Re: Issue 25 Audience Measurement Research Notes and follow up from our call yesterday

One small correction to the statement, "Ronan explained that other
approaches would degrade the user experience by exponentially increasing
the number of round trips necessary to provide the necessary frequency
patterns.": it would not increase the number of round-trips, but it would
increase the size of each round-trip by a large-ish amount (possibly on the
order of 120KB each, regardless of whether 6,000 counts are stored in
separate cookies or squeezed into a smaller number of cookies with multiple
sets of counts each.)

--ronan




On Wed, Jul 24, 2013 at 11:02 AM, Thomas Roessler <tlr@w3.org> wrote:

> Forwarded with permission.
>
> (And with thanks to Kathy, who's heading into some well-deserved vacation.)
>
> Thomas Roessler, W3C <tlr@w3.org> (@roessler)
>
>
>
>
> Begin forwarded message:
>
> *From: *Kathy Joe <kathy@esomar.org>
> *Subject: **Issue 25 Audience Measurement Research Notes and follow up
> from our call yesterday*
> *Date: *July 24, 2013 16:49:45 +0200
> *To: *"Israel, Susan" <Susan_Israel@Comcast.com>, tlr@w3.org, rigo@w3.org,
> k.joe@esomar.org, rob@blaeu.com, ronan.heffernan@nielsen.com,
> adam.phillips@realresearch.co.uk, rweaver@comscore.com, kim@esomar.org, "
> dan@eff.org" <dan@eff.org>
> *Cc: *"tien@eff.org" <tien@eff.org>, Nicholas Doty <npdoty@w3.org>
>
> Dear All
> Here are the notes from our call yesterday (with some responses to
> additional questions raised) with many thanks for participating at such
> short notice.
>
> I won't be in the call this evening as I leave for vacation so if you have
> any comments or suggestions please respond to Adam and Kim who will be
> handling this in my absence.
>
> With best regards
> Kathy Joe
>
> Notes from call on Tuesday 23 July with Dan Auerbach, Susan Israel, Ronan
> Heffernan and Chad Hage from Nielsen, Richard Weaver, Rob van Eijk, Adam
> Phillips, Kim Smouter and Kathy Joe at ESOMAR****
>
> 1.   The text sent to W3C DNT group (July 23, 2013) explains the purpose
> and importance of AMR and describes the process to answer various questions
> raised in the discussion last week and since. Page 3 answers questions re
> ‘pseudonymisation’ and proposes new wording in the normative text for
> clarification. It was noted that the wording for Issue 25 can still be
> changed providing the issue is not closed and may change again after
> consultation. As the doc has just been sent out, no-one has had the time to
> read it so the conversation did not focus on this document.****
> *Action:* Those on the call are invited to suggest improvements. Other
> proposals may come in responding to the call for friendly amendments and
> these could change or be merged with the existing text. As this issue
> crosses over with the Red, Green and Yellow tripartite concept and the
> Art29WP is drafting an opinion on anonymization, it could be better to keep
> the text open to the possibility to integrate changes later and Rob will
> send in a friendly amendment to this effect.****
>
> 2.   Why can the data not be anonymised immediately? Ronan confirmed that
> detecting fraudulent activity involves scanning patterns for robots over
> time. More might be identified when fresh data comes in as the campaign
> continues, which means reprocessing all the data to remove the doubtful
> elements before pseudonymization and then analysis.****
>
> 3.   There has been an exchange between Mike O’Neill and Ronan on
> alternative methods of doing AMR to avoid using unique identifiers. Is
> there a process of experimentation for a more privacy protective approach
> because with say 6000 cookies being placed on a hard drive, perhaps there
> are ways to reduce this number eg multi identifiers in a cookie?****
>
> Ronan explained that other approaches would degrade the user experience by
> exponentially increasing the number of round trips necessary to provide the
> necessary frequency patterns.  ****
>
> In relation to geographic detail and pseudonymisation Ronan explained that
> geographic detail was limited to DMA’s which identify very large
> metropolitan areas and the last quad of the IP address was deleted noting
> that where higher legal requirements exist they are implemented e.g.the
> truncated IP address is hashed in addition in Germa`ny, as required by
> German law.****
>
> Action:****
> Ronan will circulate a short summary of the conversation and performance
> requirements to help Dan and others understand how other approaches could
> impact performance delivery and the user experience. He will try to provide
> Dan with an example to illustrate the issue although this could be tricky
> as clients rotate ads so frequently.****
>
> Dan will review the explanatory text to see if there are any remaining
> concerns about the scope of Issue 25 and continue the conversation with
> Ronan, noting that it would not be a trivial matter to introduce new
> technology given the performance requirements and the amount of data
> involved.****
>
> 4.   IAB announcement regarding proposed new online audience metrics IAB
> on 3MS (https://www.iab.net/mmms). Whilst this in general fits with the
> activities described by Issue 25, it was noted that the text describes
> aspirations not yet in place and view-through reporting and cumulative
> social activity’ would go beyond the scope of Issue 25 wording.****
> Action: Rob will send the link to Susan who will check with IAB US to
> ascertain the status of this proposal and what it covers.****
>
> *Other questions received since or before the conversation:*
> Jeff Chester: It would be helpful to have a list of all the products now
> developed using the panel measurement method that the proposed exemption
> would cover.   This would include cross-platform and intra/inter-site
> analysis, application related interactivity, etc.  A description from
> comScore, Nielsen and other panel based research products of their products
> and use should be part of this record, at least.****
>
> One of the best ways to evaluate this proposed use on the tracking
> preference of users is to have a clear understanding of measurement's
> actual use and impact on the record.  Pending or potential 3MS measurement
> related products should also be identified (http://www.iab.net/mmms).****
> Answer: it will not be feasible to provide an exhaustive categorisation of
> all the products currently being offered by these companies and not
> possible to describe products that might be offered in future. However
> products will either fit the requirements of Issue 25 or not qualify as a
> permitted use. More information is being sought on the status of the 3MS
> potential product.****
>
> John Simpson: First, I think your opening sentence may be at the heart of
> our difference.  You write:  "The discussion in W3C DNT is naturally
> focused on the technical specifications for enabling people to control the
> extent to which their surfing behaviour is tracked for the purpose of
> delivering tailored content to them as an individual." ****
>
> "Enabling people to control the extent to which their surfing behaviour is
> tracked for the purpose of delivering tailored content to them as an
> individual" is about targeting.****
>
> My understanding is that DNT is about "enabling people to control the
> extent to which their surfing behaviour is tracked." (Full stop.) In other
> words it's about collection.****
>
> Second, you seem to say that audience measurement is essential for the
> market to function.  If that is the case, why do you provide for an
> industry sponsored opt-out?  Is that opt-out just window dressing and you
> expect nobody to use it it?  I just don't understand.****
> Answer:****
> There has been ongoing discussion about the scope of W3C DNT compliance
> and whether it is collection and use or simply use and as we understand,
> this discussion has narrowed in focus but Issue 5, the definition of
> tracking is still open.  Should the consensus be that it is collection
> and use, then we are happy to delete the sentences to which you refer,
> because we believe that Issue 25 as a permitted use is about seeking an
> exemption for collected data for a narrowly defined purpose which will not
> involve a return path to the user.****
>
> We believe that AMR is essential for the market to function but note that
> even experts on the W3C DNT group have needed explanation to understand
> this particular purpose. We believe that describing it in layman’s terms so
> that the average consumer is able to understand what data is being
> collected, the purpose and the fact that clients receive aggregated data,
> they will be less inclined to opt out. This is not window dressing but a
> consumer facing explanation to bring more transparency to the activity of
> AMR.****
>
> Kathy Joe,
>
> Director, International Standards and Public Affairs
>
> [image: Description: Description: ESO_AW_logo_b_c]****
> Atlas Arena, 5th floor
>  <http://www.esomar.org/>
> Hoogoorddreef 5
> 1101 BA Amsterdam
> The Netherlands
> Tel: +31 20 664 2141
>
> www.esomar.org
> This e-mail message including any attachment(s) is intended for the
> addressee only and may be confidential. If you are not the intended
> addressee, we request that you notify us immediately and delete this e-mail
> including any attachment(s), without copying, forwarding, disclosing or
> using this (these) in any other way.
>
> ESOMAR, the World Association for Social, Opinion and Market Research, is
> the essential organisation for encouraging, advancing and elevating market
> research worldwide.
>
>
>