00:07Siamo a Milano, agli Amazon Unbox dove Amazon Ads presenta le sue ultime innovazioni e oggi
00:12parliamo di un argomento particolare, parliamo di measurement e lo facciamo con Paula Despin,
00:17Vice President of Measurement in Amazon Ads. Welcome Paula and thank you for being with us.
00:22Thank you, thank you for having me. It's great to be in Milan.
00:26So Paula, I would like to ask you, measurement has become one of the most debated topics nowadays,
00:34why does it matter so much for you?
00:36Oh, for me measurement is the foundation of everything. I mean consumer journeys are so
00:42complex now, you think about it, consumers are everywhere, streaming, social, moving from there
00:49to e-commerce. You really need measurement to make sense of it all, without measurement
00:54you don't know what's happening. You're flying blind.
00:58Yeah, it's a way to understand more the consumer. We're gonna come back later but now I would like
01:05to move a little bit because one of the main products of Amazon is the Amazon Marketing Cloud.
01:12Yes.
01:12And I would like to ask you first, what is it and why is it important for advertisers?
01:19Well, I'm always happy to talk about Amazon Marketing Cloud or AMC. It's a privacy safe,
01:27clean room solution that enables incredibly flexible analytics. So it's a place where advertisers can
01:35bring their data together with Amazon signals and then really explore to understand the customer
01:42journey all in a very privacy safe way. It's always aggregated but you can query very flexibly.
01:48that gives you amazing insights into customer journeys.
01:52Okay. Okay. And staying on the AMC now is available for some time. What has changed recently
02:00on the AMC that makes it more accessible?
02:04Well, that's a really good question. I love writing code but not everybody loves to write code.
02:11We've launched some no code solutions where you can point and click to do an analysis but
02:17really that power is in the flexibility. So what we've been able to do with Generative AI
02:24is create an Amazon Ads Agent skill for AMC that writes the SQL code for you using natural language.
02:35So you just type in your question like, how many long-term sales have my campaigns produced in the
02:44last 30 days? And the Ads Agent skill will develop that SQL instantly for you and you just press click
02:53and then you can get the insight. That's pretty amazing. What we're really trying to do is make that rich
03:02insight available to everyone. Okay. You have already answered the next question but I'm gonna try to
03:09focus more on that because you introduced, like you said, the Ads Agent for the AMC. I would like
03:15to understand more. You have already spoken about it. How does it help the advertising in their daily work?
03:24I think it's quite important because historically you needed to have a data science team. That's quite
03:32difficult for smaller companies. Even if you have a data science team, there are just so many insights.
03:38So now being able to very, very quickly develop the queries to create an audience or do analytics or just
03:46like find out more about the products, just being able to type that in natural language is so fast
03:54and powerful. It's really a big unlock, I think, to getting more insights faster for everybody.
04:01I would like now to move on another announcement that you make is the MCP server. I would like to
04:07ask you,
04:08can you explain to us what is it and why does it matter to you?
04:13Yeah, I always make jokes. I mean, MCP, it's such a strange acronym, right? I was like,
04:18is it a regulatory body? Is it a band? Who knows what MCP is? It's actually, it's quite important in
04:25this era of generative AI. It stands for model context protocol, which again, that sounds very
04:31complicated, but it's just a pattern of communication, just like HTTP in a browser URL. So it just enables
04:40two sides to know how they're communicating. That enables other agentic systems to integrate with
04:50Amazon ads through the MCP server in natural language. We also added tools to these. Now,
04:57tools are kind of like, I was, I was joking before it's a TLDR or keep it short tools package
05:04up the
05:05logic. So it's even simpler for those agents. When they call our systems, they can learn like what a
05:11metric means. They can understand in natural language, the order of operations. So like when you start a
05:18campaign, first thing you have to say is which account am I doing it in? Okay. What's my budget?
05:24Having those capabilities packaged up in tools lets an agent at a partner system or an advertiser system,
05:32just call Amazon ad systems in natural language and get the right answer. Okay. Now we have seen
05:39these new tools on a global scale. Now, now I would like to focus more a little bit on Italy.
05:43I would
05:44like to ask you what are you presenting for the Italian advertiser? Yeah. So we made a very exciting
05:50announcement for Italy today. We released prime video insights in Amazon marketing cloud or AMC
05:57to enable really flexible analytics using those insights about content. And so now advertisers can
06:06connect the content with their product in really unique ways. So for example, let's say you sell food
06:14products. So you might have guessed that people purchase more when they watch cooking content.
06:20That's the pattern you might expect, but now you can put a number. Is it three times? Do they purchase
06:26three times more or four times more? Now you have some precision. You also can find patterns you didn't
06:33expect, like maybe people per that purchase your food brand more that watch documentaries versus romantic
06:40comedy. Those differences, those insights are very unique. This is the first time you can really see
06:47that. And it's going to change your audience, your creative, your whole marketing strategy.
06:52We opened this interview talking about measurements. I would like not to come back on measurements.
06:57Always. I'm always happy with measurements.
06:59The Unboxed Milan is showcasing many innovation for Amazon ads. What does this moment represent for the
07:05future of measurements in analytics? Yeah, it's a very exciting moment for measurement. I think you're
07:13seeing a convergence of more signals, more powerful analytics using generative AI and then simpler interfaces.
07:23Those things together really have the potential to unlock measurement in a time of frankly great complexity.
07:32So being able to simplify that down to the core insights, super powerful.
07:37And one last question. What is the broader vision for you? I mean, what would you say to advertisers
07:46that are still hesitant about investing more in advanced measurements?
07:52I mean, it can be intimidating. You know, how do you unlock all of these insights? And I think that's
07:59what we're so excited about with generative AI can really tear down those barriers so that anybody I famously
08:06say all the time, you know, it can be the CMO, but also your newest hire to gather the accessibility
08:13that
08:14you can get from generative AI is super important. And, you know, you could say, well, is it is it
08:21important?
08:22Do I have to do it? I think, yes, you do. Because if you're not doing it, the cost of
08:29not measuring
08:31is high. It might be a little bit invisible, but it's real because all those insights that you're missing,
08:39you may be over-investing in channels that look efficient, but really you're missing the big picture
08:47across all the ways in which you interact with customers. Measurement can help you distill that
08:52and then take action on it. It can be a more precise step further. Okay, same. Perfect. I think
08:59we're going to see what will happen next. And then thank you so much for being with us again, Paula.
09:04Awesome. Thank you.
Commenti