Greetings all, it has been a while since the last post and even longer since Google Analytics tracking was removed and replaced by Matomo.
The good news is that Viewthrough.org has now joined its sister site, The Tip of the Spear Blog, and is now published on WordPress…sayonara, adios and ciao Big G and Blogger.com.
More to follow including a new post about Viewthrough and Safari ITP!
Part 1 – Introduction
This will be a multi-part post about recent developments that should be of concern for serious digital marketers. In particular, large digital advertisers that use Adobe Analytics (or other non-Big G Analytics platforms) and whose media agencies (or internal digital media buying teams) also use Big G Ads/Search/AdWords beware.
The Problem
Apple’s most recent Safari release in September 2018, which effectively blocked all 3rd party tracking cookies is being blamed for what amounts to a clandestine behavioral data land grab by Big G. In going along with this charade, digital marketers everywhere risk the de facto relinquishment of digital measurement and customer behavioral data to others with different and not non-transparent interests that create a competitive disadvantage.
Through major brands’ own media agencies, Big G has been cannily pushing everyone to use a new kind of Web site tracking tag that circumvents Safari’s strict privacy features. The new code snippet replaces the legacy DoubleClick Floodlight tags that have been historically appended the brand’s destination Web site on key actions, e.g. order thank you page. The loading of that page and associated 3rd party tags signal a successful “conversion” event for analytics and ad targeting optimization purposes.
Floodlight 101
The DoubleClick Floodlight tag was created as an adjunct to the agency ad server (a/k/a 3rd party ad server) which was initially developed to enable ad impression delivery in the user’s browser to be associated with later events on the advertiser’s Web sites. In this way, a kind of campaign measurement could be performed showing cause (ad delivery) and effect (purchase event). However, the method relies a 3rd party cookie from doubleclick.net set at time of ad delivery to be safe and secure in the user’s browser and recognized by the Floodlight tag when the user subsequently interacts on the advertiser Web site.
Additionally, the Floodlight tag has the capability to piggyback other tags from media vendors like ad networks so that they could also measure and optimize their campaigns in flight based on actual performance, e.g. order confirmation pages firing. Similarly, ad analytics uses include brand lift studies and multi-touch attribution. Not surprisingly, over the years Big G has shoved more and more back-end server-side functionality into the Floodlight tag to help their other businesses like AdWords, Bid Manager and Dynamic Creative Optimization.
The problem with the new code snippet is that the tag is no longer the kludgey Floodlight but the much more powerful Big G tag manager. The latter was engineered to sending data back to Big G faking 1st party even though it is not really (similar to Big G Analytics). Compounding the challenge for digital marketers is the aggressive recommendation to install the new GTM code snippet onto *all Web site pages* instead of just swapping out existing Floodlight tags.
Within the new code snippet itself, note that the comment text still says “DoubleClick” but the actual server being called is not doubleclick.net anymore but is googletagmanager.com. Big G’s timing of the recent name change dropping the DoubleClick brand from everything is interesting and has further caused confusion with the site operations and ad operations team that work with these technologies.
Major digital advertisers that are not already using Big G Analytics/Tag Manager are particularly susceptible to the false sense of urgency in this crafty switcheroo. Especially those from industries with limited digital competency. In the end, Big G gets better clickthrough conversion tracking for search and client-owned Web site analytics platforms like Adobe Analytics, Webtrends, IBM Digital Analytics or Matomo remain as is.
Clients should consider pushing back on what amounts to a digital behavioral data cookie grab.
COMING UP: Part II – Victim-blaming with Apple ITP.
It may be getting rough for the advertiser/agency ad server marketplace with recent developments squeezing viable independent alternatives to Big G. Digital marketers and their media agencies would do well to think through the impact on their ad tech stack instead of the usual.
These challengers (and others) could still be a worthwhile alternative to Big G’s near monopoly on advertiser/agency ad serving. With the heavy hand of GDPR and CCPA looming regulators are looking to score political points and have a real need to fund overstretched governments with fines.
The bad news is that Sizmek is entering bankruptcy proceedings which itself is a roll-up of many companies. Formerly known as MediaMind, it is made up of Pointroll, Eyewonder, Eyeblaster, RocketFuel/x+1, Peer 39, StrikeAd and others. From the outside, it looks like self-inflected wounds are the cause and not a problem with the quality of their technology products per se.
According to AdExchanger, Sizmek owes around $26MM to various ad exchanges/networks. Missed revenue targets on the tech-side at the same time that they were trying to consolidate operations is mentioned. It would not be the first time that a lack of operational savvy by the financiers caused problems. Just because combining a media business model and technology model works for the dominant player in the market does not mean it is a good strategy for challengers. Mashing up a challenger technology product company with an overglorified ad network is usually a bad idea.
TrueEffect does not seem to be making much progress and is rumored to be having trouble. That would be unfortunate as their unique and innovative approach first-party ad serving would improve measurement of KPIs like reach, frequency and viewthrough measurement. It is too bad that more advertiser’s and their agencies choose ad servers based on expedience without considering the hidden costs.
Next, there is Flashtalking with their nascent relationship with Adobe. In 2018, it was announced that the two would collaborate on integrating the Flashtalking ad server with the Adobe Media Optimizer DSP (demand side platform) which would together improve identity management in part of the ad stack. On the heels of Adobe Summit 2019, there is not much new to get excited about. For their part, Flashtalking brings a novel approach to identity management with their ad server leveraging probalistic machine fingerprinting methods to boost targeting and analytics efficacy. Limited reliance on 3rd party tracking cookies seems to be a good proxy for first-party browser tracking.
The Atlas ad server is history. In 2013, Baby G acquired the Atlas ad server technology from Microsoft (an early funder of Baby G) and then ran it into the ground. Briefly, there was potentially big funding around and it was all about people-based ad serving with an obvious identity management benefit. Alas, the towel was thrown in by Baby G in 2016. Riding two different horses is hard. Prior to Microsoft it was part of the aQuantive group which also included the DrivePM and the Razorfish agency. Microsoft also let the Atlas ad serving technology go to waste which is thought to be a me too move and also to prevent Big G from getting another DoubleClick.
Meanwhile, it is all sunshine and rainbows in Mountain View. Big G continues to scam advertisers/site-owners for more data with the old DoubleClick Flooglight for Google Tag Manager switcheroo (more on this soon). Conveniently, they and their media agency lackeys blame Apple Safari ITP for the land grab under the auspices of: it is to improve measurement [theirs primarily it seems].
Good news for Web site owners and digital marketers!
The Tip of the Spear Blog is now using the open-source TMS (tag management system) offered by Matomo. In fact, I installed it myself with one auto-generated tracking tag (Matomo) in about 15 minutes. It was a little tricky seeing all the moving parts for the first time up close and personal. Still, it was good to see a new TMS option available for Web site owners.
Matomo (f/k/a Piwik) for those who don’t know has been a low-cost open-source platform for Web site analytics for about ten years. Matomo has more popular outside the US where people have not drunk Big G’s fake free kool-aid. Matomo is highly recommended for those that do not want to contribute to Big G’s data ingestion Trojan Horse a/k/a Google Analytics and Google Tag Manager. All you need is a Web server and instance of MySQL.
Matomo’s site analytics and tag management technology is highly recommended for those that do not want to share their user’s Web site behavioral data with Big G’s data ingestion Trojan Horse (a/k/a Google Analytics and Google Tag Manager). All you really need is a Web server and instance of MySQL.
The bad news is that in the process, I discovered that even after removing any and all WordPress plugins, there still seem to be phantom trackers including several for Google: Analytics, Display Bid Management, Publisher and Tag Manager. How is that possible?
TOTSB get that nothing is free, but nowhere during the WordPress blog creation process did it mention that my blog would be providing tracking fodder for Big G and other targeting wannabes. It is this kind of sleight-of-hand and non-transparency that leads to ignorant users, false expectations and then misguided calls for government regulation.
TOTSB will continue looking for ways to remove these trackers but will also look to see about alternative blog publishing platforms, if needed.
In other news, the sharp folks at Evidon have released a new tool that takes tag management up a level into tracker management. Evidon is known for their privacy consent capabilities and is where Ghostery originated. They also pioneered tracking monitoring a few years back with their TrackerMap service that allows Web site owners to understand the latent piggybacking going on from their supposed partners. The new “tracker management” service gores from the tree-level up to the forest level view. It requires a hard-coded co-pilot tag on each page that can actually suppress piggybacking, i.e. stop it cold from sketchy partners and agency noobs. Bravo Evidon!
UPDATED 2/11/19
And the beat goes on…
The UK-based Guardian’s John Naughton joins a growing line of sanctimonious journalists carping about digital advertising tracking in, “The goal is to automate us’: welcome to the age of surveillance capitalism.” The article rehashes Shoshanna Zuboff’s tropey new book Surveillance Capitalism once again just as Motherboard did a couple of months back (mentioned earlier in this post). For an ex-Harvard Business School professor, the book sounds like a post-modern Das Kapital.
With at least 54 trackers at the time of this blog post Guardian joins many other digital ad tracking hipocrites. John Naughton is not likely working for free and Guardian is generating plenty of ad revenue as a result. One can see clearly see the banner ad at the top of the page and in case you don’t read the article there are another 11 ad positions on the page! Last, Guardian harangues site visitors begging for donations with a zeal comparable to the NPR and PBS fund drives. The strategy must be working as digital revenue is up at Guardian Media Group.
Trackers revealed per Ghostery plug-in
Similar to Motherboard, it should be no surprise that although Facebook and Google are thoroughly trashed by the author, there are in fact, multiple trackers from each on the article Web page. Apparently it is OK for Guardian and Naughton to make a buck from digital ad targeting but Facebook and Google should not?
Guardian.com Diagram shows Rampant Trackers
A quick look at the tracker diagram for that article shows the hipocrisy of Guardian, who is really filling up at the ad tracking trough. As you can see John Naughton’s publishers are certainly not working for free. How long would John and the Guardian last if they took up Shoshanna Hoffman’s advice and swore-off digital advertising money?
Last, it is interesting that Guardian is hawking Shoshanna Zuboff’s book itself with a link at the bottom at the end of the page. The holier than thou Guardian doesn’t mention any interest in this book which makes the article itself smell alot like undisclosed advertorial.
Message to komrades John and Shoshanna, so-called surveillance capitalism isn’t the biggest problem. As with the other digital tracking Pharisees, government surveillance is minimized to just two sentences. For people pining about democracy, there is no mention of multiple government agencies already feasting on backdoor access and optimized APIs to more easily glom data on users. The kind of industry-government coziness that enables massive spying on the citizenry would have made King George III squeal with delight. Paul Revere, George Washington and Thomas Jefferson would never have had a chance.
Free is Not Without Cost
While, I’m no fan of Big G[oogle] and Baby G[Facebook] advertising practices they do provide useful services to consumers for free that nobody is being forced to use. It is puzzling that some users (Saying No to Free-riders) expect to enjoy the benefits of these services without paying for them or providing information about themselves in exchange.
More worrisome should be the perversion of the free market (a/k/a crony capitalism) that is happening when regulations are foisted upon an industry. Red-tape raises cost of entry and helps incumbents entrench themselves creating monopolies, duopolies, etc… Basic economics explains that reducing the number of suppliers (shifting supply curve S to S1 to the left) while holding demand constant increases prices.
Not a big surprise but disappointing then that some of the biggest players are welcoming the rent-seeking opportunity to be regulated. Railroads, pharma, insurance and banking come to mind as high-margin and high-regulation industries.
Of course bureaucrats and politicians gain but that is a feature not a bug.
UPDATED 1/23/19
Natasha Lomas is the latest writer/journalist that sure likes to get a paycheck but bemoans advertising tracking and the ad tech players that make it all happen. Just another in a string of digital tracking hypocrites in the rather verbose and rambling, The case against behavioral advertising is stacking up over at TechCrunch.
Lomas’ actual premise is an inch deep, e.g. the Digiday reference about the
NY Times International says that they are “still doing private marketplace deals.” Clearly, missing the fact that PMPs use the same programmatic techniques just different ad buying methods: guaranteed placement directly from NY Times assisted by ad sales reps vs. open exchanges buying remnant ad inventory in real-time. Sorry, the facts got in the way of the story!
Here are more facts: Lomas and her editors might be interested learn that Ghostery counts a whopping 86 trackers on the article itself; the tracking network diagram found even more.
The case against poor quality business writing is really stacking up too!
UPDATED 12/4/18
Not to be outdone, comes the steaming hot chicken-little indigence of Motherboard Vice with, Targeted Advertising Is Ruining the Internet and Breaking the World with a whopping 62 trackers, almost 3 times the capitalists-of-convenience at CBS.
Here is their tracker map diagram, which is pretty impressive in the level complexity and likely degraded user experience. It shows how dependent on digital tracking Motherboard Vice is for ad targeting and measurement.
While reading the article and waiting for the call for Workers of the World to Unite rally-cry, I struggled to find a link to the privacy policy on the forever expanding page. An interesting technique that served endless content and more display ad banners as you scroll down.
Eventually, I found Motherboard Vice’s privacy policy explained here – pretty long and complicated as per usual. I did find a link to an article on FAZ.net promoting retired HBS academic Shoshanna Zuboff’s new book about so-called “surveillance capitalism“…which…wait for it…ironically sported 24 trackers of it’s own!
Meanwhile no mention of the NSA, CIA, DOD and FBI surveillance by CBS or Motherboard Vice. Maybe Shoshanna Zuboff’s next book will be: Surveillance Government – The Fight for a Human Future at the New Frontier of Power.
You can’t make this up. In many ways, Web media publishers with ad-supported business models brought this on themselves – see Saying No to Free-riders. None of the above Web sites are using the open source Matomo analytics or Pwiwk Pro tag management product which is big in the EU.
UPDATED 12/4/18
John Stossel recently did a story on “The Creepy Line” a movie that rails against tracking and specifically Big G and Baby G.
A quick look with Ghostery shows the Web site uses both! Google Analytics and Doubleclick tracking was observed. Facebook Advertising and Connect are being used to facilitate social media promotion of the movie – ironic, no? Not to mention links to Amazon.com to see the movie. Disclaimer this blog is a Google-free zone but we’re making an exception here.
Coincidentally, it appears that Peter Schweizer is one of CBS’ producers in NYC…interesting! No mention of government tracking or spying.
Meanwhile, the tracker network diagram reveals even more clandestine tracking (130 total ) than Ghostery; this is likely due to server-side piggybacking.
ORIGINAL POST: 11/19/18
Last week, CBS 60 Minutes broadcast “Your Data” aiming to scare everyone about online ad targeting. The outrage of the story was palpable. How dare Google (Big G) and Facebook (Baby G) track people and target ads to us?
In an irony too good to pass up, a quick look at CBSNews.com’s 60 Minutes Web site showed some 23 different trackers doing the exact same thing!
In the mislabeled but scary “Your Data” episode last week, the hapless Steve has little clue about which he speaks fawning over GDPR the whole time. BTW, how can it be “Your Data” when people voluntarily access somebody else’s Web server for all you can eat news, social media fun and search engine content? Magic bunnies don’t build these Web sites for free.
Alas, CBSNews.com’s Privacy Policy looks just like everybody else’s. Perhaps TV media is threatened. Or perhaps morally superior for forcing people to trade their time instead of their Web browsing behavior in exchange for free all-you-can-eat media.
So many ironies so little time…here you can see the CBSNews.com actually filling up themselves at Big G and baby G’s data trough. Traditional media trying to claw back advertising revenue they lost or journalism at work…hmmm.
If you find any more good ones please reach out to me via LinkedIN.
I was invited back to esteemed colleague and neighbor Professor Aleksandar Velkoski‘s Fundamentals of Data Science (DSC441) class for a guest talk. The summer session class was packed with aspiring Computer Scientists, Information Systems and even a Cybersecurity professionals keen to better understand data science.
Dr. Velkoski’s Data Science Lecture
According to the syllabus, the biggest challenge that data warehousing professionals face is extracting value from vast data warehouse. One of best tools to use is data mining.
The course objectives are to:
Dr. Velkoski explaining the Gini distribution devised by Corrado Gini in 1912
If you are felling up to the challenge feel free to pick up the textbook: Data Mining: Concepts and Techniques, by Han and Kamber, Morgan Kaufman Publishers, Third Edition.
Textbook webpage: http://www.cs.uiuc.edu/~hanj/bk3/
Learn more about Dr. Velkoski’s academic profile.
Response to Martin Kihn of Gartner’s piece at AdExchanger, Did Google Just Kill Independent Attribution?
———————–
Interesting. Digital advertisers relying on those IDs for MTA sacrificed quality for expedience a long time ago. That cookie and anything relying on its stability is the poster-child for unreliable 3rd party tracking, i.e. bogus measurement and imprecise targeting.
#1: I’m glad you brought up trust which hopefully advertisers are paying attention to now in the post-ANA world. Best approach is also “to verify”.
#2: It is hard to believe that advertisers willingly uploading their data to ADS.
#3: Well-said. Some media agencies may be conflicted about this, while others are moving their clients away from the expedient choice.
#4: It should now become more clear: digital advertisers that chose expedience have lost the competitive advantage. Worse, they have been measuring and targeting with garbage data for years.
My recommendation to clients is to run from reliance on this dirty bird and find point solutions where they can leverage 1st party methods/do quality control and own their own destiny.
Stepping Over Data Dollars to Save Pennies
Have a look at this image. Sound familiar? Web analytics has held out the elusive promise of being a set-it-and-forget-it kind of thing. “Set up your reports, and the data will fill itself in.” That promise has largely held true — for every part of web analytics except Marketing. That’s because with marketing, the web page you have today isn’t the one you had yesterday. There’s constant change: new information, new deals, new parameters. What everyone wants is a system that runs itself. Otherwise, as the figure shows, you spend all your time making sure the reporting is right. Spending time on data correction takes time away from the analysis that will really help the company. It’s a necessary evil. Wouldn’t it be great if we could get marketing data to the same set-and-forget kind of place as the rest of our web analytics?
Removed the embedded story blurb as this was inviting Big G to track readers…With focus on transparency, barter agencies get more scrutiny.
Interesting article on AdExchanger, “The Rise Of Addressable Video And Banner-Ad Bans.” Seems to be the latest in an attempt to kill off banner advertising. Generally, a bad idea and kind of like throwing the baby out with the bath water.
Display ad media (banners) are notoriously hard to measure especially if your strategy is an upper funnel one where clicks do not make sense as a KPI. Most banner ads seem to be about awareness or engagement in spite of that. Digital marketers should stick with proven ad research methods, e.g. brand lift studies or dig deeper into analytics using passive brand impact measurement like viewthrough. And of course there is always algorithmic attribution that can help.
A bigger threat to display media’s viability is ad blocking, which prevents display ads and any associated tracking tags from being served up. Yes, that means no potential for viewthrough (or clickthrough) if there are no ads presented to the user.
In addition to the reduced page clutter (what most people seem to want), ad blockers are also a good way to speed up your Web browsing experience. Unfortunately, too many publishers do not use advanced tag management to reduce the strain put on browsers when 25-50-100 3rd party cookies and tags fire in the browser.
According to eMarketer this is going to be a 25% de facto tax on publisher audiences. Some are however, asking users to shut off their ad blockers when visiting their site. Here is the official word from the IAB on ad blocking.
