I wonder how Google would handle this attribution for YouTube. I was looking for stickers I ended up on stickermule I'm watching YouTube TV I get an ad about stickermule for t-shirt design. I'm going to likely buy this sticker pack this week. - ThreadSky

wilreynolds.bsky.social • 8 days ago

I wonder how Google would handle this attribution for YouTube.

I was looking for stickers
I ended up on stickermule
I'm watching YouTube TV
I get an ad about stickermule for t-shirt design.
I'm going to likely buy this sticker pack this week.

Comments

ratocaster.bsky.social•8 days ago

There’s an old Bob Hoffman post I can’t find right now, where he talks about his reading habits of The NY Times. How he reads on Sundays and a few other days, but never Monday/Tuesday. And marketing analytics would attempt to come up with all these reasons for why, and they’d be wrong.

ratocaster.bsky.social•8 days ago

It had to do with him preferring the harder crossword.

Point being, consumer behavior is erratic, emotional, unpredictable.

Web analytics are just accumulations of things that happened. Attribution is always BS because it’s attempting to say “why” at scale. We don’t know why at scale.

ratocaster.bsky.social•8 days ago

Also, and lastly, this.

https://bsky.app/profile/jessicacalarco.bsky.social/post/3lirjqzud222y

ratocaster.bsky.social•8 days ago

I lied. The real last thing is I try to avoid answering those questions. We try to make larger changes to ads in somewhat isolated time periods. But my preference is we measure ad dollars against total revenue. Attempting to attribute return by channel always involves lying.

chris-green.net•8 days ago

If they’re talking to advertisers they’re obviously taking FULL credit for that

brendanhufford.bsky.social•8 days ago

I got everything from them for years until their CEO used to their entire marketing list to start raising money for Trump.

And the proceeded to have multiple very public meltdowns.

wilreynolds.bsky.social•8 days ago

Oh that was that guy???

brendanhufford.bsky.social•8 days ago

Oh yeah - put a 100ft wide sign in their hq, too.

philmarz.bsky.social•7 days ago

My view has always been that attribution is A tool not THE tool. Too often I hear about attribution modeling being used as a single source of truth to determine what does and doesn't work with a marketing and advertising campaign... That's ALWAYS been a lie and guided many brands to poor decisions

philmarz.bsky.social•7 days ago

For starters, attribution isn't just 1 thing or 1 number like people like to talk about it - there are MANY different attribution models, each meant to show you a different part of the sales funnel. First click for example weighs more balance to your top of funnel vs last click to bottom of funnel

philmarz.bsky.social•7 days ago

The goal of attribution modeling should be to use MANY DIFFERENT models to get a better understanding of which actions are working across your WHOLE strategy. I've seen too many marketers just use a last click model, only caring about what made them immediate money.

philmarz.bsky.social•7 days ago

The issue with that logic is that email marketing will almost always be 40% - 60% of your sales for a healthy marketing strategy... BUT HOW DID YOU GET THOSE EMAIL ADDRESSES??? You need to use multiple models to properly determine what DOES and what DOES NOT work.

philmarz.bsky.social•7 days ago

This is... In my opinion... Why GA4 is SO MUCH WORSE than UA used to be. With the update to GA4, Google REMOVED all attribution models except for their own "Data Driven" model. WTF does that even mean???? A mysterious algorithm that gives no visibility is going to magically tell me what works? HOW?

philmarz.bsky.social•7 days ago

As soon as Google made that update it became more important than ever to look for 3rd party tracking and analytics tools. And there are many great ones that exist without needing to use cookies - but if you ever see tools try to advertise a single attribution model, know it's BS!

joelhaffey.bsky.social•8 days ago

Purebuttons may be an apolitical alternative: https://www.purebuttons.com/order-promotional-products/custom-stickers

wilreynolds.bsky.social•8 days ago

The ad has nothing to do with my conversion
And the ad wasn't even about stickers.

How do you folks in paid test for how prevalent this kind of thing shows up as "converting" view through ads.

randfish.bsky.social•8 days ago

Exactly what I've been talking about the last few years!

wilreynolds.bsky.social•8 days ago

Dude yeah it's that pizza and fliers example you've been using forever, I finally got aive example of that. Now I wanna figure out how to test for the "attribution".

randfish.bsky.social•8 days ago

My position is attribution was always a lie. Less so in 2014 before ad blockers, privacy laws, Apple's cookie decisions, keyword not provided, etc.

But definitely today it's a fool's errand.

The way forward is measurement based on incremental sales, not attribution.

roysteves.bsky.social•8 days ago

Yeah, my career was really built on attribution up to 2015. I loved it.

Now? Incrementality testing and retail. Still fun to be had there!

samuellavoie.bsky.social•8 days ago

With all these ads, they must undoubtedly try to recoup losses from their CEO’s political stance.

Comments

Posting Rules

Reply