From: Scott Spencer [email protected] To: "Smith, Michael" [email protected] Bee: Sara Walsh [email protected], Sunni Yuen [email protected], Kenneth Rubenstein [email protected] Subject: Re: CONFIDENTIAL: Ad Blocking, CLEAN Ads by the Tech Lab Cc: "Klass, Gabby M" [email protected], Tom Shields [email protected], Jonathan Bellack [email protected], Scott Cunningham [email protected], "[email protected]" [email protected] Michael, Thanks for your email and thoughts on the topics of ad blocking and header bidding. Both topics are timely and we look forward to working with the IAB. Your response directly ties ad blocking with the header bidding discussion in a novel way - mainly that the redirect chains of ad serving cause ad blocking. While we agree that long redirect chains are a contributor to the size of code (megabytes) required for consumers to download as well as direct ad load latency, it's not clear to what degree this impacts ad blocking installations relative to other ad experience issues, nor to what effect header bidding may alleviate this consumer load. Indeed, some header bidding implementations may require more code and callouts than a redirect chain, since header bidding is installed on every page, versus redirects which only happen on some ad queries. To date, we have not seen a neutral study of the latency impact of header bidding across an entire Website on varying devices and connection speeds. We would welcome the IAB's leadership in such a study. We do think that the attempt of some publishers to call multiple exchanges in an attempt to get marginally higher yield can have a negative impact on the consumer experience on those sites. We think it would be useful for the IAB and others to help publishers understand the consumer tradeoff for any of their ad implementations so that they can assess the trne life-time consumer impact of the ad experiences they provide. Indeed, web-site analytics products can be used to help publishers assess session usage and number of sessions so that they can correlate it with their ad experience. We would love lo work with the IAB as well as top publishers lo better understand what causes consumer annoyance with ads and leads to ad blocking. Currently, we are reaching out to the IAB and other groups to design a study to assess the factors and levels of those factors that drive consumer annoyance. Separately, we are happy to discuss different options for other exchanges to be used via DFP. Regards, -scott On Sun, Oct 11, 2015 at 1:50 PM, Smith, Michael Scott, T hope you are well. wrote:
On Friday, Scott Cunningham and Dave Moore hosted a 90 minute conference call for the executive committee of the IAB Tech Lab. We discussed two topics: (1) Ad Blocking, and (2) Header Bidding. The latter became an understandably spirited discussion between Tom from Appnexus and Jonathan from Google. All productive from my perspective. I would like to merge these two issues for your review and consideration. Also, I'm not including the entire executive committee here - instead I'm adding my EXCOM backup at Hearst, Gabby Klass, who attended the IAB Tech Lab September meeting on my behalf and is fluent with IAB Tech Labs mission, ad blocking, header bidding and Hearst's 'waterfall' based monetization.
Here is our logic and appraisal:
Mike Ht The root cause is performance; goes far beyond 'ads are annoying' At the surtace, the distraction of ads, particularly more intrusive formats like auto-play video or pop-ups, is a factor dri\ blocking adoption, though the catalyst, in our view, is a deeper issue of site performance and data costs. As a result of advertising technology industry having evolved in a highly fragmented manner with many (hundreds} of intermediarie serving, optimization, analytics, and tracking code on a site, and deploying more complex ad formats like rich media, p times have deteriorated. According to Catchpoint, sites with heavy ad tech monetization like news sites can be 4X larg, slower than sites with fewer tags as 62% of the size and 45% of the speed can be attributed to third parties. The issue can be greater on mobile where load times may deteriorate further over weaker cellular connections. Ads ca1 negatively impact battery life and data usage, which eat into consumers' data plans. According to iOS ad blocker Cryst, publishers' load times improved by more than 80% with Crystal installed while page sizes can be reduced by more tha1 terms of battery usage, according to research by UC Berkeley, ads on average consume 65% of an app's communicatio 23% of an app's total energy demands across a select group of publishers. From: Jonathan Bellack [email protected] Date: September 30, 2015 at 8:24:45 AM EDT To: "Smith, Michael" [email protected], Scott Spencer [email protected] Cc: Scott Cunningham [email protected], David Moore [email protected], Anand Das [email protected], Kelly Petersen [email protected], Tom Shields [email protected], Srini Venkatesan [email protected] Subject: Ad Blocking, CLEAN Ads by the Tech Lab It is great that the IAB and the Tech Lab are stepping up here. I am adding Scott Spencer to this thread. If you don't know him, he is the founder of Google's Ad Exchange. He is now leading Google's work around ad blocking, as well as other critical issues like fraud/ TAG, and SSL. In general -- we are aligned that there needs to be a new standard, that it should be defined by the industry and not ad blockers, and that the IAB should be a big part of this process. We have some ideas in that direction already. That said, there are other parts of a response plan that we think are equally important to finding a successful new equilibrium. Scott (Spencer) can share more and would be happy to work with this group, and/or any of you directly. HIGHLY CONFIDENTIAL GOOG-DOJ-15643122 Case 1:23-cv-00108-LMB-JFA Document 1247-9 Filed 08/24/24 Page 2 of 6 PageID# 92117 On Tuesday, September 29, 2015, Smith, Michael [email protected] wrote: Scott, thank you. What would be required to involve myself in this particular effort please: our private legal investigations continue as well as IAB Europe and others wanting to beat the drum in unison. With kind regards, Mike From: Scott Cunningham [email protected] Date: Tuesday, September 29, 2015 at 10:11 PM To: David Moore [email protected], Jonathan Bellack <jbellack@ google.com>, Anand Das [email protected], Kelly Petersen [email protected], Tom Shields [email protected], Hearst Magazines [email protected], Srini Venkatesan [email protected] Subject: Ad Blocking, CLEAN Ads by the Tech Lab Hi Folks. I wanted to update you before our ExComm call on the Ad Blocking front and give you the latest from me, but also solicit some feedback. There has been quite a bit of press lately as you know on Ad Blocking and on IAB's site discussing our initiatives in this space as well as announcing work groups to cover a few areas. It's been overwhelming actually. We are highlighting all of the migration to HTMLS ads as a step in the direction to help User Experience, performance and security. We are referring to our endorsement of SSL/HTTPS delivery. We are saying our ads are bloated, with too many data calls. We should only request an ad when in view. Even with many data calls, our engineering power in the industry should be used to imagine a better data collection process than all of the data calls. What that is, I don't know. As ExComm and Board, I think we should look at how that gets commissioned. We are ready to beta test a detection script for publishers as announced today and I am looking at data from some publishers on their experience experimenting with user tolerance levels. We have been evangelizing experimenting, and in some use cases, users are responding. I urge all publishers to share, anonymously of course, any data they can share with me. So please spread that to your network to provide this intelligence if they are willing. I also had an hour long conversation with the Chairman of Eyeo GmbH https://eyeo.com/ in Germany and they emailed me again today stating they are starting a standards board for all ad blockers who join HIGHLY CONFIDENTIAL GOOG-DOJ-15643123 Case 1:23-cv-00108-LMB-JFA Document 1247-9 Filed 08/24/24 Page 3 of 6 PageID# 92118 them in the "acceptable" ads program. To me this is like robbers forming a union. And I will probably say that in the press when I am asked. And our private legal investigations continue as well as IAB Europe and others wanting to beat the drum in unison. This program has become bigger than Fraud or anything the IAB has faced I am sure. Trying to synthesize all of the work that is happening, most of it in the Tech Lab. The Tech Lab staff came up with a rallying name for the program. Some principles built around our standards and best practices and how we are approaching Ad Blocking and most importantly User Experience. It is called C.L.E.A.N. Ads C: Clutter Free Experience, Cross Screen L: Lean. Standards should be light in file size with strict data call guidelines reducing latency. E: Encrypted. All supply chain transactions should be https compliant. A: Ad Choices Support. All Ads should support consumer opt out programs. (some don't want to conflate ad blocking with privacy, I get it. But we need to have more teeth to our programs too I believe.) N: Non - invasive/disruptive. Standards based Advertisements should not interfere with user experience. This includes covering of content and sound enabled by default. Some of this we have already started with https, htmlS ads. We need to kick the DAA in the ass a bit to up their game. And we need to think in terms of what it means to be clutter free. All of this gets executed in Tech Lab work groups with releases on standards and guidelines. I am privately socializing this, but want to get feedback here first. It describes our work as a body in support of user experience while also supporting a ad supported, open world wide web with standards. Thanks folks, and please share any thoughts on the CLEAN acronym or anything else you may have. Scott Scott Cunningham General Manager, IAB Technology Laboratory HIGHLY CONFIDENTIAL GOOG-DOJ-15643124 Case 1:23-cv-00108-LMB-JFA Document 1247-9 Filed 08/24/24 Page 4 of 6 PageID# 92119 SVP Technology and Ad Operations, IAB INTERACTIVE ADVERTISING BUREAU 116 I:: ast 27tt1 Stroot, 7th I toor. New York, New Yor'k 10016 Mobile Skypo Scott Cu1111inghRl"LDigitai http://www. iab. neUtechlab fonathar Bellac~ / [email protected] Jirecto1ยท1 Produc l\fana(;cmt:'nt / r1,b!isher r,d P'atforrns HIGHLY CONFIDENTIAL GOOG-DOJ-15643125 Case 1:23-cv-00108-LMB-JFA Document 1247-9 Filed 08/24/24 Page 5 of 6 PageID# 92120 Ill. The root cause is performance; goes far beyond 'ads are annoying' At the surface, the distraction of ads, particularly more intrusive formats I ike auto-play video or pop-ups, is a factor driving ad blocking adoption, though the catalyst, in our view, is a deeper issue of site performa nee and data costs. As a result of the advertising technology industry having evolved in a highly fragmented manner with many (hundreds) of intermediaries inserting ad serving, optimization, analytics, and tracking code on a site, and deploying more complex ad formats like rich media, page load times have deteriorated, According to Catchpoint, sites with heavy ad tech monetization like news sites can be 4X larger and 2X slower than sites with fewer tags as 62<:;10 of the size and 45% of the speed can be attributed to third parties, The issue can be greater on mobile where load ti mes may deteriorate further over weaker eel lula r connections. Ads can also negatively impact battery life and data usage, which eat into consumers' data plans. According to iOS ad blocker Crystal, select publishers' load times improved by more than 80% with Crystal installed while page sizes can be reduced by more than 60%. In terms of battery usage, according to research by UC Berkeley, ads on average consume 65~10 of an app's communication energy and 23% of an app's total energy demands across a select group of publishers. HIGHLY CONFIDENTIAL GOOG-DOJ-15643126 Case 1:23-cv-00108-LMB-JFA Document 1247-9 Filed 08/24/24 Page 6 of 6 PageID# 92121