[Music] So the dark ages of DSP were like, here was your sixty-thousand-dollar spend and

here's the five million impressions you got. Right? That's the story. There's more

than that. No, I mean, way long—ten years ago, right? Right, like 2010, that's really

what it was, right? And uh, before AMC, the conversation was a lot more interesting. We just broke down those metrics that actually, they do have

a story. We didn't talk about lift tests, but I think lift tests are hard to do,

right? And now we have AMC. So please, like, introduce AMC for people who don't

know what it is and what what's the value. Then sure. So let's let's maybe

compare it to a lift study as a nice photo comparison. So when you're doing

a lift study, you survey a bunch of people and you said, 'I know that you saw this ad, did it do anything for you? Did you learn

about my brand? Did you go buy something?' Tell me, based on what your survey results, what you did. AMC does the

opposite, right? So AMC says, 'I know that I showed this ad to this person's unique

user ID, which is Amazon's like anonymous identifier of the person that you show that to. Then I can actually see, based on

being shown that ad, what behavior did they take? So, not what was their self-reported behavior, what did they

actually do? Did I click on a different ad? Did I click on five ads in a row? Did I add something to the cart and then

abandon it? Did I actually buy products? Which products did I buy? Um, but I'm giving you the ecosystem to

Say, rather than panel data, get the actuals of how people actually interacted with my ads and what downstream effect that had.

Yeah. I mean, AMC, in a very quick, short summary, is just a means to track the entire customer journey. We always

had to make assumptions before. Um, and we did. Many assumptions were made, but you never had the ability to say,

like, "Okay, right now I'm running this display video campaign, maybe it's OTT or

streaming TV." Right? And with that, you mostly just get like video completion rate and impressions, so it's really

hard to justify this type of advertising. But we've always made the

assumption that if you run a streaming TV ad, those impressions—those people know your brand; they'll go

interact with other parts of your program. Assumptions. Now we can truly tie those views and those impressions back to the

rest of your program. And this is, you know, replicable. We can replicate those across a variety of different ad types.

It's not just with video, but video is a great example. And so it's really just a

means to take out the assumptions and give you accurate data. [Music]

A Quick Introduction to Amazon Marketing Cloud (AMC)