you.
Authorization is a file, not a promise
app-ads.txt is how a buyer verifies that the inventory is really yours and that we are really allowed to sell it. It is the first thing a serious buyer checks and the last thing most sellers set up.
What the platform does
Capabilities
- Pod-aware auctions
- OpenRTB 2.6 A break is auctioned as a pod, not as four unrelated impressions that happen to be adjacent.
- Deal types
- Open · PMP · Preferred · Guaranteed Programmatic guaranteed for the inventory your sales team has already promised, open auction for the rest.
- Floor management
- Per placement, region, device Floors are yours. Unfilled at your floor is a valid outcome and is reported as such.
- Supply path
- Declared, one hop We buy from you and sell to the DSP. No chain of resellers between your break and the money.
- Traffic quality
- Pre-bid and post-bid Traffic we cannot verify is not sold; rejections are reported to you rather than silently dropped.
We publish no impression-volume figure. We are early, and a number chosen to impress would not survive your first week of reporting.