TurtleDove: Ambiguity in level of decision making.

[JoelPM]

TurtleDove seems to have been written with the expectation that a publisher works with a single SSP to request bids. While this used to be true, publishers now work with a number of ad providers and use a solution like prebid.js to decide which ad/bid they will choose (to then send to the adserver and compete there).

It would be very helpful if we could come to consensus on what level of decision TurtleDove is making. I believe there are two options:

1. A global decision over all bids (contextual and IG) that are eligible for the given page (from all SSPs)
2. A local decision over all bids (contextual and IG) that are eligible from a single SSP

If the intent is a global decision then TurtleDove will end up having to implement much of the Prebid.js functionality in addition to the described functionality needed for preserving privacy. If the intent is a local/SSP decision, then TurtleDove can stay fairly focused on its mission and publishers (and prebid.js or similar solutions) can work with the APIs provided by TurtleDove to implement their preferred business models in a privacy preserving way.

This is a request to clarify the level of decision that TurtleDove intends to make.

(I intentionally avoided using the word auction because it is a more narrow definition of decision and can potentially confuse the issue.)

[michaelkleber]

Yes, you're quite right. We should think of TURTLEDOVE-so-far as having scope 2, a local decision over the demand coming through a single SSP.

I think #59 is in some sense the discussion of how to use a sequence of multiple on-device decision-making events, with logic controlled by multiple parties, to recreate the scope-1 global decision. This seems like a more complex goal, of course, but I think feasible as well.

[JensenPaul]

Closing this issue as it represents past design discussion that predates more recent proposals. I believe some of this feedback was incorporated into the Protected Audience (formerly known as FLEDGE) proposal via #59 and #251. If you feel further discussion is needed, please feel free to reopen this issue or file a new issue.