← The Marketing Millennials
The Marketing Millennials
322 - How to Market AI Properly with Kylee Peña, Product Marketing Manager at Adobe
00:37:50
Loading summary
Transcript37 lines
- [00:02]
Daniel Murray
Welcome to the Marketing Millennials, the no BS Marketing podcast. I'm Daniel Murray and join me for unfiltered conversations with the brains behind marketing's coolest companies. The one request I tell our guests stories or it didn't happen. Get ready to turn the up. We are back with another episode of the Market Millennials podcast in full trans. I had a cold during this episode, but there was no way I was skipping the conversation. Today I'm joined by Kylie, the product Marketing Manager at Adobe. And we're diving into something every marketer and product team is figuring out right now. How do you market AI that's built into your product? In this episode, we get into how to advocate for the right AI decisions internally, how to position AI in a way that feels useful and not scary, how to tell a story that brings users with you as your product evolves, and how to build trust when tech is moving faster than your roadmap. If you're navigating how to bring AI to market right now, you better listen to this episode. Let's dive in. I'm excited to welcome Kali to the pod. Thanks for joining.
- [01:20]
Kylie Pena
Yes, thanks for having me. Excited to be here.
- [01:23]
Daniel Murray
One thing that has been popping up like wildfire is AI tools, AI products. I feel like everybody says their tool is AI powered because that's the nice thing to say. So how are you thinking about positioning yourself and messaging in a sea of all these AI tools?
- [01:48]
Kylie Pena
Yeah, it's been very interesting to be part of the first generative AI tool inside of a video editing product. And of course there's a lot of, there are a lot of AI tools in the market, but there's also a lot of customer sentiment around AI tools that can be anywhere from really excited to very fearful. I'm located in Los Angeles, so the economy and the state of the business here has been very challenging. So having AI rise up during that has been tricky. And I think something that I always come back to when I am writing positioning or messaging for products with AI tools is that it's not about replacing human creativity, it's about enhancing it and making these workflows faster or smarter or more intuitive so you can really get into what you want. And I think something that makes an AI tool stand out is having that transparency in your marketing. Not just saying and with AI, woohoo. But rather leaning into kind of responsible messaging. So it's not a black box, but rather you're talking, you're being upfront about the ethics, about provenance, about data transparency and how things are trained. I think that helps to stand out, especially with creative tools or anything where your data is really involved in a sensitive way.
- [03:19]
Daniel Murray
I also want to take a step back because I know people might not know definitions of what AI means. So when you say generative AI, could you just describe to people what generative AI actually means?
- [03:37]
Kylie Pena
Sure. So I define generative AI as something that's creating like net new pixels or new text or something like that. At Adobe we have, at least in the product I work on Premiere Pro, we've been doing assistive AI for like 10 years and that's something that we're pretty used to having that machine learning type of thing where it's sort of more algorithmic or just it's not creating anything new, but rather be like automation. But generative is creating something brand new based on your input.
- [04:13]
Daniel Murray
Yeah, that was a great definition. I also think that things that Adobe is doing, I think what's great is when you have a foundation of how usage of a tool and how people have used it and what they need to help on. I think one thing I think you can probably debunk for a lot of people right now is I think a lot of people, you, even me for a long time thought Adobe was this hard tool to use. You had to be like a top tier graphic designer or you would have to spend years learning the tool. But I think generative AI and what you're doing has made it that barrier to entry way easier for people to edit photos, maybe not do top tier graphic design, but edit photos, come up with sketches, come up with what they're thinking by using AI.
- [05:10]
Kylie Pena
Yeah, I think that with a lot of the tools that Adobe has been showing in research phases and then introducing to tools like Photoshop or Premiere Pro, it helps with kind of getting what's in your brain out and into the computer a lot easier. It's interesting because if you really dive into the craft of graphic design or video, I wouldn't say that AI has necessarily made it like easier to learn, but it's made it more accessible to get into, to get started and to get into and then to learn and to feel like you have kind of support by the software. So there was a sneak that we did where you could take a sketch, just kind of sketch something out and then it would bring it into a 3D design or a photo. And then like in Premiere Pro, just being able to just drag to extend a shot because you need a little bit more. If you're trying to cut to the beat of a Music video now you can just drag the edge and with Generative Extend you can add a few more frames. So things like that, that would have required more fiddling around and finesse and kind of non creative work to figure out how to just get the thing done, that's all being brought to the side. But then you have the real spark of things. And I think if you look at how people prompt with like text to image, you can see that human creativity and that like specialness that the ideas that you have are still very necessary. Because if you don't have any ideas and you go to prompt, you're not going to have as good an experience as somebody who's got like a clear vision. So I always think that's very interesting too.
- [06:57]
Daniel Murray
I want you to also I know you're working on positioning and stuff at Adobe for the product, but I want, how do you educate non technical stakeholders to avoid both like the overhype or the skepticism of AI?
- [07:14]
Kylie Pena
That is definitely a challenge. And as an inbound pmm, I do a lot of work directly with product marketing or product management, engineering design and it's people with all different levels of experience. So it's really important to just get into the table stakes of what this thing does and what it doesn't do. And it's really important to be clear about the expectations. Like when we were building Generative Extend, Generative Video was and still is evolving very, very rapidly. So as we were trying to build that tool, I was always very, very clear about the fact that when we were communicating to stakeholders of different levels, especially executives, we needed to show demos, not cherry picked demos of Generative Video where it was working, but the real test cases where it's not. And I think being honest internally and not trying to spin anything so that you get your product shipped is really important. And just telling the full story of what this is going to do for us and where the pitfalls are going to be just helps set expectations. And it's evolving so rapidly that I feel like it's like a weekly cadence that you have to update. Because when we were building Generative Extend it felt like one week it was kind of creating nightmare fuel and then the next week it was beautiful and perfect and it was. That's how fast this is evolving. So yeah, honesty and over communication and then just drilling into the finer details of what this is and isn't going to do for a customer, really important.
- [09:04]
Daniel Murray
I know one thing PMMs do a lot is communicate with a lot of different stakeholders and customers and talking about what's happening. But since AI is moving so rapidly, how do you improve those feedback loops and then also like take those feed lab loops and decide, hey, what these people are saying we need to build or we need to message about versus over. Because someone may today say like that feature is good, but then in a week it could change because AI is moving so fast.
- [09:38]
Kylie Pena
Yeah, we have a lot of regular cadence set up where we are talking with our research teams and then communicating up any progress or regression or anything like that. But I think it's important from my perspective in my role, I feel like it's important to stay up to date with what's happening in the world. So a lot of my time is spent understanding all of the progress or the happenings in the AI world and then also with the broader creative world. And because I'm not only working on AI tools, it just so happens that that's kind of pervasive with a lot of different components. Right now I work on the core workflows in the video editing tools as well. And I'm also thinking about how could generative or assistive AI really bolster those. It's a lot of self guided education and learning and then finding ways to package that up and communicate it out in little increments that people can digest. Because not everybody is going to sit down and listen to five podcasts a week. And when you're communicating with executive level stakeholders, bullet point lists and things like that are even more important, like the cause and effect, the value prop, things like that. I feel like managing all of that helps to prevent people from spinning out on any particular breakthrough with AI and just hold the course forward. I think it's very easy to start shaping a feature or a product and then sort of get knocked off course. You're like, oh, what about this thing? What about that thing? Or Veo did that, or Sora did that. And just know what the core problem is that you're trying to solve and who it's for and just keep on doing that and working with the people that are very close to the research without trying to run to the next shiny object. And I think that's what we've been doing and it's been working pretty well.
- [11:42]
Daniel Murray
Yeah, I think you made a good point. Just like for marketing in general is note what other competitors are doing, know why they're doing it. But also at the end of the day, you're creating for your audience, you know what they like. So if you pivot because someone else is doing, they don't have the same. They might have a different audience or they're trying to go after someone else and you're staying off the course. So I think that's a great point of stay the course on audience and note that competitors and other people are going to do, are going to come out with different things. But you know your audience, you're doing research, you know that this is what they need. One thing I know that is hard with AI is transparency of the data, transparency of the imagery, transparency of the creative. It's. It's taking from. And I know that Adobe is big on transparent AI and how it's used. So can you describe how that goes into your messaging and how that eases the stakeholder? Because obviously marketers also don't want to be dealing with legal and all these problems as well.
- [12:55]
Kylie Pena
Yeah, you don't want to be dealing with problems after the fact. But I think having a strong relationship internally with legal during all this is really important. And so I got to know a lot of our legal representatives, which is great because they're awesome. But yeah, the huge piece is transparency and trust. People trust certain products and if you start introducing aspects that feel a little bit squidgy or, oh, I don't know how I feel about this, that trust erodes and then they don't want to use your product for anything. So when we launched Firefly, which is Adobe's generative AI model system, we made it a point to always highlight what it's trained on. It's trained on things that we have permission to train on. It's trained on Adobe stock license content, public domain content. So I was not part of the original Firefly team, but I have brought all of that into the messaging when I am bringing Firefly into the products that I'm working on at Adobe, including Premiere Pro. So whenever we're talking about generative AI, that's always we talk about the value prop of the tool and then it is trained only on things that we have permission to train on. It's designed to be commercially safe because we are designing AI tools that are creator friendly. We want to do right by creatives and not take their work. Obviously, we are creatives ourselves, but we also want what you create to not cause any problems for you. So if you try to go in Firefly and generate type in Spider man, you're going to get something real scary because it's never seen Spider man as we know it. And you can kind of describe very long prose of what Spider man looks like, but it's still never going to get it right because it's never seen that ip. So that's one aspect. The other thing is Adobe co founded the Content Authenticity Initiative, which is a group of organizations, journalists, things like that, that are coming together to support provenance and trust around AI. And they rolled out content credentials, which attaches metadata to AI generated content. So there's always a clear chain of custody for how that was created and then how it was altered. And so I think leading the charge with things like that before the generative AI boom really happened and then bringing that in has been really important. So one thing that I brought in, I wanted to make sure that we had when we rolled out generative AI in Premiere Pro was the ability to attach content credentials in Premiere Pro and any tools that could export video, such as Media Encoder. So part of my job is not only the messaging and positioning, but also the advocating for certain aspects of tools to work a certain way so that I can write messaging that is truthful, such as you can attach content credentials when you add generative AI in Premiere Pro and that helps to provide a nutrition label for what went into this. And it helps to provide an understanding of how it was altered. Because I think that sort of thing, not just messaging, but also believing in it and doing something for the betterment of the creative ecosystem all goes hand in hand because it's sort of like it's not just what we say, it's also what we do.
- [16:35]
Daniel Murray
One thing, I know you said this earlier on how you send bullet points to leadership, but I want to know what is your process to get buy in internally for a decision that hey, this is the path we want to go down and not fall under pressure of someone saying hey, this needs AI or that needs AI or this needs to message AI. So how do you one get, what is your process to get internal buy in for an idea or something you're advocating for versus also how do you also fight back the pressure of someone saying oh this, I think this needs to have AI or this needs to say AI or this needs to say this.
- [17:23]
Kylie Pena
Yeah, we have. Adobe has multiple layers of leadership, so it is a lot of communicating upward and having meetings early enough in the process where you can communicate things before sort of the rest of the marketing train is less left station. One thing that I spearheaded last year, or maybe the year before time has no meaning for a marketer was really gathering a core group of people that work on the product internally and coming together to align around product and release readiness. So I work Very closely with PM engineering design. Like I mentioned in program management. Very, very important partner in program management to have all of our work streams come together on a regular basis and report it through, you know, a simple deck. It's like an executive summary, but it's a holistic view of what are we doing to make sure the product is ready, what are we doing to make sure that the messaging is tested. Every component that would go into the success and us meeting our goals for that release is in this one place. And we use this as sort of a way to align all of our decision making so that when we have a recommendation, whether it's we should continue on this path or we should rethink this or we need to change the messaging, all of these core stakeholders that are on tactically working on the feature are going to recommend it as a group. So we even get down to a go no go decision like is this ready? Is all the components ready? And here's showing all of our work. So we show all the work and then we recommend it. We co sign it together. And I think having that sort of unified front among all the people that work on, I mean it's like, you know, six to 10 people that work on this tool is really important. If the decision is early enough that it's like I don't think this is the right path. If we're researching something and we decide I think we should go a different direction. It's a similar group of people that come together and kind of lobby that upward and go to these positions. A lot of it is using my manager to help shape the expectations so we don't get to a point where we really have to then say, oh, we need to pull the plug. We're part of the conversation from the very beginning with marketing requirements documents and then all the way through to the end. So I think that kind of communication and just seeking alignment and partnership with these technical and program management groups is really critical to having success or choosing the right path.
- [20:16]
Daniel Murray
What are some things you do to equip your manager with the right materials to be able to help advocate for you? So like I think that's like something that people struggle with is you want to make it easy for your manager to go talk to other leaders to advocate for you and your team. I think one thing you said which was great is group buy in that like there's a group has said this, but what does that look like? What does the end product look like when you're giving it to your manager to go message us to other leaders.
- [20:53]
Kylie Pena
Yeah, a similar thing where we're creating this sort of living document deck where we have a roll up of our current progress. And I will use that when I'm talking to my manager or as a tool for her to see the data that's laid out about the research or the scientific evaluations or the feedback on messaging. And then, you know, it may not be as polished when it goes to executive level, but using that throughout the process to show it's not just me saying that it's pm, it's engineering. Like we're coalescing around this idea. Because she'll often ask like if I have an opinion or an assertion, she'll say, are you aligned with PM on this? And I want to be able to upgrade, be upfront and say, yeah, definitely. The other thing that I do to support my manager and really the whole PMM team is having very comprehensive marketing requirements documents written. I established a new template that we're using where it's very, very driven on identifying all these trends and doing very deep internal and external research. So it's, it's kind of like writing a paper. It's work cited, research cited and having that with a very concise executive summary. But then being able to jump off into that research has been really important too. We're very research driven. Obviously we have a lot of gut instinct, but we like to back it up with things that we've been documenting from customers. So I provide very well researched, comprehensive documentation, but also make it very concise to skim up front to understand the direction and then deep dive where she needs to.
- [22:46]
Daniel Murray
Yeah, I love the data angle. I think like you have to have actual data points to tell a good story. I think good stories are led with emotion and then backed by data, which I think the motion is your passion towards this, working the story you're telling, why it's going to work. But then you have like, hey, this trend says this, these customers say this, this, say this. So your boss has leverage. And I think one thing we're talking about right now is like the art of storytelling. So when it comes to, when it comes to AI, what kind of storytelling have you found most effective? Is it like case studies? Is it impact on a marketer? Is it the future vision? Like what is like something that is really helping with storytelling for AI?
- [23:43]
Kylie Pena
Yeah, I found that AI messaging resonates most when it's tied to a real human story or something that everybody can resonate with. So showing how someone saved time or was able to express A creative idea in a very specific way or unlocked a new idea. So with Generative Extend, it's a tool that is going to be used by editors who really care about the finesse of editing. But when you're bringing an AI tool that's Generative into a creative tool, it can feel a little bit threatening, like, oh, you're taking away something. So part of it was also influencing the trajectory of how the feature was created. From the beginning. I was part of the design Sprint saying this should only extend, like, up to two seconds, because that's not fundamentally rewriting a whole shot. It's not something that takes away a creative decision. It's about these little fixes that you would have spent a half hour on and now you can spend five minutes on instead. And showing that. I was at a trade show last week, and I was showing this tool in the context of a real situation that I would have faced when I was a video editor. I was showing it not just like, oh, this is the shiny demo, but rather, here's where I would use it in real life on these little things that really matter if you're a real video editor. And that kind of messaging of grounding it in these real use cases really, really resonates, especially with something that can feel up front a little bit like, oh, I don't know if I want to use this. You see where it is and how it works. You're like, oh, that is not threatening my livelihood. That's not taking anything away other than the stuff I don't like doing anyway. I also think the AR tools become more. The messaging becomes more powerful when you position the tools, like, in a way that the messaging flexes for different audiences. So this is certainly true with our assistive AI. The messaging that we did when we introduced a tool that automatically tags tracks in your timeline as the kind of audio it is like dialogue or music. We wrote that messaging so that it would say it's easier for beginners to understand what kind of tools to use on these things because you can click on it and it opens a panel automatically. But it's also just faster for pros to just reduce the number of clicks, the amount of mouse mileage that you're using, so you can get to the creative part sooner. So I think that's also a big component of positioning it for these different audiences, so people will look at it and say, oh, that is, in fact, for me.
- [26:39]
Daniel Murray
Yeah. I think one thing you also said that's very useful is I think you coming in from that I used to be a video editor and this is how it's making my. If I was doing this with AI, this is how I would. So you're coming from also experience driven too. So people trust you. Instead of saying, hey, this could do this for video editors, it's like, hey, if I was doing it, it's like show, don't tell instead of from a place of expertise.
- [27:10]
Kylie Pena
Yeah. And that's been important. I think that people know when they're being marketed to and they kind of can. I've learned to tune it out a little bit and I think that's marketing is also just being real with people. I as a. I was a video editor for 10 years and then I went into workflow technology and things like that. And now I work on the tools that people use in that ecosystem and I'm able to use my subject matter expertise to kind of position it as this is what I have done or this is what I would do. And I think just being real, authentic with the messaging not being fluffy. I hate fluffy marketing. I hate it when it's like really over the top. I just want to be real like this does this and that's why it's great. And then here, look at it. And if the feature is built well, if you've been involved in shaping it too, then it's going to be successful because the promise meets the actuality.
- [28:10]
Daniel Murray
When do you decide of embedding AI in the messaging versus just the experience? Because I think sometimes you can embed when you get into the product, hey, we have this. To do this, you don't necessarily have to say it up front. So when are you thinking about saying, hey, this is AI in your messaging versus like once they get into the product, it's a nice experience to figure out the AI.
- [28:40]
Kylie Pena
With generative AI, I have leaned more into being very upfront with that because I don't want anyone to feel like they were misinformed. Otherwise, it's been a little bit of a balance because there's certainly a lot of drive from everyone to want to say AI in your messaging so that when you're in the news cycle, everyone knows that you're doing AI. But I think really it comes down to if adding the AI component to the upfront messaging is important for people to know, it could be important for people to know that we are innovating and we're actually using AI and we're making the app modern and things like that. But there is also a level of saturation around AI and people are like, oh my gosh, again with this. So when I am writing messaging and positioning, I do lean into using it more often, but when I'm speaking with customers directly, I'm not necessarily positioning that in live demos or things. I'd much rather when I'm speaking just say, hey, you can do this and this and this, and then they will come to understand that. So, for example, we just launched something called Media Intelligence in Premiere Pro, which automatically identifies the visuals so that you can just use natural language to search for things like objects and locations and stuff like that. That's a brand new thing, obviously powered by AI, not generative. But when I'm talking to people, I just kind of don't say this is powered by AI, but rather this. It analyzes the visuals and they know they're like, oh yeah, that's AI. And then they ask the questions about it, like, is this looking at my content? Is it sending anything to the Internet? And then I can have that conversation where I'm like, no, it's all local. On device. It's an on device AI model. So I think it depends on the context. And in broad, broad marketing we are tending to do more AI as a signal to the ecosystem. But in the reality, I think a lot of customers know what they're getting into.
- [30:49]
Daniel Murray
One thing I wanted to also ask is what do you think is the biggest myth about gen AI you wish more marketers or product leaders understood?
- [31:02]
Kylie Pena
I think one of the biggest myths is that it's going to make everybody equal somehow. Like it's going to level the playing field in a way where your expertise will not matter anymore. And I think it's really not the case at all. I think more and more we're going to see small AI models like generative or assistive models pop up that people are able to create for themselves that are maybe trained on their own voice or their own material, and they're going to be able to use that to kind of scale up their own efforts. And I think that I hope that ethical AI that's trained on things that you have permission to train is more common as well, which will change the kind of landscape too. But I think if there are people who think that their marketing jobs or their creative jobs are going to be devalued because of AI, I don't think that's the case. Because your personality and your spark and your insight around what people want and how to give it to them is not something that generative AI is going to be able to do for you. And I think that you'll be using this tool to ideate or do the things that you don't really want to do. If I'm going to sit there and think about a bunch of messaging and positioning, I could sit down and just brainstorm 50 different tags, or I could use an AI tool like Copilot to iterate on that and all of those that I get. If I did that, I'm not going to get anything that's really spectacular, but it helps you open up a different pathway and go, oh, I'm going to go this direction or that direction. That's the real magic of this. Getting rid of the stuff that you don't want to do and then focusing all of your brain power on the really special stuff.
- [33:02]
Daniel Murray
What are some signals you're watching to stay ahead of how Genai will evolve for product and brand teams or marketing teams?
- [33:12]
Kylie Pena
I am in a lot of user forums. I think that just being close to where people are having conversations about this, I'm in a lot of forums that are related to my specific product in particular, because people, they're asking all kinds of questions about their workflows and their projects and I feel like that indicates to me what kinds of challenges and problems they're having. Like not just challenges and problems directly with the software, but rather broadly within the whole creative ecosystem. I also listen to a lot of a number of podcasts and just I think that it's really interesting because I want to be like, oh, I go to these particular blogs, but I'm getting the most insight and up to date information just from going to events where there are people that are in my position or use my software and I'm able to kind of triangulate their personal experiences into where I see things heading as well as just trying to stay ahead of the news. On something like Blue sky, it's a lot of different signals and it's very messy right now. And it's really just about how you sort of synthesize all of that into a story of like, where is AI going to lead us? What is the sentiment and what's the feeling? And what are people really wanting and what they want? Is that what they're telling you? Yet I think that's also an important component to kind of predict the future.
- [34:48]
Daniel Murray
The question I want to ask too, the last question and I asked everybody in this podcast, is what is a marketing hill you would die on?
- [34:59]
Kylie Pena
A marketing hill I would die on, is that PMMs need to be involved in the entire product development process. PMMs are not just I'm going to figure out where to put Facebook ads. PMMs are about, at least in my world, bringing user sentiment and real trends and business objectives all the way up front before you even write a line of code. And you're going to have a better outcome if you have that kind of cross functional alignment.
- [35:37]
Daniel Murray
Yeah, people always forget that PMM is like a very strategic role where I think that people mistake that they see everything because they're with the customer, they're with the product, they're with the users so they have a full picture of what's happening that they can spill over to the whole company, which I think that. And then also once they've got all those details now they can message it and create audiences off of that. But I think people forget that about PMMs and I think more people, I think PMM sometimes gets messed up because of the name more than what the role is. I think it's so broad of what people think PMMs do.
- [36:27]
Kylie Pena
PMMs are ultimately storytellers in every aspect, whether it's internal or external. So a big gap that I see sometimes is the ability for another team to be able to explain why things are or what they need. So leaning into PMM to help to tell that story is very important and very useful.
- [36:49]
Daniel Murray
Lastly, where could people find you, what you're doing? All that good stuff you can find.
- [36:55]
Kylie Pena
Me on across social media, Kylie or LinkedIn and my website Kylie pena.com and then of course the Adobe Premiere Pro blog.
- [37:07]
Daniel Murray
Well, thank you so much for coming on and I really appreciate it.
- [37:11]
Kylie Pena
Yeah, thanks for having me.
- [37:15]
Daniel Murray
Thanks so much for listening. Keep tuning in to hear more great insights from the coolest marketers from around the world. If you haven't already, make sure to subscribe and follow the Marketing Millennials podcast on Apple Podcasts, Spotify, YouTube or wherever you get your podcast us. And if you like what you hear, I would greatly appreciate you giving us a five star rating. It helps bring more marketers into our community.