[Music] [Arthur Lacroix] All right. Good morning, everyone. Welcome to the session. Thank you for joining us, as early as 8 AM.

So Frederic and I are really delighted to have this session today with you. And we are going to talk about how Generative AI can boost your creativity and productivity. So I am Arthur Lacroix, Senior Product Marketing Manager at Adobe. I've been at Adobe for 10 years, and I started on the campaign product with the client console, which was desktop based. And I have seen this client console on the desktop, go from desktop to the web. That was amazing. And then I have seen the custom code that we were doing on the product to connect to third-party product go from custom code to out-of-the-box connector, and now we are infusing Generative AI on our customer journey product. So that's truly amazing and that's what we want to talk to you about today. I'm thrilled to be joined with Frederic Mary. Frederic, a little bit about yourself. [Frederic Mary] Yeah, sure. Hello, everyone. So I'm Frederic. I'm based in Paris. I also have a similar background as Arthur. I come from the campaign acquisition, some time back, right? And we built... I was owning content, and we built the Editor, the authoring experience for email and the other channels. And overtime, we reinforced the features there. And I'm really glad today to be here to talk about some of the latest features with GenAI powered capabilities.

And so, yes, thank you very much for being here. All right, we're going to start with a few questions. Who knows what event is happening in Paris, in July this year? All right. So it's not the rally. It's not the color run. It's the Olympic Games. So we choose this theme with red because we are both from France, and we built an analogy between the games and GenerativeAI. All right, so who in the gents are using a Smartwatch to track their run, track when they are playing tennis, or at the gym? Raise your hand.

All right. That's the view as expected. And... So nowadays, Smartwatch are using AI to estimate a lot of things, health rates, heart rate, variability, blood pressures, the glucose level, and all of that to gather data in order to help you refine your training. So AI is everywhere. It's on your Smartwatch. It's in sports, as we will see, but it's also in marketing, as Fred will demo today. So today, we are going to focus on how Adobe's AI Assistance will help you on your day-to-day work.

Before we start, I have to go through this disclaimer here. But I want to tell you that every AI assistant features that, we are going to show you, and Fred has a great demo on that, they are all features that are in Beta.

All right. So let's again begin, right? So think about this session as a sprint towards innovation and join us on our very own Adobe Journey Olympics. Today, for our journey, we will identify key challenges athletes and marketers are facing. Fred here will be our coach, our torchbearer, as we navigate through Generative AI features for the customer journey management platform. All right, so let's begin the Adobe Journey Olympics.

Just as athletes, they are training to push the limits of performance in Olympics. They want to break records. They're really pushing the limits. Generative AI is doing the same. It's pushing the boundaries of what's possible in the realm of creativity, innovation, and productivity.

As athletes, they always want to break a record in speed, strength, agility. And similarly, AI algorithm is constantly evolving to push the boundaries of what is possible in arts, music, literature, in marketing, as we will see. But speaking of music, four years ago at the Tokyo Olympic Games, just as an athlete will set a new record, AI produced a groundbreaking work of art, creating an innovative music called "Make The Beat!". So the AI played a pivotal role, acting as a maestro behind the scene to orchestrate a symphony of music with more than 1,000 samples and 5 rhythm to evoke emotion, inspire athletes, inspire the audience, and resonates with a worldwide audience. So Generative AI, as you can see is reshaping the way we think about creativity.

But it's also reshaping every aspect of marketing, from planning to execution and analysis. I personally think there is three focus area for marketers as it relates to Generative AI. One, is the increase in productivity to innovate faster, especially with content. And demand for content continues to explode against pressure budgets.

With the use of Generative AI, as we'll see, you can start now as a marketer, create content at the speed of your imagination with prompts.

The second main focus to me is around upscale. There is a need to reduce the gap, the skill gap for the practitioner in the product. We need to supercharge the practitioner through AI Assistant in our product.

And an integrated AI assistants in the user interface will allow the practitioner to perform tasks faster, providing anyone in the organization with the ability to perform their tasks faster, to increase their performance and at the end of the day, reduce the cost.

The third focus area to me as a marketer is that you must assure that Generative AI data usage is safe to use, is also consistent with your brand image, and that's really important, and we will talk about that at length. It's safe for commercial use, it's ethical, it's high quality...

Because it's one thing to say that you can now create campaign and journeys very fast, but it's another to be able to be confident on using all the data generated by AI.

All right, so any judo practitioner in the room? Not a lot. Oh, here we go, one. So it's been a while for me, honestly. I did not make it to black belt, unfortunately, so I won't do a demo. But just as judoka will maneuver his opponent's energy to achieve victory on the mat, harnessing the capability of Generative AI for a marketer.

He will be able to excel in the digital arena to achieve victory against his competitor. So with Fred here, our coach, our torchbearer, we will learn how to boost our creativity edge as marketer, how to turbo-charge content velocity, how to empower personalization at scale, and also, last but not least and very important, measure and optimize content. All right. So, everyone, tighten your judo belts. Let's go through each of these points one by one.

Boost your creativity edge. I don't know if you know that, but break dance is being introduced in the Olympic game this year, for the first time. And break dancer, they must create unique routine to stand out in front of judges, in front of the audience. It requires creativity in choreography, costume design...

And overall performance. Thinking outside the box is critical for a break dancer to be able to have a competitive edge. Similarly, marketers need to create engaging content. It needs to resonate with the customers. It needs to stand out against the competition. So here, both break dancers and marketers must reinvent themselves for every experience they deliver.

So Coach Fred, can you enlighten us on how you can help marketers here in the room boost their creativity edge? Yeah. Sure, Arthur. So let's talk about how the AI Assistant is going to help marketers get an edge over competition when it comes to creativity.

So the AI Assistant is really infusing GenAI powered capabilities for all the channel content creation experiences across our customer journey management products. So think of building an email, a push notification, an in-app channel, an SMS channel. When you're building it, what the AI Assistant is going to help you to do is to ideate, iterate, refine with great levels of control. So it's truly acting as a brainstorming buddy that you can work with, and it's helping you spark some creativity to ultimately create more engaging experience, more impactful experience for your consumers.

It will act on your existing content that you've already worked on for your campaign, or it can also create new content based on prompts.

And all of this creativity is done by keeping in mind the brand consistency. So how you can make sure that your content is on brand, that you're using your own template with your own styles, and that the tone of the text and the content is actually compatible with your brand.

So creativity is really important, right? So thank you. But you also need velocity, especially in the Olympics, to win the gold medal.

So if we take the example of cycling here, we see that a lot of modernization has happened in the cycling area. From the helmet to the bike, the shape and the weight have been improved to reach levels of velocity that were, thought, impossible before. I was not cycling myself. I live in Boston, and I'm just using the blue bike and it's very heavy and old. And when I met my partner a few years ago, she was with this beautiful shaped bike and very, very light, and that really was astonished to me and changed my approach to cycling. So this is really, really important. And similarly, marketers need to modernize their tech stack, as a cycling, will need to modernize his bike and his helmet. So marketer needs to modernize his tech stack in order to execute content faster, to get the latest and greatest experience to their customer out of the door using multiple channel in different geo and for different markets.

So velocity is really important here. So, Coach Fred, can you tell us how you can augment marketers' velocity? So as Arthur mentioned, the battle for attention, the customer attention is fierce and content is a bottleneck, right? So the AI Assistant can help with this problem.

What we are doing with the AI Assistant is that the cards have been reshuffled, right? So things that were not possible before are now becoming possible with this type of technology. So we see a tremendous opportunity to get substantial productivity gains and we will see that through different features that we will demo today, whether it's about creating a full email from scratch through generation, whether it's leveraging brand assets to refine the generation, so existing brand approved assets, whether it's refining each individual content element to have, like, great level of control while we work, and finally, keeping really all this in mind in context of the different channels we have. It's not only about email. It's about every channel that a marketer is operating on.

And ultimately, what this allows you to do is to reduce the reliance you have on creative teams that are resource constrained, right? So the goal for us is to empower marketers based on their knowledge, on what the campaign is about, what content is there, what data is there, empower them to make an impact, get to the next level when they build the content.

So it's great to get content faster.

But now, Coach, we need to personalize the content, right? And we need to do, to be able to do that, the personalization at scale.

To prepare to the Olympics, coaches need to tailor the training of their athletes based on the body type, based on the strengths, weakness. They need the personalized training.

But coaches, typically they have like three to five athletes, but marketers, you do have millions of customers to take care of.

So similarly, marketers have the need to deliver personalized experience at scale.

But good Fred, how can you help marketer deliver this kind of scale with highly personalization content when they need to take care about millions of records, millions of profiles, millions of users? That's a great question, Arthur. So first, with AI Assistant, what we are doing is that, the cost to produce content is really being reduced. The speed to create content is much faster, right? So this really unlocks an opportunity for us as marketers to tailor the experience down to the individual level. Something that was not possible before because it required building a lot of content, which was really resource consuming, is now becoming possible.

So when you think about content generation and you combine it with other features that we have in customer journey management solution, such as the ability to create dynamic content, like conditional content for different segments, and think of it that each content for those segments could be generated. Or even if we go to the next level, say it's not segment level, really individual level, think about offer decisioning capabilities. We can have a catalog of offers that are boosted with generated content and that can really deliver the next best offer at an individual level also based on AI algorithm that will find the best offer for each individual. So when you combine this set of features, it becomes a very powerful tool in the hands of the marketer to take the experience to the next level and really resonate with our consumers.

It's awesome, Fred. Thank you. Thank you for that, right? So be able to be more creative, create content faster, achieve personalization at scale. But I think we are missing one key element here that we need to address, the ability to measure performance. So at the beginning, I've asked, like, who was wearing a watch tracker? Athletes are using trackers, a lot of trackers to measure heartbeats, performance, collect all this data to learn from it and test different approach to optimize their training and optimize their performances.

Similarly, marketers need to be data driven to develop the winning strategy and win the gold medal. So, Coach Fred, do you also have a solution for that? Yeah. Yeah, yeah, sure, Arthur. So as we mentioned, we talked a lot about generating content, ideate, and so on. But at the end of the day, why are we doing all that, right? It's to drive better engagement, make an impact with your consumer, drive revenue, right? And so it's one thing to believe that the content that we created is good. It's another to have data to back it up, right? And so what we are doing with the AI Assistant is that we connected it with our experimentation framework that you can find in Adobe Journey Optimizer. So what it means is that when you generate an alternate content, you have two choices. One is, you can, you believe it's better than your baseline content and you can replace it, right? But the second choice is, maybe it's a good option, right? But you want to test it.

You want to compare it with your baseline, right? So we have created a shortcut that just allows to add that generated content as a treatment for an experiment. So it is going to be very easy to test your content, deliver that. You will be able to choose a certain part of the population to test the generated content and really measure the lift created by that content over your baseline content. So with that, there's no guesswork, right? We will have data that will turn into insights that will inform your content strategy.

All right. So... Awesome slides. Thank you, Fred, for that. But I think now, you just need to roll up your sleeve and show us in action what it means. - Yeah. Sure. - Are you up for the challenge? - Yeah. Sure. - All right. Let's go to the demo now.

Okay. So for this demo, what we're going to do is, I am a marketer working at WKND Fly. It's a brand, the company selling flights. And in this context, let's say what I want to do is, I want to promote flights for the upcoming Olympic Games in Paris, and I also have a secondary goal. My secondary goal is to also promote my WKND Fly Club program. So it's like a fidelity program...

So that you can see here. And I want like, to promote that so that more people can get exclusive benefits by joining this program.

So what I will do is, I will start here with the email designer, right? So I'm tasked to create an email to promote that. And what I will do is, I have access to a set of brand templates, right? So WKND Fly has this template. It has its own WKND Fly styles, but it doesn't have actual content, right? It's like Lorem ipsum text or placeholder images. So at this point, what I will do is, I will just use this template. We are inside the Email Designer. There's a lot of features here around templating, fragments, reuse content, and so on. But what I'm going to do here is, I have this AI Assistant on the right rail.

So I'm going to pull that up, and I'm going to put a prompt in there, right? So if we see the prompt quickly, we see that we have, as a marketer working at WKND Fly, I want to promote flights for the upcoming Olympic games. I also want to promote the WKND Fly Club program.

So I want to create a full email. I want a subject line. I want a pre-header to be generated. I can choose a type of communication. So let's talk about exclusivity here. I can choose a tone. So let's say I want an inspirational tone. And I also want some images to be generated. So let's say, photo type. At this point, I can click on Generate and the AI Assistant is going to give me options. Right. So let's go through these options. So I have two alternates here. And what we can see here is that we have a full email generated. You see on the top, I have a subject line. I have a pre-header. And when we go through the email, it used the template that I selected, but it also created cohesive content. If you look here, you have an image, you have a title, you have a paragraph, and a call to action.

And it all makes sense together. This section is promoting flights to Paris with a book now, Call to Action, but we also have other sections. So we have here a section specifically around the WKND Fly Club program. And here again, it's a very cohesive content block that has been created. And this is really the power of the AI Assistant.

So we can also explore the alternate option that we have here. But the idea, what I like here is really the first one. So I'm just going to select that one.

So what do you think about that, Arthur? Well, I'm impressed. That was very fast. I think it's going to save marketers a lot of time, so truly, truly amazing. And I guess you can never run out of ideas, right? You can generate, you can generate it very, very fast with a lot of design. So that gives you, like, a good head-start in the game against your competitors, right? But this content is great, but I need it to be tailored to my brand, right? I have, like, specific guidelines. I have imagery, vocabulary. Do you have a solution for that? - Yeah. - Sure, Arthur. So in this case, what I'm doing is I have a brochure at my disposition, right? So it's a PDF document that's talking about this marketing operation for the Olympic Games. It has brand approved content, imagery, text, and so on, right? So it's really talking about a lot of this, detailed information that I have available, even like travel tips. If I come to Paris, how can I move around and any tips like that, right? So let's say I want to use this document, right? So you have here a section called Brand Assets. I just uploaded this brochure. And what I will do is I will generate this again. So what's happening here is that the AI Assistant is combining generation with also pulling content from the document. So it understands semantically what information is in there and is going to do a combination of those. So here, again, I have two options. And this time, when I start looking at the content, I see it is becoming much more prescriptive, like it's more aligned with my initial goal, right? I wanted to promote the Olympic Games. I think it's inspirational communication that I want to do. And so over talking about just flights to Paris, I think, like, witnessing history with an Olympic asset like this is something that is resonating more with my consumers. It also kept the different content blocks and it also pulled more precise information. So I mentioned we had some travel tips, right, in the document, the PDF document was able to pull that.

So that's something that is done because also we can really like fine tune the prompt. So if in the prompt, you say, pick up a section, create a section around travel tips, it's something that it will be able to do.

So at this point, I have two alternatives.

And again, I will go with this first one. I like this one better.

And, yeah, I'm sure that's how I leverage brand assets. That's amazing, right? I have my content generated in a matter of minutes. My content is now based on my brand image.

So as a marketer, I'm really happy with the outcome, right? But I also want to... I feel I want to be more in control, right, and controls the outcome of all the text content, and etcetera. Do you... Can you give me a solution for that, Fred? Yeah. Sure. Sure. So we are talking about full email generation, right? Like, this shows a lot of capabilities. But at the end of the day, maybe I like a certain section. Maybe I like some part of the content, but I need to remain in control. I need to be able to edit each content section individually. So in this case, for example, let's say I want to edit the subject line, right? That's like the most important aspect of my email, the only thing that customers will see before opening it. And I want to refine it. So I can just go on the subject line, again open the AI Assistant, and here I have my original subject line. And I can say, okay, let's change the tone. I want something a bit more exciting. I also want to include emojis, and then generate again. So I can iterate like this many times, but you see now we have a text generation assistant, right? And so here, let's say I like the Variation 3 better. It's really like appealing to me. And so I can just select that for the subject line.

Amazing again. And text is, text generation is, is great. But as a marketer, I'm really struggling with my designer to get image very quickly, right? I need to wait for four days. I'm sure a lot of you can relate. So did you have a solution to generate image as well, like text? Yeah. Yeah. Sure. Sure. I mentioned that we can generate different elements, right? So I talked about subject line, but let's just say I want to change an image here, right? Actually, let's say I don't just want to change an image. I will go a little bit further. I want to bring personalization in the mix, right? So here I have an action to say enable conditional content. So these are existing features in our Email Designer, and I can start adding variations, right? So here I added two, but I can add as many variations as we want, right? And for this variation, what I'm going to do is, I just go pull the AI Assistant on the right. I'll type a prompt. So I have this content that I find appealing to a young audience, but I also want to create another content more for a VIP traveler. So I have a VIP traveler sitting comfortably in business class, and I want to generate another image for that.

So here, the AI Assistant adapts to the selection. So depending on what you select on the editor, you're going to have Text Generation, HTML Block Generation, or Image Generation. So in this case, what we are doing is we are leveraging Firefly to create images, right? So think about the keynote and the information you've seen there. Firefly is evolving. Now we have structure reference, style reference. Those type of capabilities will really allow you to create a very accurate content, based on the information you have. So I have here another image that I like. It's pretty good. So I'll just select this. And now we see in the editorial that we have two variations, right? So this is how you can build personalized content but also like generate images. Okay. That, that's great. That's gonna really help me, for sure. And so personalization is key. We have we have seen that. So we can do personalization content for segment audience. But how can we achieve personalization at the one-on-one level? Yeah. Yeah, yeah. So we talked about individual content, but here at the end of the day, I was still working on segments, right? So let's go further. So what I'll do is I'll deactivate conditional content. I'll remove this image and just delete it. And instead I will use an offer decision. So that's also existing features, part of our editor. We can just add an offer decision. And what we do here is that we are going to configure this decision. So there are two things that we need to do here. One is select a placement. So there can be multiple placements in across your different channels. And here we're just going to select the WKND Fly email section. So that way the offer that we are going to pull, are going to fit the template. And we also have preconfigured the decision. So what is a decision? It's a set of offers that you pull from and it's also, details of how you pull it, right? What are the eligibility criterias? What is the ranking algorithm to pull those offers? And that can be an AI algorithm.

So at this point, I just set the offer. I have a preview of that. So now it's not just a few conditions. It's really like offers to promote the WKND Fly Club, lounge, the lounge specifically and here I have a bunch of options. Maybe you like cocktails. You like drinking a coffee at the lounge. Maybe you prefer to have some rest. - Or maybe you like to listen music. - Listen to music. Maybe, Arthur, you like music. - I do. - I don't know.

Personally, I prefer just sitting comfortably there, but the idea is you see infinite possibilities here, right? It can be 50 offers. It can be more, right? And here, we've combined generated content, but we're also combined with offer decisioning that can deliver the next best offer based on AI algorithm.

So, yeah, Arthur, that's what I wanted to show regarding personalization. Again, amazing. At... This, all of that is through the email channel, right? Yes. Are you, because as a user, I don't check my email daily, right? I'm on my mobile phone using apps. So are you able to do that with another channel, like mobile? Yeah. Sure. So, Arthur, we showed like how we can level up our email game. But I think it's not just about email, right? So let's move on to Push notification here, for example, right? So here, I didn't use the AI Assistant. I just built my push notification again to promote this WKND Fly Club. I have an image. It's a rich push. I have some, title, and a message here, and this is our editor for Push notification.

It's working both on iOS, Android. I can preview that. But again, here, I want to pull the same AI Assistant. So it's always there on the right rail. It's there to help you. And what I can do here is I will say, unlike what I did for email, I have this little toggle that says, enhance current content. So I'll enable that. And so what it means is that, now the content that I have is going to be used as the baseline content.

I will say, okay, make it more appealing for business trips...

And let's say, just generate.

So in this case I have variations, right? I can choose, like, also which type of field I want to generate. So, again, it can generate a full Push message, or it can generate just the title, just a paragraph of it.

And here, I can very quickly just apply it. So I have an alternate variation. Let's say I like this one. So my content got updated.

So it's that easy to just use the content.

If we recap so far, it's amazing. Like, I'm able to create content with Generative AI very fast. I'm able to tailor this content based on my brand image, thanks to brand assets. That's great. I am able to use personalization with audiences, but also at the one-on-one level. I'm able to do email. I'm able to do push. But I think we are missing one thing here. We are missing how to measure the success of all the experiences we are sending here. So do you have a... Can you demo? Do you have a solution for that? Yeah. Sure. Sure. So we were talking about measurements. So here, let's say I do another scenario. So I don't want to enhance my content. I actually want to use a brand-new prompt to generate brand new content, right? So in this case, again, I will click Generate. I will have a few options. And I check every options, title, message, images, so I will have full push notification, a few alternates here, right? I can preview them, and I see, like... I like for example, this variation.

And this time, I'm not just going to use it to replace my content. I will use this enable experiment toggle on the top. So if I do that, what happens is that I can select this variation and add it to an experiment. So view it as a shortcut to our existing experimentation feature.

So what this does is now I see this left rail where I have multiple treatments. I have this treatment A, that was my original content and I can switch to the treatment B, right? And this is the generated content that I just selected.

So very quickly, I have the two at my disposal and I can click here on experiment settings. I can configure this experiment.

While configuring an experiment, you have to choose a success metric. I have unique push interactions here, but I have a lot of metrics available based on which, I can optimize.

I have a distribution of my population. So in this case I have a 50-50 distribution. But, if, for example, you feel less confident about the generated content and you want to test on a smaller amount of the population, you could very well do that as well.

You have all the groups as well. So this is really allowing me to optimize my campaign.

So here, let's just save it. And we're going to activate this push campaign.

So we already have an audience setup. We have everything to activate it. And so at this point, the campaign is live. Right. But the idea is that after sometime, what I want to do is I want to go check the reports. And we have a report on experimentation. So overtime, people have opened. I have clicked on the push notification and we're going to have metrics. And you get a report like this one where you can select your baseline, which was treatment A, and you can start seeing which one is the winner treatment. In this case, it's treatment B, meaning my generated content was performing better. I have a lift measurement. I have also a confidence level. So we see 92% confidence. And the confidence level is really also based on the size of your audience, how much data we had to backup this lift. Right. And we also have, like, confidence intervals, so you can go into details of really analyzing like, how confident we are on these results.

So this is like giving you the data to back your initial content strategy. Maybe you thought about... Here in this case I wanted to make it more appealing for business travelers. Is that something working in general, better on my audience? That's what I'm finding out here.

So, Arthur, this is, I think, like the end of the demo. So we've shown all the four points that we were talking about at the beginning, creativity, productivity, personalization, but also how we can measure the impact of the creative content. Yeah, you really rolled up your sleeves here, good friend. So thank you. Thank you very much. You truly deserve the gold medal here in the Adobe Journey Olympic. So that's awesome. Thank you for that.

So we have seen here all features that Fred showed. They are in Beta. They are... All features to me are really reshaping the way marketers plan, execute, and, measure their campaigns, journeys, and really think about the AI Assistant as, as an assistant that will help you in your day-to-day task to achieve your task faster, better, and really can have outcome on your organization, like reduce in cost as well. It's really like the image here is passing the baton to marketers, mean that as a marketer, you are still in control. There's our AI Assistant features that are helping you, to give you a competitive edge, and boost your creativity and everything that Fred mentioned. So thank you, thank you for that. Also, goal with this AI Assistant features is we to equip you to, to get this gold medal in the market against your competitor.

So the gold medal is really at your fingertips here and all the AI assistant features that we have seen with Fred will help you achieving better levels of engagement, through better content and improved creativity.

Our Beta users right now are really amazed by how easy it is to be more creative and it's giving them more ideas while remaining in control. The second point that, that is really important here is streamline content creation process, increase efficiency, and reduce cost, right? Marketers can, marketers before, with everything that Fred show, like, took hours. And now as you've seen, like, you can deliver that in a matter of minutes. That's really, truly amazing. Drive revenue growth through targeted content strategy and personalized experience.

And the fourth point, really important to me, measure content performance with precision and optimize, for maximize, for maximum impact.

Yeah. Just one thing to add, Arthur. So we've talked about... You can wonder about content created way upstream, right, before even coming to join an optimizer, Adobe Campaign. Content is created. Maybe you have a brief, right? The content is coming from there, right? And that is a perfectly valid use case where more content centric personas would work on generation. But what we believe in is that, as part of the content supply chain, the last mile when we deliver, right, where content and data are assembled to really create personalized experiences. There's also there an opportunity to leverage GenAI, right? And this is what we've shown you today because at that moment, if you think about that persona, right, you're inside your campaign. You might have already an audience, a schedule. You have the world context that is informing you. And you as a marketer have a different vision as well, right? It's not just about content. It's about your goal. It's about experimenting. And that is a unique opportunity where this type of feature is generating content. It's assembling content from the sources that you have. Let's say you have your assets, maybe that were generated and that are available in AEM. You can pull them as well, right? So this flexibility of pulling assets that are there, generating, combining, and driving intelligence, we view that as a unique opportunity.

All right. Well, thank you, Coach. Thank you for participating at the Adobe Journey Olympics. [Music]