Bosko Milekic 🌶

This week my friends @vstesin, @maiduong, christianerochon.com launched @swabtheworldnow and you can probably help. It's much easier than you might think! swabtheworld.com -- take a minute and have the privilege of giving someone their life back.

@aripap Don’t fall to its fawning compliments

I think Claude is impressed with the quality of my prompting

@EricTilbury_RTB If you mean matching in the DCR then can separate operations and separate outputs to split types of treatment into multiple deals or LIs depending on whether activation happening in SSP or DSP. If you mean how IDs targeted by DSPs are resolved to pub traffic that’s a diff story.

@bosko How is a probabilistic match of those IDs and a deterministic match of those IDs handled? How can you designate between the two at the line item targeting level?

I don’t live in a world where targeting IDs is a priority, but there’s something I’m trying to wrap my head around: How are clean room IDs governed in a DSP setup? What prevents you from understanding more about the user ID you served an ad to via the clean room set-up? Are there certain types of clean room IDs integrated with the DSP where you can’t access detailed information? To the clean room experts—this isn’t an attack, just trying to educate myself.

@EricTilbury_RTB Tldr Signals need to flow through the RTB pipes to make things work. No silver bullet way of avoiding that. The signals can be labels on groups of users (eg KVs/deals), IDs linkable to other user level data (eg UID2), or IDs decoupled from other user data (eg PAIR pub IDs)

@EricTilbury_RTB … the list of IDs/users coming out of the clean room need not correspond to matched users either, it can include other users based on eg modelling done in the cleanroom…

@EricTilbury_RTB And, with standards like PAIR, the IDs coming out of the clean room/match operation and going into ad requests to enable the activation via DSP are engineered in a way where parties outside of CR cannot tie them back to individual user data (like the email used to generate them)

@EricTilbury_RTB What comes out of the clean room to enable the targeting is often a bunch of IDs. These then go either into 1. Pub side system where they are targeted to a deal/KV and then addressed at a cohort level from DSP, or

@EricTilbury_RTB … 2. DSP where they are targeted with the expectation that the DSP will see corresponding IDs in the bidreq stream and/or be able to link uploaded IDs to other IDs it sees in the bidreq stream…

@mikeosullivan 😅😂🤣

proving to my wife that we have a vibrant marketplace of tasks and responsibilities in our marriage

@GalebNikacevic @njuznet Ovo je bilo odlično 😅😅

U novoj septembarskoj epizodi družili smo se sa dragim prijateljima iz @njuznet i to u punom sastavu. Sva sreća imamo dovoljno mikrofona, ali ne baš i dovoljno mesta. youtu.be/CB8qKBKHKdE?si…

@gyardley Maybe Nxc7+ Kd8 Qf8+ Kxc7 Bf4+ Kb6 Be3+ Ka5 Qc5+ Ka4 Nc3# ? Though after Be3+ black Kc7 possible and could lead to stalemate due to repetition. What am I missing?

Absolutely insane blitz game today. Played a Scotch, black did the 4...Qh5 thing and I played 5 Nb5, hitting the f7 square. After 5...Qxe4+ 6 Be2 he drops a piece with 6...Be6? I surprise him with 7 Qxe6!, but then he plays 7...Qxg2? Now, find the mate. 😅 #chess #chesspunks

@ericfranchi @eric_seufert Agreed. I am guessing the acq includes the Inscape biz. Will be interesting to see if they close that up now.

Odds of Walmart's acquisition of Vizio passing regulatory scrutiny seem good. Walmart doesn't have operating entities in EU (h/t @eric_seufert for doing the digging here). Amazon has FireTV and the largest commerce media business, so W + V = pro-competition. Thoughts?

@alextcone @aripap @ericfranchi Insightful Q&A up here ☝️

Thanks for having me on @aripap and @ericfranchi! It's not every day one gets to be a guest on a favorite podcast. Keep up the good work providing the ads ecosystem with deeper understanding (and good humor). I hope listeners walk away with some new insights into Privacy Sandbox.

@shails If RTB was the newborn, what were those of us doing targeting, measurement, and optimisation in digital advertising pre-RTB doing? There was digital advertising before ad exchanges, before data got decoupled from media and 3rd party data marketplaces proliferated.

That was a newborn absorbing, learning everything, forming and growing. Now it’s a grown up with some fully formed ideas and needs

Good point

My colleague @vstesin delivering the state of data collaboration and privacy ad tech in a packed room of Canadian collaborators. Proud to report that the 🇨🇦 ecosystem is at the forefront of positive change in digital advertising. Thanks to our friends at @Deloitte for hosting us!

Apparently, the Privacy Sandbox APIs are “really complicated” and require some rethinking about how things are implemented. In contrast, in 2008 rewiring the ecosystem for RTB was a walk in the park. Ad tech gonna ad tech… 🤔😅🤣

@alextcone 👋

Made it to Vegas #CES2024

@AdtechGod I am guessing your question pertains to what is more important wrt CTV advertising — it’s the OS and software. Whoever controls that controls the platform and service layer, including the AVOD/FAST programming and app distribution. Platforms set the rules, with few exceptions.

@AdtechGod TV models at $500 vs $2500 from the same manufacturer have a big diff in terms of the hardware inside, memory etc. Margins are thin in the CE industry and extreme optimization is required resulting in big perf variance. This in turn makes a consistent software xp difficult.

Does the smart TV or device really matter? Or is it the OS that matters? Is it a battle for Device dominance or OS dominance?

Inside the House of Unskippable Podcast Ads 😅