How Adobe’s Marketing Organization Scaled Content Creation with Generative AI

Transcript

Thank you for joining today’s event. Overwhelmed to empower all Adobe’s marketing, organization scale content creation with generative AI. I would like to introduce a speaker for today. Louise McCord, senior generative AI strategist for Gen Studio and Adobe. Please, you have the floor. Thank you. Carl. Hello, everybody. My name is Louise Marcotte. I’m a junior strategist at Adobe, and I have with me here today. I have a look. Fernandez, who is a paid media SEO director here at Adobe as well, and our marketing organization, and Danielle Fehrenbach, who is a senior LCM marketer. So you first hear from me. And then at the second half of this, we’re going to have a roundtable with Alok and Danielle to talk about how Adobe, has been using generative AI to power our marketing campaigns.

So quick agenda here. We’ll start talking about the content challenges that market organizations currently have, including us. I will give you guys an introduction to Adobe generated for performance Marketing, which is the newest application that we are launching now at Max on October 14th. I will talk to you about the use cases that we have solved at Adobe using generative for performance marketing. Then I’m going to give you guys a demo of of our performance. And I’m very excited about that because it is a really interesting product.

Then we’re going to have the roundtable with Daniel and Alok, and finally at the end we’ll have room for Q&A. So please post your questions in the Q&A chat. If you have a. First, I want to start by asking you guys this question like, what are the biggest challenges that your organization faces when creating personalized content for customers? There’s four options there. I would love to to learn which one is, you know, your biggest challenge? Seems like creatives are overwhelmed and cannot meet the demand. Is the top one for now. And then marketers have to kind of go back and always be asking for new versioning and things like that. Second one.

Yes. Well, thank you for answering. So I, I feel like your, your, your, your, your challenges are very similar to ours.

So if we can move. Thank you. So your channels are very similar to ours, right? The challenges that we have at the Adobe, Adobe’s global marketing organization, GML. So we are a very large organization. We have more than 1000 employees. We have annual marketing spend of more than $1 billion. We operate across multiple channels and markets. For example, as a you know, as an idea, we send, you know, billions of emails across more than 40 languages, right? So we are a massive marketing operation.

And to run this operation content. Right. We need a lot of content if you, you know, just sort of back of the envelope, calculations here. Right. So if we have 15 channels that we operate, right, Facebook display, TikTok, and then we have eight products. We have more than that. But let’s assume, you know, let’s just start with eight, right? And if we operate across 35 languages and if we’re trying to be best in class, meaning we are personalizing this content or we are optimizing it using, you know, algorithms, then we need a pretty variance for this content. And, you know, sometimes you can go with three variants, but, you know, platforms like meta ask you to go with ten, at least ten, so between 3 to 10 variants. And if you’re trying to refresh this content because content fatigue is real, attention spans are just smaller and smaller lately, right? So if you try to refresh that every month, that is a content need of between 37,000 pieces of content to 126,000. So that is a lot like to give you an example, at Adobe last year for, Black Friday to run Black Friday globally, our studio team and our agencies had to create 52,000 pieces of content to serve Black Friday. And that is just a lot. And it does not leave enough room for, you know, our studio team and our agency partners to to do anything else. Right. So what happens is there’s this ginormous demand for content, right? And it just keeps increasing and increasing and increasing. And there is a part of it that our agencies and our internal studio team can fulfilled. But there’s a huge unmet need for content, right. And then when generative AI came about, we had Adobe. We thought, you know what? We should definitely leverage that, right? To try to close that gap, try to close the the this unmet need. And we’ve done this in two ways. One of them is by supercharging creatives, right. Supercharging studios and agencies with generative AI technology. And we’ve done that in Photoshop and Illustrator. And you’re going to see a lot more of that at max. And but the other way that we thought we could also address is by using generative AI to empower marketers themselves to self-serve, on brand content, right? Creatives as well, but also marketers. And hopefully they can help us close that gap. And here I’m and here I am to tell you the story about how Adobe has done that.

And we have done that by creating this new application called Adobe Gen Studio for Performance Marketing. And I will read you the tagline for the product. So Gen Studio for Performance Marketing is a Gen I first application that enables the modern marketer to self-serve on brand content with enterprise guardrails and governance for high performing experiences. And we’re going to release this product on October 14th at max. And then we will. So we will be able you will be able to create Engine Studio for performance, for marketing paid social ads, emails and also display ads.

The key value proposition for Gen Studio is to allow marketers and creatives to reuse existing content to generate brand new content, and to create content variations across multiple languages, for example, or audiences.

So Gen Studio, will this content will be on brand, it’s going to be high performing and it’s going to be personalized and all of that within enterprise guardrails of review and approval.

Now, what have we at Adobe done with generative for performance marketing? What has our global marketing organization been able to accomplish? Right. Because since it’s our product, we have been using it for a very long time. Right? We have been test driving this product. We’ve been kicking the tires for more than a year now.

So and these are the six use cases that we have been able to address so far. So testing at scale, we’ve been able to increase our testing velocity for both email and paid social. We have it has allowed our marketers to jet to serve more audiences that they couldn’t before.

It has given creative independence to our regional teams that have always struggled to generate their own content in their language, for their cultures. It has allowed us to refresh and reuse content much more quickly. It has facilitated real time responsiveness. So if something happens in the market, we don’t have to wait. You know, weeks and weeks and weeks to react. We can do it faster. And then also it has created a brand new process. You know, I call this like a creative loop. So between like creative performance measurement and then informs the creation of new content and, and that’s a Loic and Daniel were here to talk to us about some of these use cases. So I do want to ask you guys another question, which is, so what are your for you guys? Right. I just told you the Adobe ones. But what are your main primary business cases when you consider Gen I like using Gen I for content creation. What would be your primary use cases? We’ll give 30s accelerate in-house content creation. That seems to be well, seems like everybody has the same the same challenges that we do. It’s great. Can all work on this together? All right, ten more seconds. All right, so accelerate in-house content creation. We’re just trying to empower our internal team to create more content. And that’s definitely again, our biggest challenge as well. So, going back so now I want to talk to you guys about four in more detail about four of the use cases that I just presented before. So the first one is testing at scale. And we did this for email and for paid social. So before we off, I mean, 90% of the time we could only really our campaigns only had one version of everything, right? It was very hard to really test at scale and create multiple versions, which is really the golden standard of, you know, of personalization, right? The technology and big data and these other rhythms, personalization algorithms like Multi-armed Bandit, it’s like they’ve existed for a long time, right. But they could not really be, you know, we couldn’t take full advantage of them because we didn’t have the content really to, to to do that. Right, to take advantage of them. That’s redundant. But so, so that has been the challenge that we’ve had. So our approach has been, you know, using density for performance marketing and engine AI, right, to create multiple variations. And you guys, we’re going to see this in the demo, how you can use it to create multiple variations very quickly. And, and we’ve saw some really great impact. So for example, in one of our tests we did last year for Max last year. So a year ago, we, we did a test, an email test, and we saw a 57% boost in CTR, by, you know, sending five different versions and, and serving the highest performing one to, the whole audience. Right. We always it’s also allowed us to do something we call this always on test. We have these always on tests now, in email, for example, for subject lines where, you know, we’re, you know, every two weeks we are just switching the subject line, you know, using Jen. So you’re creating a subject line in the studio and switching the subject line in production. And a new one, and that has increased one of our campaigns increase the open rate by 8.5%, which is a lot when you think about lifts. And, you know, in page. So show it has allowed us again to create multiple like Facebook, like meta. They recommend that everything you every campaign you run there should have between 8 to 10 different creatives so that then their algorithm can work its best to serve the right content to the right audience. Right? Most of the time we could only give it one creative. Sometimes two. Right now we can give it a lot more. And in one of the campaigns we did this for, we got 10% higher ROI, versus a comparable campaign that we could run before. So, like testing at scale has been a big unlock for us. And we will we will speak to this, from, Allo. Continue. Another thing that has allowed us to do is it has allowed us to because we have more content, we can create more quickly, a lot more quickly. It has allowed us to, serve audiences that we couldn’t before. So our email team, our LCM team, which Daniele was here to talk to, they, you know, they have generally they, they have 12 audiences that they’ve been trying to serve forever. Right. But because we didn’t have enough content, they generally would have to create, you know, one version of an email and then serve it to, you know, most of these audiences. Right. But now with just sort of a performance marketing, we can create one version for each audience, right? One like one personalized version for each audience. So it’s really enabled us to serve 100% of our audiences and for email and, you know, in general would speak to this like this quarter in Q4, our LCM team are 100% of our acquisition emails in the US are being created engines. We offer performance marketing. We’ve run the entire program on just for for performance marketing. And we’re able to serve all of these audiences now with personalized content.

Another thing that it has allowed us to do, because we can create content so much faster, right? Because marketers like Daniel can now create their own content.

Then what has allowed us to be a lot more responsive, in the market? So when something happens or something is happening, you know, before something happened in the market, because it would take our agencies or studio teams, it would take like a creative brief, and then it would take a meeting and then it would take a lot of back and forth, and we’d take between 2 to 3 weeks to get to the final content. We couldn’t react quickly enough. By the time we had the content, the the event, like the thing was gone, right. But now, because marketers can self-serve this on brand content, it we can be a lot more responsive. So, so two examples of what we’ve done, for example, has been, you know, we’ve been able to do a fall campaign last year, which is something we’d never we’ve never been able to do seasonal campaigns like that. It was a question we always had like, well, what if we did a seasonal campaign like we don’t have time, so we’re not going to do it right, but we did it, and we saw a 9% ROI increase relative to unpaid social relative to, to previous campaigns. Another example was this year. We saw, you know, in one of our international markets, we saw, like a pocket of strength and we’re like, oh, man. Like, let’s, let’s, let’s go, let’s react to that. Right? The Olympics was going on. It’s happens to be a country that loves the Olympics. And we’re like, let’s create a campaign that, like, involves the Olympics. It’s like, no time. But the marketers themselves like using studio for performance marketing to create these campaigns. And, they are they’re pretty cool.

The other thing, the other, use case I want to highlight for you guys is, is, this creative loop that has existed. So one another thing generally for performers marketing does is not only creates great on brand marketing content, but it also will measure the performance of this content at a very, very granular level. Right? So if you have a piece of content, we are going to run it through our generative generative AI algorithm. And we’re going to feature, we’re going to say what is the background color. The foreground color. What are the subjects. What is the theme. Right. All of those different things. And then when you have enough volume of those insights, then you can see like what is the most highest performing color? What is the highest performing theme for each particular product. And that can inform the, you know, the generation of your next campaigns. Right? It’s so it’s this sort of cycle of insights between creation and between measurement, that we’re very excited about. So at Adobe, you know, one of the things we did with, with this new capability is, one, you know, on our website, we have a page with a lot of templates that we wanted people to download. And we ran we, you know, all of those pages got feature ized. All those those, templates were feature A’s. And we’ve realized that one color in particular, had a very high CTA, right? A very high download rate and CTA. So we were like, well, let’s create more of those templates and let’s put them on the page. And that resulted in a 35% increase in downloads for templates. Right? So, so that is the type of insight that you can only get if you if you look at those insights across everything right across like aggregated across enough volume of data. Right. And and Allo is going to speak to us about a different use case as well that his team is working on. So now these are the use cases that, that, we have addressed. But now I want to do a demo and show you guys generative for performance marketing.

Okay. So we’ll start this by showing you guys the what I call the foundations of Gen Studio. And the main foundation of Gen Studio is the brand. Gen studio is built so you can build on brand content for your customers. And so let me show you guys what the Adobe brand looks like for the Creative Cloud and Document Cloud. Here we have our brand voice guidelines like our tone of voice, our editorial guidelines, our editorial restrictions, for example, our editorial guidelines. We use the active voice, our editorial restrictions. We avoid all caps. That’s the way we speak. That’s the way we talk. That’s our brand. We also have channel guidelines. So for email, for example, our subject line is between 35 to 50 characters for meta ads. Our headline, always ends with a period. Like those are the things that make up our brand, and those are all the things that Gen Studio needs to create on brand content. Another foundation that we have here is products. So here is where you have all the information about all your products and that you. So then studio can create content about these products to your customers. Then you have definitions of your customer segments, your personas. And this is where that information goes. So now that I’ve shown you guys the foundations, let me show you what you can create with these foundations. And we are going to start creating an email. So we will create an email here for the Creative Cloud. So we’re going to search here in our templates for a Creative Cloud template. We are going to do a Creative Cloud three pod template. So three pod templates. We have the first pod over here with a headline image. And then we have another image with another pod and another image with another pod. That’s the template that we have. So we are going to create this email by selecting our brand, which is going to be the Adobe brand for the Creative Cloud. We’re going to select a persona. I want to create a campaign for our creative professionals in the UK. Then what I want to do is, since this is for the Creative Cloud, I want to highlight a different product in each of those pods. So for the first pod, I want to highlight Creative Cloud All apps because that’s what we’re talking about here. For the second pod, I want to highlight Photoshop. And for the third pod, I want to highlight Adobe Express, which is one of our newest Creative Cloud applications. Then I will select an image for each of these pods. So for the Creative Cloud, I will select this awesome psychedelic token that was created in Firefly for Photoshop. I will select this amazing portal image that I myself created. Firefly for and for express. I will select an image from our content, which you know has been previously used and that contains product expressions. We want to highlight the product. So this is the one that I’m going to select over here.

Now comes time for prompting. This is where I’m going to write my prompt. But first let me point out that there’s a lot here already. There’s a lot. The gen studio already knows and knows about your brand and knows about who you’re talking to. It knows about the products that you’re highlighting. So there’s a lot of things that it’s doing here. But I have my prompt here, which I wrote previously. So in the prompt I’m just giving a little bit more direction to Gen Studio. And I’m saying the purpose of this email is to engage our potential customers, introduce them to the new features in each of these products. We want them to try these products and use them to express themselves. And then I’m being even more prescriptive here. I’m saying for the pod one, which is about creative Cloud, I want to highlight how Creative Cloud all apps and names are full creativity helped by AI. For the second product for Photoshop, I want to highlight the generative AI features in Photoshop. And for express, I want to highlight how easy it is to use Adobe Express and also highlight the new features. So those are the things that I want this email campaign to have. So now we’re going to create an generate and C Gen Studio.

Now what’s happening is Gen Studio is processing the request. And you can see here in real time as it produces the copy produces the subject line. Now it’s producing the pre header. And then soon we’re going to see the rest of the creation.

Awesome. So I created these four versions of this email for Creative Cloud consumers in the UK. And you can see here that it targeted this this email towards this audience. So it’s for example it’s highlighting the new features in the product because it knows this this audience in particular already knows the product. It talks about their ideas being fully realized, unleashing their creative potential. If you come over here, you’re going to see that it followed my directions in terms of highlighting Photoshop and highlighting gen AI features in Photoshop. You know, it talks about professional results because these are professionals. So you can see here that it’s it’s targeting. It’s tailoring the message to the persona that I want. But what if I want to actually go even further and use this exact same prompt for a different audience? So this is what I can do. I can use the exact same prompt, only change the the persona. And for example, here I selected the hobbyist persona. So the this is exactly like the the email series. We just created. I just changed the persona. And here you can see the results of here you can see the result, that we get and we get something that’s different art hobbyist audience. It’s not a professional audience. They don’t know our products very well. They’re still discovering these products. So you can see here that the language is about discovering. It’s about enhancing. It’s about exploring new tools for your imagination and things like that. So it’s targeted the message to this persona. And I literally did that. I had already this loaded to save us time. But, you know, this would have taken just an extra minute to for me to create a whole email series for a completely different persona, leveraging the exact same prompt. So this is pretty cool, but there’s more. So now that we have this copy here, we can we can take a look at it first of all. And I can see here, for example, that this copy this headline, there is a widow line here. So you know that your line is in the middle. So there’s both a mobile view and a desktop view. I can look at this what this email is going to look like in a desktop which is great. So you can see this in different ways. But most emails are open and mobile phone. So I don’t want to have this widow here. So what I can do is in Gen Studio, I can click on this and I can say I can either ask just to rephrase this, I can ask you to shorten it. I can ask it to lengthen it, or I can prompt directly just for this headline. In this case, let me ask it to just shorten it.

Perfect. Also, let’s take a look. I’m going to go with Ignite Your Creativity and then replace. And there you go. I have a brand new headline here. Another thing I can do, like I said, I can prompt specifically for each item in the canvas. So for example, this image over here is about generative expand in Photoshop. And over here in the copy it is talking about sky replacement and content-aware fill, which are great features, but they are not generative. Expand. So if I want this to be about generative expand, I can come over here and say focus on generative expand and I’ll click on generate. And it’s now generating three different options for me to choose from.

And I can come over here and choose one of these. So this one this one for example, describes Genesis one pretty well. And like unlocking know creative possibilities. Seamlessly extend your photos by describing what you want to add. Begin a creative adventure so I can select this one and I can click on replace. And there you go. It replaced it for me. So this is in incredibly powerful and just ginormous amounts of time saving, leveraging gen AI. Now another really powerful thing about Gen Studio that I want to highlight for you guys is Gen Studio Insights. So with Gen Studio Insights, you can answer two questions that markers have been asking for a very long time. First of all is the what what is currently performing out there? What are customers seeing. Because traditionally in marketing analytics we have amazing dashboards. But these dashboards, they really focus on the metadata of the campaign. They really focus on the statistics, but they do not generally do not show you what is actually out there. What are your customers actually seeing. So in a traditional dashboard where you would see, you know, this, you would see Adobe CC Pro, Web A and Na and you would know like, oh, it has this many impressions and there’s many clicks. But like what is actually people seeing and this is what Gen Studio is doing here. It’s bringing that into the analytics. So Engine Studio Insights you can click in here. You can click in each of your campaigns. You can see each experience within this campaign. You can see how each experience is performing. You can see the copy of the experience. You can see some featurization here, which I’m going to mention later. You can see all the statistics across time. It’s great. You can see exactly how each of these things is doing. You can also go even deeper and go into each asset within a campaign and within each asset, you can see how is this particular asset performing. Because if all you saw is this, if all you saw is just the name, then you don’t really know what your customers are seeing. You just it’s just a very in personal analytics. So Gen Studio is bringing this visual aspect into this. Now the other thing, and one of very powerful things that Engine Studio does is answer the why. Why are things performing better? Why is one campaign from better than the other? It’s giving you insight into that. And we do that by for each asset that you have, both for copy and for images, we are featuring using it. So so for example, for this asset, the scene, it’s a bakery. The visual content density is high. The orientation is portrait. The tags. What is in this image? It’s pastry. It’s food, it’s cake. The visual attention spread is low. We are featuring this this image, and with enough of those feature eyes, then you can get into the why. Why are things performing well in your marketing campaigns? For example, here at Adobe, you can see that if I order by CR to click through rate, what has the highest CTR? It seems like it’s street photography. Street photography has the higher CTR. Now that insight in mind, let me show you what meta creation is like in Gen Studio for performance marketing. So let’s create some meta ads. So we will go here some very similar flow to email. I’m going to select a template for Adobe Photoshop.

And same way we’re going to select a brand. We’re going to select the Adobe Red brand. We’re going to create this for hobbyist persona that we created an email before as well. And this is for Photoshop. So I’ll select Photoshop. And here I’m going to select what images do I want to go in this campaign. And here I previously put these images here in our content. So I have this image here which I created in Firefly. I will say that again I also created this image in Firefly, which I love, and I will use these two other images that were used in previous campaigns by our studio team, by our amazing studio team. So there we go. So we’re going to use these four images. And then I have my prompt, which is different than the prompt for email because this is a meta ad. So for this one I am saying the purpose of this ad is to engage our potential customers and introduce them to the new generative AI features in Photoshop. So let’s click on Generate.

Great. Here we have all of these results for different ads.

I really like this one. Collaborate on fun okay. So what can we do here. We want to take these four ads and we want to have them in different sizes. Because meta allows us to have different sizes. So all I have to do in Gen Studio is I have to select the different sizes that I want. In this case, we do four by fives and nine by 16. And I will click here. And it’s going to duplicate and resize.

Awesome. We have now each of those ads was resized into two more. So now we have 12 different ads and looking at each of them, it looks like the resizing worked very well.

And let’s see. Yes, I would say so. Let’s say yes. I think this works really well, I love this. Yes, this is great. This is really awesome. So, I really like, I really, like, collaborate on fun, discover new edits. Those look really, really great. Now, one thing I also want to highlight, besides the amazing scalability of content creation here, is, you know, I’ve said before, I said at the beginning of this that the foundation of Gen Studio really is the brand. So one very powerful thing that we have for email creation, also for meta creation is our brand guideline checks. We call these the brand checks. So you can see here the different scores that each of these ads received. So I can click here. And I can see that 21 of our brand guidelines seem to be met seven need attention. Doesn’t mean that they were not met. It just means that, you know, we should check them. We should. They need attention. So, so if we look over here, one of them that, comes up a lot for us is that at Adobe, we avoid using the possessive of any of our products. So we don’t want to, like, personify the products. So we don’t say, photoshops. Features. We say the features in Photoshop. So one thing that I can do in this particular ad is I can come over here and I can edit this and I can say, create and play, with the new eye features in Photoshop that’s more on brand in actually in Adobe Photoshop, because we should always mention the product name, the full product name the first time we’re mentioning it. So, there we go. So I can make the edit, the brand check is going to rerun, which is great. You can use this to check that things are on brand. So and then finally what we do here is we can send this for review and approval. And in this case I can send this to our amazing Gen I, creative director and our amazing J writer. And they can take a look at this. We can collaborate on edits that they want me to make. I can send them comments here. They can send me comments back and forth. So that’s really great. Now, once things are published, then I can come over here into experiences and I will see that, you know, an Adobe, gen studio is going to keep all of the metadata related to that creation once it’s approved. So there is the title. What campaign was it associated with? What’s the brand that we use? The product, the persona, the channel when it was created, who created it? The prompts that was used, the assets that were used. And it also will store who approved everything. So really cool. Really great. I really think Gen Studio is a amazing unlock for content scalability.

All right everyone, I hope you liked this demo. I love showing you in studio. Okay, so now we’re going to go into our roundtable discussion. So and like I said before, I have with me a logo Fernandez and Daniel Fehrenbach and Daniel Angelo and me, I are we well, they are part of Adobe’s global marketing organization. I used to be a part of that organization until two months ago, when I decided to help our customers do the same thing that Adobe has done. So they’re here today with me. Hi. Hello. Hi, Daniel. And, and we’re going to get started with our questions.

Awesome. So first question here, goes to a look. So I look we talk about this like increase in content need right. What’s causing it.

Yeah. No that great question. So you covered some of it but I’ll, I’ll, I’ll touch on probably three things that are converging that are driving this, this trend. The first one is just Adobe size and scale, right. We work in like 70 or 80 different languages across 10 to 15 different products. We need content to be refreshed regularly. So you touched on that, Luis? I think the second thing is actually the the platforms and the change in platforms that we’ve seen over the years. Right? So you need product, you need assets, especially in the paid media side. You need assets for tick tock, and you need assets for Instagram and need assets for YouTube and YouTube shorts. And so you need different sizes, different scales, different types of assets, statics, a video, all of that. And you need them at that, that frequency. And the third thing I think that’s that is change over probably the last 4 or 5 years is really the platforms themselves. So Google and meta specifically and other platforms have followed suit, which was they move to using machine learning algorithms to basically target customers. And these systems thrive on on volume and velocity of content and content. Now is the new form of targeting. And that’s, that’s that’s really interesting. And that’s what’s driving the need for us to constantly refresh creative, find gaps and creative, see where creative is actually fatiguing.

And then driving the need for more content where we see those, those gaps or need, and then we refresh and we actually see improvements in the ROI. So in Mediavine an email drive, a ton of, of, of revenue for our business. And so by refreshing these creators, we actually able to drive a higher and higher return to the business. So it that’s that’s that’s why we need more content.

Yes. Thank you. Okay okay. Next question is for Danielle. So Danielle we know the content need. But why are we not able to meet this content. Yeah. That’s a good question. I think we asked that question a lot, with an email marketing. I think the first thing I want to touch on is. And we’ve covered it, you covered it in the position, but is speed to market? You know, when do you think about email marketers? You know, when we need new content or we’re trying to fill a content gap, we, we need to develop a brief. We need to deliver that brief to studio or agencies. We need to review it, have a kickoff call, go through multiple rounds of review, make sure that everything looks great.

And the process can take up to 3 to 4 weeks just for a single email or a couple of versions of a single email. And so when we’re talking about, you know, being able to cover content needs, that’s a huge constraint is speed to market for us.

And the second piece I think is, which is very related, is around scale and personalization.

As that sounds like there’s so many different audiences in the landscape with an email is becoming more and more complex over time. And so we’re there’s a desperate need to, create more and more content and versions for the different audiences, whether it’s persona based, category based, what we know about their purchases or search history.

And so given the time and time it takes to create just one single email asset, we’re only able to serve, you know, 2 or 3 of our top segments, right, instead of everyone. And also that that also is a constraint in terms of our ability to personalize content. You know, we we would love to serve, you know? 1224 5000s of audiences if we could. But given the constraint on content as well as, scale, we’re not able to have that 1 to 1 personalized experience at a user level. Yeah. Thank you Daniel. Okay. So next question will go for both of you. Let’s start with a look. And then Daniel, like, so how has generative AI and generative for performance marketing help address these content challenges? Yeah, I think just building what Daniel’s saying is that because we are struggling to get the assets and the content we need from our regular creative studio, angles, I think, Jensen has helped us to get fill in the gaps in our content need. Really. Right. So, Louise, you give it you could you give a bunch of examples on we’ve never done found creative. We could actually run a campaign for fun creative. We can run things that are more seasonal, more, you know, specific to the moment. We can seize on moments and generate content based on what’s what’s what’s maybe trending. We can so we typically, if you think of it as we get, big chunks of content, for, for Adobe is when, when we have release cycles. Right. So we typically get a lot of content for the release cycles, but then that comes at a very different cadence versus the cadence that we need content for.

And so we need to fill in those gaps. And density is a, is, is really helping us fill in the gaps in our content needs, where we see our creativity. Thank you. Then. Yeah. Then on email, I think it’s, it’s very similar. I think speed is our number one. Unlock and watch gen performance marketing has helped, improve. You know, I mentioned earlier that, you know, from creation of brief to and to creation of assets, it might take up to three weeks now with, Gen studio and which I it gives the marketers control to go in to prompts to generate these assets. And now we’re able to create. Net new emails and versions of emails within 2 to 3 days. Which is super powerful because that was before, and that also unlocks scaling, because we’re saving time during the creative briefing process, and filling those content needs, we’re now able to say, hey, if if it only takes, you know, myself or my team, 2 to 3 days to create one Gmail asset or four versions of an asset, we can start to scale across segments, which is super powerful and something you know, our team is always trying to do. So yeah. Yes. Yeah, yeah, I’ve definitely seen that happen. With Awesome. Okay. So yeah. So Daniel, next question is for you again. You know, like, you know, your team, right? Like, you know, and your team, you are in the, you know, the Americas acquisition team, right? So your team in Q4, you’re creating, you know, a lot of emails in Gen Studio, right? So can you speak about that? Can you speak about, you know, the purpose of that program and how it yeah, you guys. Yeah, absolutely. I think at the beginning of the quarter, the biggest question was trying to understand how we can scale Gen Studio, across our ML programs. And so we set up a pretty ambitious goal to rely on Gen Studio to produce 100% of our batch emails with, North America, which is, which is a huge goal. And so, you know, part of what we’ve been able to do over the quarter is across our entire team, learn more about Gen Studio, be able to have that control to go in and prompt across all of our segments. And previously, you know, we would only be able to serve about six of our total audiences. Whereas now, given the speed and the scale that Gen Studios helped, accelerate, we’re able to serve, you know, upwards of 12 audiences. And these are, you know, audiences that were actually able to personalize content for, you know, if I know this user is likely to engage with design products or photo products, we can give them that specific message versus having a blanket message or something that might not be as relevant. And so I think, you know, given the control, the speed and the scale that density does equipped us. Well, that’s been a huge unlock in Q4 as we’ve worked towards scaling across 100% of our, batch emails in North America. So. Yeah, that is amazing. And you’re like, you’re in gen studio every day. Yes, absolutely. Awesome. Okay, so, next question is for you, Alok. So, you know, your team has been working on, on this initiative, right? Which you guys call it. Think the efficiency layer, where you’re using Gen Studio Insights to inform the next content, like the next content, or we’re going to put in some market for paid. So can you, can you speak about that a bit about that. Yeah, I think so. So the specific product what is the Gen Studio Insights product within Gen three. And that’s really, enabled us to actually understand the why are certain assets performing well and that enable us to then run new hypotheses to test. Right. So I’ll give you an example. So to set the context that Gen Studio Insights essentially tags each piece of creative, the content, the text, what is in the image or in the video with attributes. And then we can essentially go in the gen from, do you show this in the demo? You can just sort by what are the top performing assets. And one of the most interesting insights is we saw that these these baking that’s that’s for one of the strongest performers, both in terms of click through rate and in terms of of of driving performance for, for us in the pay media side. And that then enables us to say why, firstly, we wouldn’t have been able to get to that insight if we hadn’t had these great. And so so it’s really helped us with that piece, because we have hundreds of thousands of assets.

But then from there, it enables us as analysts to understand, like, what is the what is the why behind this? Right now, you can come up with different hypotheses. You can say, maybe it’s the color, maybe people really like food. Maybe, maybe we need to try landscape photography. Maybe it is the presence of a UI, elements in the image and on those are all different hypotheses that we can just run from that one single insight and then go out and use it to actually generate new creatives to test against those hypotheses. Learn from that. And that’s the whole funnel that essentially enables us to drive higher creative performance, which which we then measure through to, to in terms of returns and improvement in ROI. So, really, really full funnel tool from, from the, from going from the insights back to creative production.

Yeah. Yeah I, I think yeah, I think it is so cool what your team is doing. And I think one of the, one of the coolest parts for me is the fact that because of Gen Studio Insights, you and Adobe Firefly engines you for performance marketing, you now have a senior business intelligence analyst that is collecting those insights and also creating the creatives and the coffee and everything. So it’s I feel like it’s such a huge unlock, right. And it also talks about, you know, like organizational change and things that, you know, that gen AI is just enabling. Right. And we certainly have had to change some of our organizations and change some processes, like having a senior buy an analyst, you know, now, like, you know, he’s a part of the content creation process as well, of course, under the direction of our studio team. Right. They review and approve everything. It’s not like he’s just publishing whatever he wants, but like he’s the one that’s creating it. So it’s I just I just think that story is so cool. Yeah. Early.

Okay. All right, so this is the end of our, roundtable. So I really want to say thank you to you guys, for being here. So, a few things that I wanted to talk about as well.

One is, everything we told you here today, we actually wrote a white paper about it, and it is already published, so it’s just a great white paper. I wrote it. So, And and here is the QR code. The link to the white paper is also going to be in, the resources part of the webinar. So please read it. It has a lot of lessons, I think in particular, you know, it has some of the content you saw here, but I think a part that we didn’t address this time and we will be addressing in our next webinar is like the change management aspect that has been required, right, to implement Gen Studio at scale. Right. Because one thing we’re very proud of, like everyone is playing around with Gen I, everyone is doing little PMQs, everyone is doing these things. Everyone is trying it out. But like we’re actually using it at scale. And to do that right, we’ve learned a lot of things and we put all of those in this white paper for your benefit.

And the other thing I wanted to also point out to you guys is that Gen Studio Academy. So we are fully aware that generative AI and prompting and all of these things, they are new. Right? So Daniel for example. Right. Like has, you know, like there’s been trainings that we’ve done with our marketers. Daniel has been a part of that. Right. And so it’s it requires, a certain degree of upskilling. Right. And we are fully aware of that. So we created the Gen Studio Academy to help our customers and potential customers and the whole world really, to become better acquainted with AI, to know how to take, you know, better, how to take advantage of it and how to leverage it. So generally, the academy is, coming. So.

Now, we have ten more minutes for Q&A, and you guys have submitted a lot of questions in the Q&A chat. So, I will try to go through all of them if I can. So let me, let me start at the top. So first question is how long did it take to populate this system with all the background information before it becomes really useful? Seems like a lot of setup. Okay, so how long would it take to for a company to set this up? To the scale of Adobe? I would say it would probably take you between 1 to 2 hours to set up your brand. Like if you take someone like someone that understands the brand, like someone in your brand team, in your studio team, and you say, do this, they can do that in 1 or 2 hours. Why? Because Gen AI is helping with that as well. So the way you set that up is you upload your brand documents and then Gen Studio will basically take it all in to understand the information and place it in all the different places. So, so all of this set up and the same thing is true about products and personas. You upload the documents and then studio itself will create those things for you. You have to go through and then like do some fine tuning. But I think like the initial setup, like it, you know, if you take if you take a person and you empower that person to do that, it will take for everything for brands, products, personas, I would say between 4 to 5 hours because the product has been set up to make that easy for you.

Someone ask, Is Gen Studio only available in beta currently? Yes. It’s going to be released in at max on October 14th 14th. Is there a way for me to download it if I have Creative Cloud? No. So Gen Studio performance marketing is not, a part of it’s it’s, you’re not entitled to it in a Creative Cloud. It is a it is an enterprise solution. So, it’s not a part of your Creative Cloud subscription.

Another question is. Okay. So basically the question is, I think is a question about insights, which is how, how are how are we collecting the data to power the, Gen Studio Insights? And the answer is directly from the platforms. So the way you set it up is you just put your platform credentials, engine studio, and then we will pull the data, you know, into Gen Studio, do all the featurization and everything. So it’s coming directly from the platform right now. We support meta. We’ll be supporting more platforms shortly.

Next question. Does it offer automatic translations of copy? The answer is it’s not fully supported yet in the product. However, this is just between us. It you can you can create copy in multiple different languages. Our Japan team is using Gen Studio for performance marketing. Already. And they did, you know, they did a test, for example, with subject lines using Gen Studio that gave them a lift of 10% and their open rates, just because it’s allowed them to, you know, be on brand content so much faster. So it does work. But it’s not fully supported in the product yet. It will be fully supported, after GA shortly after GA. That’s what I know. Next question. Okay. So like I said, like, how long does it take to get set up? Right? I would say honestly, if you if you take one person, if you take one person, if they watch like three of the Gen Studio Academy videos and they collect all the documents they need, like the brand documents, the products documents, the personal documents, it should take them between 4 to 5 hours to set everything up. It’s it’s really it was built to be very intuitive.

Okay. You showed meta ads in email what other ads destinations are available or being worked on. So display Display is the next one that is. So display is going to be available available by GA. So it’s going to be available on October 14th. Next question. Could this personalized email content based on attributes like sector company for each recipient. So for each recipient not yet but you can certainly create broader audiences. Right. So for example you could create a a persona engine studio for, you know.

Like health care, large health care company. Right. For the CMO, then you could create that persona and you can create content for that persona, but for a particular person. Not yet, but that is certainly in their roadmap, right? It would be amazing. It will be amazing when Gen Studio in the future is, you know, for example, connected to things like agile, for example, and allow you to, to do personalization, like on a person by person, basically.

Okay. And then we have another question here, which is yes. Does the content generated have the ability to be stored back to Adobe Experience Manager? Yes it does. So Gen Studio for performance marketing has a connection to, Adobe Express Manager. So you can access all of your Am assets from within Gen Studio for performance marketing, and also save it back to your Am instance. And then when you save it back, like I showed in the demo, it will save it back up with all of the metadata. Like who created this? Who approved this, what was the persona? Was the product, was the prompt that was used, all of that. So like let’s say you create something today and then six months later you want to refresh it and resend it. You could literally just use the exact same prompt. Right. And just like maybe just do a few tweaks or something. So it’s just really empowering, for, for be able to re utilize content.

And all right, this is it. This is the end of the Q&A. Thank you guys so much for being here. Thank you for asking your questions. And this was great. I hope to see you again in our next webinar.