Sizmek declared bankruptcy at the stroke of midnight on Friday. Come Monday, will any of their impressions still be showing? They owed $65 million - is that just delayed payments? Will payments flow to keep ads on? This feels like the ad bubble - who else has to close very soon?
When you look at the creditor balance sheet for Sizmek it should scare many folks watching markets- you could probably pick any of those companies and find similar books on their ends. They are all doing shades of auction arbitrage w/ cookie changes in browsers gutting data flow.
The subprime crisis occurred bc too many banks bet on each other and layered risk into assets that stopped being revenue multipliers and just started failing. Cookie consent rates w/ safari/Firefox changes (&chrome coming) are failures crossing that ~7% threshold like subprime...
Ppl need to look one step deeper into Sizmek & recognize that they *lurched* into bankruptcy w/ $65 million in open books. That’s a lot of payments owed to identical peers - and it also means less or no more Sizmek impressions and payments. The foundation of the $pyramid is 📉
How many ad networks were seeing diminishing returns and counting on the few big vendors like Sizmek to ensure they could fulfill ad fill orders and keep their business models. It’s a piggyback soup out there and most folks don’t realize how many pixels piggyback via Sizmek.
I’m going to share some screen shots of Sizmek javascript deployments that should truly scare people who understand javascript piggybacking and what it means if Sizmek goes under and stops firing other people’s pixels ......
Here's Sizmek's own website -- you'll notice that nearly all the pixels fire from Sizmek's own "Sizmek Tracking" pixel -- but" it's actually the SUPER OLD legacy RocketFuel pixel - this one @ c1.rfihub.net/js/tc.min.js // Notice it's insecure - great start.
Wanna know how you can tell the Safari 3rd party cookie blocking totally gutted Sizmek? Sizmek updated their legacy RocketFuel pixel with *new* details to try and get consent on Safari.

Screen shot receipts:
There are clients still using Sizmek's super legacy insecure RocketFuel javascript -- there have been huge declines (bc obviously normal websites want to deploy an SSL and can't fire insecure javascript..) - There are still 41,000+ U.S. websites that have RocketFuel JS... 🤯
I suspect someone would be able to analyze a bunch of Sizmek client accounts and compare the data flow to the debtors and see which of those companies *mostly* piggybacks data via Sizmek. Piggyback javascript for ad networks is the subprime crisis of the internet.
I'm gonna say this again because it's important: Piggyback javascript for ad networks is the subprime housing crisis of the internet. If 1 ad network is piggybacking javascript for lots of other networks it creates the potential for waterfall failures if a couple ad nets go down.
I almost feel bad for some of the Sizmek sites i'm analyzing. They obviously have user data slipping through their fingers, revenue is ∂ropping & they are adding more & more javascript & ad trackers to increase revenue. That slows down pages, sucks out value, and then what?
Some sites with Sizmek seem to almost only use it for conversion tracking - the data flow is quite sparse to/from Sizmek directly, when compared to the other big players. But again.. you see the debtors in these screen shots. You see their relationships.
But then you see some sites with Sizmek and they are *all in* with Sizmek Tracking -- it's pretty safe to say that @krispykreme is a Sizmek data business. Like in a super serious way. What will their marketing team due on Monday morning? Uhh folks, we gotta switch all our ads??
And even software vendors with good reputations like @Mimecast have gone all-in with Sizmek // You can really see the data flow to their partners on these types of installs. This is subprime javascript folks - sorry to break it to ya'll:
And this also impacts nonprofits - like Care.org - a huge Sizmek partner who fires most of their pixels via Sizmek piggybacking. If Sizmek stops operating, many companies will *lose access to publisher websites & data* and be forced to quickly pivot or...?
Just imagine you've been using Sizmek for years- a company like @Burlington coat probably has years of data there. AND they are piggybacking countless pixels for other organizations. So if Sizmek goes, all the ad networks in the screen shot lose access to Burlington coat data. :(
I don't mean to pile on the bad news for Sizmek, but this is *very scary news* for the online advertising industry. As someone who spends a ridiculous amount of time looking at the javascript publishers fire, & the javascript waterfalls, Sizmek is the Bear Stearns of javascript.
Maybe i'm being alarmist, but javascript piggybacking hasn't only been a user security/privacy problem, & it's not only increased the speed of data transfer worldwide as javascript piggybacking became popular, "ok" & widely done... It's connected businesses together..tied debt..
To wrap up:
1) Sizmek bankruptcy is scary
2) Javascript piggybacking is the Subprime Lending Crisis of the internet
3) JS Piggybacking intertwines businesses like mortgage backed securities
4) Dog shit wrapped in dog shit wrapped in a JS wrapper = current state of ad networks
With the unfortunate news that @dataxu is laying off 30 employees (adweek.com/programmatic/d…) it's a good reminder we *did not* see this chaos in the ad markets around GDPR - this new chaos is coming on the heels of Safari + FireFox blocking 3rd party cookies - it hurts revenue..
I know a lot of folks in the advertising market have a lot of perspectives, & may not always see the forest for the trees, but access to audience data in the ads industry *relies on business partnerships* for most companies + *relies on 3rd party cookie drop conversion rates*
it only takes analyzing a website like NBA.com to see overlapping businesses, duplicate pixels for the same ad networks triggered by different vendors, numerous piggybacking javascript & JS errors... JS errors everywhere..
You can see in these screen shots to and from highlighted nodes -- you can also see this in network logs obviously but this is a helpful way to see unique vendors triggering unique versions of the same pixel. Multiple vendors firing their own unique DoubleClick accounts... :/
And in most of these javascript advertising waterfalls, the more respectable the company (google in the screen shot below) the less data partners they are working with -- the data to and from google is always light since they only take on limited data partners
Amazon is a late player to the advertising game, but you can see who they quickly partnered with in the screen shots below. You can see they are already playing "The game" so they can share data with ad networks that have external publisher networks outside amazon / google/fb
This is what it looks like when Adobe Audience Manager is angry and vomits javascript errors on your website. I've got no clue what's going on there but you can see their partnerships via JS errors which is still nice.
No surprise that Pubmatic has some solid deals, and a few unique companies they work with
DataXu is another big player in the javascript piggybacking reverse mortgage infrastructure of the internet - you can see they too have many partnerships but also a few unique ones. With news they are laying off 30 employees and "focusing on TV" .. are their pixel partners too?
Krux Digital is a huge player and helps to distribute a lot of unique pixels as well as many of the duplicate ones that fired via other parent pixel orgs // LiveRamp is similar w/ big partnerships and a few unique
Here's all the requests to DoubleClick from any vendors or pixels from NBA.com - there are multiple DC pixels here from vendors and you can kinda make that out via the roundabout trips and unique one-ways to DC
Here's just the Dataxu requests on NBA.com - you can see that many companies rely on each other and when a company like Dataxu lays off 30 people on the heels of Sizmek filing bankruptcy, this should alarm data partners across the data supply chain
Here are the Krux Digital-only requests for NBA.com - you can see Sizmek tracking up in the top right there too. If you don't understand the data sharing and audience overlaps (and business logic overlaps) from this I don't know what to say. It's compounded risk.
I like this one because Snapchat.com is just sucking up audience data from various vendors who piggybacked their data to the snapchat JS on NBA.com -- but notice there's no direct connections between SnapChat & DoubleClick. Just roundabout data flow..
Rubicon is interesting because it's a bit sparse with data flow on NBA.com - but it's also literally the first thing in the source code on their website for prefetch requests. There are a few duplicate Rubicon account/requests here too.
I like the Pubmatic pixel because it's fired from numerous other vendors -- it's also piggybacking pixels and a lone pixel/company called "Netming" which claims to have Bentley, Verizon and Dre Beats as clients due to their audience data. Piggybacked audiences for sale anyone?
I want to wrap up this original Sizmek thread now extended with some Dataxu analysis... it's important for everyone to understand that the advertising industry can get caught up by greed & complexity to build a house of cards just like any industry. It's happened before...
If you take $1 (hypothetical value of a user visiting a website) & then share that value w/ 10 companies, who then share it with 50 more of their partners, & then everyone gets similar user/request data + sells similar data they try to maximize... it's still $1 worth of data...
Many folks in the online advertising industry see themselves being part of a JS architecture that takes the $1 user value and then inflates that by 10x then some fraction of 50x and don't see a bubble in the overlapping value/sales of the same person across multiple companies.
Just because folks in the advertising industry are able to convince companies to buy their audiences and trust them with their advertising money, doesn't mean those companies aren't selling the same data that could be bought from dozens of other orgs -- the same users / targeting
The one thing that's bothered me is that I've been trying to figure out the threshold for a collapse in the advertising industry -- I don't think it's based on like 8% of companies closing or something. I think it's the cookie consent rate. If Chrome blocks 3rd party🍪 it's over.
imagine if the subprime crisis could have been stopped by preventing some regulation or some action from going into effect. Honestly, can *any* of the companies listed in the screen shots above w/o direct pixel access to NBA.com survive Chrome blocking cookies?
There's a lot of smoke / mirrors in the advertising industry + everyone is on a sinking boat hoping to sell before cookie-apocalypse so no one is talking about how much Safari/ Firefox cookie blocks hurt, and how bad it's going to be (for them) w Chrome changes.
The chrome cookie block is similar to the interest rate changes that pushed the housing crisis into a global recession - an inevitability. Too many people have built too many orgs on top of 3rd party cookies (& built w/ debt), and publishers houses are over-leveraged w/ pixels.
Publishers can't pay back their page visitor debt w/ more pixels & more 3rd party cookie data transfer. Publishers must stop deploying more & more pixels, & instead start to "short" their pixels to offset their position w/ 1st party data & more prompts for users to login.. 🤖🍪🖖
Missing some Tweet in this thread?
You can try to force a refresh.

Like this thread? Get email updates or save it to PDF!

Subscribe to ℨ𝔞𝔠𝔥 𝔈𝔡𝔴𝔞𝔯𝔡𝔰
Profile picture

Get real-time email alerts when new unrolls are available from this author!

This content may be removed anytime!

Twitter may remove this content at anytime, convert it as a PDF, save and print for later use!

Try unrolling a thread yourself!

how to unroll video

1) Follow Thread Reader App on Twitter so you can easily mention us!

2) Go to a Twitter thread (series of Tweets by the same owner) and mention us with a keyword "unroll" @threadreaderapp unroll

You can practice here first or read more on our help page!

Follow Us on Twitter!

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just three indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3.00/month or $30.00/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!