[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)

Pacvue

DSP 광고의 과거와 AMC의 등장 배경

과거 DSP 광고는 막대한 비용 대비 효과 측정이 어려웠음.

2010년 이전에는 광고 효과 측정에 대한 데이터가 부족했음.

AMC 소개 및 리프트 연구와의 비교

리프트 연구는 설문조사를 통해 소비자의 자가 보고 행동에 의존함.

AMC는 아마존의 익명 사용자 ID를 활용하여 실제 행동 데이터를 추적함.

AMC는 클릭, 장바구니 추가, 구매 등 실제 행동 데이터를 제공함.

패널 데이터가 아닌 실제 사용자 상호작용 데이터를 제공함.

AMC를 통한 고객 여정 추적 및 가정의 배제

AMC는 전체 고객 여정을 추적하는 수단을 제공함.

과거에는 여러 가정을 기반으로 광고 효과를 추정했음.

스트리밍 TV 광고의 경우, 시청률과 노출수만으로는 효과 측정이 어려웠음.

AMC를 통해 광고 노출과 프로그램 참여 간의 연관성을 정확하게 파악 가능함.

다양한 광고 유형에 적용 가능하며, 가정을 배제하고 정확한 데이터를 제공함.