---

Connect with Karina on LinkedIn: https://www.linkedin.com/in/karinatukanova/

About RocketShipHQ:
Newsletter: https://intelligentartifice.kit.com
rocketshiphq.com

FULL TRANSCRIPT BELOW

Shamanth: Welcome to the Mobile UA Show. Today’s episode is part of our series Intelligent Artifice, a deep dive on how AI is transforming performance marketing. Since 2018 and over 250 episodes, we’ve been deconstructing how the best performance marketers actually win. Every week I sit down with the operator shaping it, or we tear apart the ad systems behind the highest-performing advertisers in the world.

My guest today is Karina Tukanova. Karina is a freelance creative strategy consultant working with health, wellness, fitness, and lifestyle apps. She made the jump from performance marketing to creative strategy right around the iOS 14 shift, and she’s been operating at the intersection of UA and creative ever since.

I thought this would make for a great episode because a lot of people talk about audience research as a nice-to-have, especially post-AI. Karina’s argument is that it’s the big unlock, and the same product can speak to a completely different, untapped audience once you understand the language they actually use, not the audience you assumed you were building for.

She also talks through how you can do audience research without having back-to-back Zoom calls with your users all day long, how she uses AI as a second brain to store and surface creative insights, how AI spotted a hidden script formula across hundreds of UGC ads, and why she thinks handing 100% of ad production to AI produces slop.

For all these reasons, I’m excited to get into this episode with Karina. Let’s get into it.

I’m excited to welcome Karina Tukanova to Intelligent Artifice. Karina, welcome to the show. I’m excited to chat with you because you made a transition from UA to creative, and you’ve worked on some very high-scale subscription app products. You’ve done this for a number of years, from just before iOS 14 to afterward, so there’s a lot in your experience I’m excited to dive into.

Let’s start by talking about the iOS 14 shift. You said your big shift, career-wise, happened around iOS 14. Talk to me about how much of it was actually about iOS 14, and what the big shift was from a creative perspective.

Karina: Yeah, I think for me it’s kind of both a personal and professional question. Starting with the professional side of things: iOS 14 really shifted how creative is perceived in performance marketing.

Before, creative was really just a puzzle piece, something performance managers did on the side. When I started as a performance marketing manager, “creative manager” or “creative strategist” didn’t really exist as a role. iOS 14 shifted the perspective on how much creative can actually make a difference and allow you to reach specific audiences.

That’s when I started thinking, actually, I enjoy the research and ideation part of it a lot more. When I transitioned into the more creative part of the role, I think back then we called it “creative marketing manager,” but I was still working with UA very closely. So I think it was a combination of luck, right place right time, and also me always being more in touch with the creative side of things rather than just looking at the numbers.

Shamanth: And obviously part of it is perhaps luck, but part of it’s obviously you noticing what was shifting and embracing that opportunity. One of the shifts was that a lot of manual targeting is pretty much gone. So as a creative-focused marketer, what are some of the ways you look at reaching a very specific audience?

Karina: Yeah, I think that’s the piece where creative strategy, and specifically audience research, really comes into play. When I start with a new client, I really focus a lot on understanding who the target audience is, what the people using the product’s trigger words are, what language they use when they communicate day-to-day about their specific issues, and whether or not they’re problem- and solution-aware.

One interesting example came up quite recently, when I started working with a new client, a stretching app. It has quite broad appeal, right? We all know stretching is good for us, and not a lot of us do it. When I came onto the account, a lot of the top-performing content was focused on generic stretching benefits, on getting the splits, getting the backbend, something with a tangible result you can see.

It was attracting a lot of younger audience too, which wasn’t an issue, it was working, but they wanted to expand beyond that 25-to-35 target. So I looked at a lot of organic trends around where people use stretching in their life. What kind of people use stretching? Is it mobility, dynamic exercises? Is it for athletes? Is it for senior people?

What came out of it is that for senior people, it’s actually quite an interesting topic, and a lot of them try to incorporate stretching into their life. But it’s not about getting the crazy backbend or nailing it in 30 days. It’s about retaining their mobility, being agile, and still being able to enjoy their 50s, 60s, 70s because they’re able to move their body.

So I built creative concepts around that idea, much more long-form and explainer-based compared to the current top performers, which were flashy and fast-paced, appealing to the younger audience.

That’s how creative is able to speak to a specific audience: when it speaks their language, when the format and execution style resonates with that audience and looks like the content they’d actually consume. It really allows you to unlock a different audience you might not necessarily have targeted before.

Shamanth: Yeah, that’s a very interesting example, because a lot of product owners would think, “let’s go after the Gen Z TikTok crowd,” who also don’t have as much disposable income as older users, and the aesthetic of TikTok is very easy to replicate.

But from what you’re saying, you had to look at what older users really wanted, and you made longer-form content tailored to them. Can you talk about how you actually surface these insights, or even just found out that older users were a potential audience worth going after?

Before you answer, it’s a pretty common thing to say “go talk to your customers,” but in practice, I don’t know if a lot of people have the time and energy to set up thirty minutes per customer on their calendar. Is that something you do, or is there another way you find insights like the one you just described?

Karina: Yeah, I think it’s always great to talk to customers, and if you have a user research person on the team who can spend the time, that’s great. But it never gives you a big enough sample size to work at scale. So what I like to do is work with what I call the “big three:” TikTok, Instagram, and Reddit, for organic research, not just to find the content people see there, but to really understand what they’re talking about and what kind of creators they engage with.

It’s basically a lot of scrolling on these platforms, then lifting the messaging, the hooks, and the overall ideas they discuss into a customer file, alongside reviewing product insights, competitor ads, and competitor messaging, and putting it all together into what I call a big “voice of the customer” file.

One underrated element I use there is comments, whether it’s comments on your ads, comments on competitor ads, or comments under Instagram and TikTok posts. They’re a treasure trove of hooks, trigger words, and pain points; that’s where the actual discussion happens. It’s still a bit of a manual research process at this point, but it’s really worth digging into, at least when you start working on a new project, because you can lift so many good insights.

You can understand what people are worried about when it comes to your product and to competitor products, and really use that to your benefit.

Shamanth: Got it. So from what I understand, you’re maintaining a file or repository of all these insights and using them as needed. Just to follow up: what form does this take? Is this a Google Drive with all these testimonials and insights? Can you talk about what this looks like in practice?

Karina: In the last couple of years I’ve been relying more on AI to be that repository for me, almost like my second brain. A lot of this lives in markdown files in a Claude project, for example, and then it’s very easy to lift specific insights or that customer language and use it in the copywriting for hooks and scripts, to make sure the copy isn’t just generic marketing fluff but actually what customers respond to.

I’d say Claude Projects, or Gemini, with markdown files, are also very useful for analyzing videos and using those to inform your strategy.

Shamanth: For sure. So there are multiple files in a folder you reference as needed, customer insights, potential hooks, sounds like that’s how you categorize them.

Speaking of AI, what are specific ways you use it, and are there examples where AI spotted something in your creative data that you manually missed?

Karina: Yeah, as I mentioned, I usually put together a big project file that’s constantly fed with creative test data and any other relevant product context updates, so it stays up to date and useful. I also make sure it writes up any learnings and evergreen results that would be useful further on.

But when it comes to using AI, I think it’s really good at spotting patterns and lifting things you might miss, especially when you’re working with a lot of creatives and shipping hundreds of assets a week.

One example: I did an audit on a client where a lot of their content was UGC-based, and UGC content is very hard to analyze because there are so many variables. Different creators, different delivery styles. With that specific client, a lot of the UGC content was also very varied in terms of what topic they were discussing.

So manually, it was really hard to analyze why each creative worked. I could understand why a specific creative performed well, a specific hook or topic, but there were tons of creatives that were similar. After I analyzed it with AI, it helped me understand the overall framework these scripts were following.

As much as they were all different, they all shared a similar structure, which was roughly: a complaint, an app introduction, and then you show the solution. We then managed to lift that structure and use it for more varied scripts, and also adapt it to different execution formats, since one of the goals was to diversify the creative production and move away from just UGC formats while still retaining performance.

We used it for things like animation style, and adapted it a bit for statics too. It was a quite successful collaboration between AI and creative strategists.

Shamanth: Yeah, and you’re right, here’s a complaint, here’s the app introduction, here’s the ABC. If you have literally hundreds of ads, AI can really spot those patterns in ways that would be really hard for humans.

Speaking of collaboration, the other piece that’s critical is between creative and performance marketers. What have you noticed around the most successful team structures, specifically between UA and creative? What are some underappreciated aspects that could potentially break when they’re not really talking to each other?

Karina: Yeah, I think the answer should be one team, one goal. I’ve seen cases where UA and creative teams are split, and that really doesn’t work, especially when you’re both working on the performance side of things but reporting to different people, so you end up chasing different goals.

In my perfect world, a UA manager knows a bit about creative production, and the creative manager knows a bit about UA. I don’t think you as a creative manager need to know every technical aspect of campaign management or everything about attribution, but you do need to understand the key metrics. You need to understand why and how UA managers are making decisions, and equally, UA managers need to understand any reservations you might have around creative production capacity and limitations, and the fact that you can’t just go from 50 concepts one week to 150 the next just because the budget scaled.

So it’s essential that your creative lead and UA lead are talking to each other every week and are looped into each other’s plans, because it makes everybody’s life easier. If I know the UA team is planning to scale specific deals or increase budgets in the US, I know I need to increase production. But we need to have that discussion at least a couple of weeks before it happens, so we’re prepared on the creative side. And equally, they need to understand how we’re producing content, that we can’t just use certain UGC because we don’t have the rights for it, or if they have specific requests.

The insights from UA managers are invaluable for creative production, because we both look at the same data but have slightly different understanding of it. I’m a big fan of shared brainstorm sessions with the UA team, so everybody feels heard and can put their ideas on the table, and we can create the best possible ads from that.

Shamanth: Yeah, you’re right that it’s not always easy, because they’re reporting to multiple people. Creative reports to a different organizational structure than UA, different lead, and the real challenge is getting them to think and work as one team.

I know a number of companies have a pod structure, where you report to different people, but together you work as one, designers, creative strategists, and UA on specific products or projects, and that’s something I’ve seen work as well.

Switching gears, you talked about finding ideas earlier, working with AI. Before you spend anything on a concept, how do you decide what’s worth making? You could technically make hundreds of ads with AI. How do you decide what’s worth making?

Karina: Yeah, I think that’s a combination of the context I have from the client or from working on the app, whether we’re trying to unlock a certain audience or certain targeting, plus a combination of competitor research and a creative hunch, for lack of a better word.

I’m a big fan of making sure there’s always a variety of concepts in each batch. That means always having iterations, to support existing top-performing concepts and make sure campaigns can continue to perform while addressing creative fatigue before it sets in, but also some new concepts based on what competitors are doing and what we know is working, and then a smaller percentage reserved for out-of-the-box, throw-stuff-at-the-wall ideas.

I do this because a lot of companies run into issues where they find a top-performing concept, iterate on it, build all these formulas around how to make more of the same content, and then end up producing the same content over and over. It becomes very hard to move away from that framework. Eventually performance drops, and you’re stuck trying to move away from something that worked two years ago. So it’s very valuable to have out-of-the-box ideas in the mix, whether you’re producing them with AI or it’s just a generic ad you brief a designer on. It’s always valuable to keep in your production pipeline, to make sure you keep experimenting with execution styles and messaging.

Shamanth: What I’m hearing is: have a portfolio of ideas going in. And you’re right that there’s a risk of being pigeonholed into what’s working. I’ve also seen the opposite risk, where it’s “we want brand-new ideas every single week,” and there’s not enough building on what’s working. I’ve seen that far too often, but tell me if you haven’t.

Karina: Yeah, I’ve definitely seen it, especially last year with the Andromeda craze, when everybody was like, “okay, now we need to move from creative volume to creative diversity, and we need new ideas every week.”

I think that’s just not sustainable on a human level. As much as you have several great strategists on the team, there’s only so many ideas any person can come up with at any time. That asks for burnout, and if you move everything to AI, it becomes AI slop, and you’re not able to learn or iterate from anything, you’re just producing volume for the sake of it.

So it’s a fine balance, but that’s why I go with a mix of new concepts, iterations, and net-new ideas.

Shamanth: And speaking of Andromeda, what would you say to somebody coming to you and saying, “Hey Karina, Meta says here’s Andromeda, I want 100 completely unique, completely new concepts that have never been made by anybody else in the world, and I want this every single week?” How would you respond?

Karina: I’d say, “that’s crazy,” in a maybe more politically correct way. I’ve definitely seen clients try to do all-new concepts, or shift to completely new production, I think around last September, and honestly, I haven’t seen that work for anyone.

It’s still a mix of working off what you know works and iterating on that, and having new ideas. What I think shifted in the last year is having more diversity visually. Before, you could have the same concept with five different hooks that kind of looked the same, but certain messaging stood out and that’s how you knew it was a winner. Whereas now I tend to brief designers to make sure the visual hook at the beginning is slightly different, whether it’s using a different creator for UGC concepts, different text treatment, or different transitions and UI overlays, just so Meta doesn’t treat it all as the same concept and punish good messaging because it looks similar to a previous concept.

Shamanth: Yeah, and I also feel “punish” is a bit of a strong word, and I feel Meta reps are incentivized to propagate certain messages that don’t always translate to account performance. And you’re right, not to discount what you said about visual variety, visual variety has definitely become more prominent, but there’s a real danger in avoiding what’s working and just chasing completely new ideas.

As you pointed out, it’s kind of crazy and not sustainable, and you really need to rein in the people who say “let’s just do completely new stuff,” because what’s proven is actually working. There’s no reason to throw that away.

Karina: Yeah, and you can always try stuff that’s not new to your product, even looking at competitors, especially if they’re not a direct competitor. You can adapt ideas from another app outside your vertical, and it’ll be new to you, and it might work well. But you’re wasting opportunities if you’re seeing something work for others and not using that to your advantage, because if it works for somebody else, it might work for you. Not a hundred percent, but it’s an easier adaptation than coming up with a completely new idea.

Shamanth: For sure. And if it’s working for yourself, you should absolutely build on that, make a variation on it. And speaking of AI, how do you think about having AI actually produce the ads? Do you let it rip, run hundreds of ads and just run whatever it gives you, or is it more like having a human review everything, or a human do everything? Where on that spectrum do you lie, and why?

Karina: I think I’m a bit more old school when it comes to AI creative production. As I mentioned, I’m a big fan of having AI act as my second brain and help with research and learnings, but I haven’t seen cases yet where fully AI production works completely. I think there’s still some human involvement necessary to make that AI workflow work for your team and product, and I haven’t seen the magic tool where you post a link to your product and it makes perfectly ready ads.

Whether it’s creative strategy, copywriting, or a designer editing elements of the video, I think it still requires human involvement. But when it comes to statics, that’s where AI is great, you can save a lot of time and energy, especially if you want to bulk-test copy and messaging with some static templates or popular social media templates. That’s when you can get out 100 statics in a few minutes and test them, and it really speeds up the workflow.

Shamanth: For sure, and even for that, I’d imagine you still need human research to figure out what those 100 ads need to look like.

Karina: Yeah, exactly. It still requires input, in terms of the hooks you want. AI can create some hooks for you, but you still need somebody with an eye for what works, what people resonate with, and what sounds natural, to curate that AI collection and run those ads.

Shamanth: Yeah, and we talked about creative and UA being very different skill sets, and to be effective you need to bridge that gap. For people trying to bridge that gap, creative people who want to read the data, performance people who want to make ads, where would you suggest they begin?

Karina: I think by talking to people who are already doing those roles, whether it’s a creative strategist going to their UA manager and saying, “hey, can you explain how you look at the data?” Or somebody in a similar role you’re trying to move into. Creative strategy is especially one of those jobs where you learn on the go, and you can listen to a lot of podcasts and read on LinkedIn, there’s always useful information there, but a lot of the skills and understanding of the job comes from actually doing it.

As I mentioned, when I started I was a performance marketing manager, I didn’t really know much about creative strategy or how it was going to pan out, so I just did it with different clients, trying to understand how it worked for different products. A very valuable skill to have on the job is being open-minded and flexible, because you work with different apps, there are tons of different products people are using, and you need to be flexible in adapting to how you work with people. Performance marketing also shifts very quickly, so don’t get too attached to an idea, a messaging, or an angle, because one day you wake up and performance drops, and suddenly you need to find something else. But that’s also what makes it so fun.

Shamanth: Yeah, indeed. Because it’s easy to look at something on LinkedIn or a podcast, this is a podcast too, but you’d still want to… I don’t think a podcast alone can capture the learnings from thousands of ads. So you’d use a podcast or LinkedIn as a starting point if you want, but really learn from who you’re working with. That’s what you’re suggesting, yeah?

Karina: Yeah, exactly.

Shamanth: Excellent, those are wise words. I know we’ve covered a lot of ground today, Karina, and this is probably a good time for us to wrap up. But before we do, can you tell folks what you do and how they can find out more about you?

Karina: Yes, so I currently work as a freelance creative strategy consultant, working with health and wellness, fitness, and lifestyle apps. You can find me on LinkedIn, and if I accidentally miss your message there, you can email me, my email is also on my LinkedIn. I’m looking forward to working with anyone or chatting with any potential new creative strategist if they want advice or just want to brainstorm something together.

Shamanth: Wonderful, and we’ll link to all of that in the show notes. Thank you so much for your time.

Karina: Thank you for having me.