– DRAFT –
Improving Web Advertising BG

Wendy: Our agenda for today; I did not see new agenda requests for today
… for Dashboard highlights, encourage people to agenda+ any items that would benefit from synchronous conversations
… if there are the right people here for that discussion
… possibly some conversation about mapping standards work
… where are various proposals in their possible proposal to recommendation track
… what do editors need; what do we need as information in thinking about the standards landscape
… and any other business items people would like to flag for today's conversation?

Agenda-curation, introductions

Wendy: Let's start with introductions for anyone who would like to introduce themselves
… remember that we queue on irc
… on irc channel #web-adv
… and "q+" to add yourselves to the queue

Paul Farrow: I'm new
… I'm a product manager working for Xandr, a sub of ATT
… I'm in Hamburg

Adrian Mason, Also Xandr
… product manager overseeing ad server and publisher SSP product

Wendy: Welcome Paul and Adrian
… Karen is scribing and she notes that we scheduled a one-off discussion on FLoC

Dashboard highlights? https://w3c.github.io/web-advertising/dashboard/

Wendy: information is available on the mailing list

Wendy: Dashboard highlights
… we have been collecting in Github repository, links to a great deal of proposals

<wseltzer> https://github.com/w3c/web-advertising/

Wendy: the ReadMe in repository links to those proposals and dashboard pulls in all the issues raised against them

<wseltzer> https://w3c.github.io/web-advertising/dashboard/

Wendy: You can see there has been a lot of discussion in lots of different places
… Especially where things are housed in the Web Platform Incubator Community Group (WICG) and the Privacy Community Group

<Brendan_eyeo_IAB> I have the FLoC one-off call at 10 AM Eastern this Thursday. Clarifying that this is correct time?

Wendy: those offer places for discussions as well
… Here we focus on business needs and use cases intersections
… I wanted to open it up to anyone who has an issue or question to raise about the proposals we have seen?
… Some issues have been going back-and-forth

Aram: I see a question in Slack about the FLoC call on Thursday
… Is it 10 or 11am?

Wendy: Slack?

Aram: I meant irc

Wendy: Thank you for that clarification; it is 10:00am, not the same time as this meeting

<Brendan_eyeo_IAB> :)

Wendy: that was the most popular time
… one hour before this call on Thursday, 1 April

Brian: Are there any updates on the FLoC origin trials? How are things proceeding with that?

Wendy: Is anyone available?

<AramZS> Does someone have a link to the FLoC call we can put in here?

Michael Kleber: We are getting closer, but origin trial is not turned on yet
… we're at 'soon' but not this minute

Julien: Continuing on FLoC origin trial and announcement last week that FLoC origin trial will not be done in Europe

<wseltzer> [one-off call on FLoC: https://lists.w3.org/Archives/Public/public-web-adv/2021Mar/0020.html ]

Julien: Is there any timeline, or questions on what is holding it up
… wondering it is not there yet, but might be at some point

Michael: Marshall, are you there to address?

Marshall: no update from what I posted last week

Julien: Can we get clarification on what kind of question needs to be answered before moving on? Why is it not yet running

Marshall: we are just doing stage roll-out to make it be able to roll out to rest of world

Arnaud: A follow-up question. Do you still expect to have some trial in Europe once GDPR issues are solved?
… Or do we go straight to whatever stage we are?

Marshall: We are working to get origin trial rolled out worldwide

Arnaud: And same for FLEDGE? Any updates?

Marshall: I don't have any updates on FLEDGE origin trials at this time, or on status of regional origin trials

Michael: I can give you a lower bound for FLEDGE
… the code for FLEDGE is not yet checked into Chrome, as can be seen in source code
… there is some work on FLEDGE, but it is not code complete
… that means an origin trial cannot start before Chrome version 91
… the one that becomes available in June
… So there definitely will not be origin trials sooner than June
… but doesn't mean there will be one in June
… certainly no later

Arnaud: Something in 2021; is that a reasonable target?

<Zakim> wseltzer, you wanted to ask about other side meetings and to

Michael: yes, we expect to have something in 2021

Wendy: I wondered, and note we have been keeping
… other side meetings

<wseltzer> https://github.com/w3c/web-advertising/issues/104

Wendy: there was recently one on PARAKEET and one on FLEDGE and conversion measurement
… Wondered if any leaders of those meetings wanted to report anything further here for input?
… share updated pointers on where we should look for the latest

Kelda: For PARAKEET I can speak to that
… we plan to continue alternate Wednesday meetings at 11:00am ET

<wseltzer> https://github.com/WICG/privacy-preserving-ads/issues

Kelda: a couple topics have been opened in Github; so look there and continue diving in
… we welcome feedback and collaboration on this

Wendy: there is link to issues raised on PARAKEET

Ashwath: On PARAKEET, to add one more thing
… we have some parts on brand safety that we will be adding
… please keep an eye for that, hopefully the end of this week
… And other areas that came up in a similar manner

Brendan: baseline and not directly associated with the proposals, questions
… but just to make sure my understanding is correct
… Right now FLoC implementation expects each browser is assigned a cohort at any given time
… changes weekly
… and above that, each cohort is not expected to be intrinsicly valued
… like soccer moms....drive into set of relatively similarly sized cohorts without intrinsic value, but apart from attaching meaning
… And if we have that baseline
… and I as party develop understand a cohort and choose to start to sell that who have access to same identifier for that cohort
… would that gain a certain level of resistance by people operating FLoCs
… first question, is my understanding correct; and second, are there inklings of business being done
… associated with a type of audience
… is there potential resistance to that part of business?

Wendy: Anyone who wants to answer that?

Michael: I think your understanding of FLoC seems like a reasonable summary
… Cohorts are people who look similar, but don't come with a browser interpretation of what they mean
… up to you or others to understand what the meaning or utility is a FLoC to you
… I don't think I understood your second question

Brendan: week to week my browser will ID a cohort
… if a party can view and make inferences about one cohort
… and another party has access to the same cohort
… could party one, me and party two, someone else, synch data server side
… to share data developed
… is that philosophically anti-thetical to this function?

Michael: From Chrome POV, tracking individual users across sites is not possible
… what you said doesn't sound like a way to track users
… but Chrome is not in role to decide about other people's business models
… what you described doesn't sound like it is in the browser sphere

Brendan: Sounds like it's more about an audience; thank you

Angelina: I had same question as Brendan
… is there a reason a browser is only assigned to one cohort
… could they be assigned to multiple different FLoC

Michael: Assigning a user to one cohort is core to FLoC
… cohort assigned by browser has to make sure we are not finger printing information
… if assigned to different cohorts then it might enable cross-site tracking
… Browser groups people together that are kind of similar, but only one notion of FLoC

Angelina: I guess question is around...just because I am interested in entertainment
… I have other interests
… and advertisers and publishers have definitions
… and so many other behaviors that a browser captures
… so B2B, B2C, my entertainment habits would be different; I fall into different cohorts in society

Michael: What you are talking about is the intended use of FLEDGE/Turtledove
… so anyone who wants to build own audience segments
… that is what TD is all about
… FLoC is about building one segment attempting to make useful for some range of applications
… You are right that it only captures a small amount of browsing information; so useful for somethings but not all as you described

Angelina: Thank you for clarifications

Aram: Off topic of FLoC
… wondered about adding to agenda
… IAB released four standards
… the shortest public comment ends next week
… I see Alex on the call
… Would he be willing to answer questions; and if group is interested to talk about these specifications
… either at end of this meeting or next weeks' meeting before comments period closes

<AramZS> https://iabtechlab.com/standards/addressability-accountability-privacy

Wendy: Thanks, Aram

Alex Cone: May I respond on timing, point of procedure?

Alex: everything Aram said is correct; public comment period ends in a short time period, but we will continue to take your comments
… we wanted to send message to industry that we are working on a transport mechanism for privacy
… If it's not on people's agenda before comment period closes, I won't be a stickler on that
… I won't say no to public comments that come in past the April 8th deadline on that one

Wendy: Thank you
… I see Aram's link
… that the global privacy platform; @
… best practices for privacy tokens

[missed]

[Wendy reads out proposal names]
… Helpful to get pointers here
… This group is not equipped to make a comment as a group, but individuals are encouraged to talk with IAB Tech Lab
… Anything else to share with group here, Alex?

<wseltzer> [[ Global Privacy Platform

<wseltzer> Accountability Platform

<wseltzer> Best Practices for User-Enabled Identity Tokens

<wseltzer> Taxonomy and Data Transparency Standards to Support Seller-defined Audience and Context Signaling ]]

Alex: I appreciate Aram for bringing this up
… if there is web browser feedback on this forum
… we always have at least one person on browser side

<arnaud_blanchard> i have to leave the call, but here is my question: do the google folks have an idea of how long before a flock_id loses half of its original users (like a sort of half life) ?

Alex: not sure what Aram had in mind, but we are flexible to all sorts of forums for feedback
… Let the group decide if you want to take agenda time in future, we would be open to it

Aram: Just a suggestion to offer another forum

<robin> +1

<kris_chapman_> +1 to that too

Aram: I know we discuss Google's proposed standards, so would be good to discuss other group's proposed standards

Alex: plus one to that

Wendy: happy for presentations on upcoming calls on that

<AramZS> Let's add it to the top of next week's agenda?

Wendy: I see some "+1s" on irc so let's see how to arrange that
… next week would be great

Aram: I guess we can add to top of next week's agenda
… not sure what process is

Wendy: Process is bring it to my attention, and since I have not heard other agenda requests, it goes to the top
… I don't see anyone else in queue
… There was a question on FLoC, we should raise it there [on Thursday's call]
… Any other general questions or comments to raise on issues in the dashboard
… hearing none

Mapping the standards work

Wendy: I will come back to the question of mapping standards work
… The goal at W3C is to find areas that are appropriate for platform-wide standardization
… and we can put onto recommendation track for cross-platform, interoperable features for the web
… we are listening for routes to bring those to standards track
… There is incubation going on in CGs, discussions going on here, which are pre-standards, testing and experimentation
… to answer questions if this would work as a standard
… does it satisfy use cases; meet needs of platform; is it implementable
… as we get answers to those questions
… we consider whether to charter working groups
… to take the next step
… There is currently among the proposals we are discussing, none is yet in a working group
… but wonder about people's thoughts if things are close; are we seeing convergence among the proposals; and how can we help

<wseltzer> https://github.com/w3c/web-advertising/blob/main/support_for_advertising_use_cases.md

Wendy: to assess that standards readiness
… one place we look is the use cases document that has been developing
… Link in irc
… where we have some good analysis of use cases contributed by various advertisers, adtech, publishers, user needs
… and doing mapping about how proposals match up against them
… Encourage people to look again at this document and see if there are new use cases
… or how proposals meet those use cases
… and use as an assessment of how we are doing
… and what we might want to recommend a proposal towards a standards track

James: thank you, Wendy for explaining that
… I joined group in March 2020 and it took me time to understand that
… For everyone on this call, there are a hundred more who are not
… communicating what you have described to the wider industry community
… that wider industry has a perception that W3C may be blessing this
… and when words like 'collaboration' sounds like people agreeing to these things

<wseltzer> W3C Process Document

James: So now I understand that is not the case; and CGs have no requirement for consensus
… and many of components are not being put forward for working groups yet for consensus
… Would you be willing to put out a press release, or do interviews with journalists to understand that significant position
… and W3C does not risk its brand being mis-represented

Wendy: You are referring to the W3C process document about the development of W3C standards track work
… which applies to Working Groups and Interest Group procedures, but not to the community groups and business groups
… regarding advertising the status, I'm always happy to address inquiries
… I don't think there is something to put a W3C press release out about, but if people have questions, they can be in touch

Angelina: Going back to business use cases
… in case people weren't aware, the Partnership for Addressable Media
… has put together @ use cases
… written by several of the PRAM partners, IAB...[names several]
… we broke them up into sections around use cases around planning, measurement, hundred-page and optimization

<angelina> https://www.responsibleaddressablemedia.com/businessusecases

and another section on [missed list]
… there is this doc written from advertiser, publisher and agency perspective
… It's a good read
… not sure how to get that into this environment, to get their business side perspective on how things work
… regarding the deprecation o fcookies

<joshua_koran> Given the membership of global advertisers, I recommend PRAM team brief this group on their requirements

<wseltzer> https://docs.google.com/document/d/17geRxY4Wko_yzLkezlbGMIeD7s-DcaJOQfsCMMw8lxE/edit

Wendy: Thank you for that link; I had not seen the document
… Sounds like a good and lengthy read

Wendy: If it has other things for us to consider

Angelina: we had about 100 contributors

Wendy: If there are pieces we have missed in the use case document in this group, would be great to add a reference
… and pointer to this one
… and consider a future conversation between the two
… Who do you think would be the right people to invite for that conversation?

Angelina: I can reach out to 4As or PRAM to see who would lead that conversation; I am not the right person

Wendy: other comments or questions?

<jrosewell> +1 joshua_koran - it would be good to hear directly from advertisers

Wendy: any other business?

AOB

<nics> +1 for PRAM document discussion

Wendy: for our next Tuesday meeting, we will queue up some conversation on IAB Tech Lab's draft documents
… and see about a future meeting as well to queue up some discussion of the PRAM use cases
… and later this week

<angelina> publisher uses cases included: audience acquisition, monetization, content development, monetization, engagement/retention

Wendy: Thursday at 10am EDT we will have a one-off discussion on FLoC
… and if we have not exhausted that, we can raise questions on Thursday
… So today I will give you a few minutes back
… Good discussion and lots of new pointers
… we now have 150+ pages of reading material to look at
… Thanks all

<nics> bye

<wseltzer> [adjourned]